Cooper v. Brown ( 2009 )


Menu:
  •                                                 Volume 1 of 2
    FOR PUBLICATION
    UNITED STATES COURT OF APPEALS
    FOR THE NINTH CIRCUIT
    KEVIN COOPER,                              No. 05-99004
    Petitioner-Appellant,            D.C. No.
    v.                         CV-04-00656-H
    JILL BROWN, California State             Southern District of
    Prison at San Quentin,                        California,
    Respondent-Appellee.            San Diego
            ORDER
    Filed May 11, 2009
    Before: Pamela Ann Rymer, M. Margaret McKeown, and
    Ronald M. Gould, Circuit Judges.
    Order;
    Dissent by Judge W. Fletcher;
    Dissent by Judge Wardlaw;
    Dissent by Judge Fisher;
    Dissent by Judge Reinhardt;
    Concurrence by Judge Rymer
    ORDER
    The panel has voted to deny the Petition for Rehearing and
    Petition for Rehearing En Banc.
    The full court was advised of the petition for rehearing en
    banc. A judge requested a vote on whether to hear the matter
    en banc. The matter failed to receive a majority of the votes
    5429
    5430                   COOPER v. BROWN
    of the nonrecused active judges in favor of en banc consider-
    ation. Fed. R. App. P. 35.
    The petition for rehearing and the petition for rehearing en
    banc are DENIED.
    W. FLETCHER, Circuit Judge, dissenting from denial of
    rehearing en banc, joined by PREGERSON, REINHARDT,
    PAEZ, and RAWLINSON, Circuit Judges:
    The State of California may be about to execute an inno-
    cent man.
    From the time of his initial arrest until today, Kevin Cooper
    has consistently maintained his innocence of the murders for
    which he has been convicted. Cooper was convicted of capital
    murder and sentenced to death by a California court in 1985.
    The California Supreme Court affirmed Cooper’s conviction
    and sentence in 1991. People v. Cooper, 
    53 Cal. 3d 771
    (1991). The California Supreme Court denied Cooper’s state
    petition for habeas corpus in 1996. A three-judge panel of the
    Ninth Circuit affirmed the denial of Cooper’s first federal
    application for habeas corpus in 2001. Cooper v. Calderon,
    
    255 F.3d 1104
     (9th Cir. 2001). That decision was called en
    banc, but the call failed.
    In 2004, on the eve of his scheduled execution, Cooper
    sought permission from the three-judge Ninth Circuit panel to
    file a second or successive application for federal habeas cor-
    pus under 
    28 U.S.C. § 2244
    (b)(3)(A). Among other things,
    Cooper claimed that he had new and previously unavailable
    evidence that the State had violated Brady v. Maryland, 
    373 U.S. 83
     (1963). Brady requires the State to turn over exculpa-
    tory information to a criminal defendant. Based on the
    claimed Brady violation, Cooper claimed actual innocence
    under Schlup v. Delo, 
    513 U.S. 298
     (1995), and
    COOPER v. BROWN                      5431
    § 2244(b)(2)(B)(ii). The three-judge panel denied permission,
    but an en banc panel of the Ninth Circuit reversed. Cooper v.
    Woodford, 
    358 F.3d 1117
     (9th Cir. 2004) (en banc). We
    stayed Cooper’s then-pending execution until his new federal
    habeas application could be addressed.
    Two days before the murders, Cooper had escaped from the
    minimum security section of a nearby California state prison
    by walking across an open field. Shortly before Cooper’s
    scheduled execution date, Midge Carroll, the now-retired war-
    den of the prison, provided a sworn declaration in which she
    stated that she had learned from her staff that shoes issued to
    prisoners “were not prison manufactured or specially
    designed prison-issue shoes,” but, rather, were “common ten-
    nis shoes available to the general public through Sears and
    Roebuck and other such retail stores.” Carroll stated that she
    had learned this information during the investigation and con-
    veyed it to investigators before the trial. This information
    would have been critical to Cooper’s defense, for a key to the
    prosecution’s case at trial was that identical shoeprints at the
    crime scene and in the house where Cooper had been staying
    were made by Pro-Ked “Dude” tennis shoes, and that these
    shoes were distributed only to prisons and other institutions.
    Warden Carroll’s information, though clearly exculpatory,
    had not been provided to Cooper prior to trial.
    In granting permission to file a second or successive appli-
    cation, the en banc panel noted, “Once a Brady violation has
    been established, a federal habeas court is required to evaluate
    all information in the case, not just information relevant to the
    Brady violation.” Cooper, 
    358 F.3d at 1122
    . That is, every-
    thing in the new habeas application was properly before the
    district court. Woratzeck v. Stewart, 
    118 F.3d 648
    , 650 (9th
    Cir. 1997) (per curiam) (“If [a petitioner’s] application makes
    a prima facie showing as to one of the claims, he may proceed
    upon his entire application in the district court.”). In addition
    to this general instruction to the district court, we specifically
    directed that two tests be performed. We wrote, “As soon as
    5432                       COOPER v. BROWN
    Cooper’s application is filed, [the district court] should
    promptly order that these two tests be performed in order to
    evaluate Cooper’s claim of innocence.” Cooper, 
    358 F.3d at 1122
    .
    First, we directed the district court to conduct further test-
    ing of a bloody tan t-shirt that had been found beside the road
    leading away from the house where the murders took place.
    The tan t-shirt was found soon after the murders. Initial test-
    ing of stains on the t-shirt showed that they contained blood
    consistent with one of the victims and not consistent with
    Cooper. Cooper presented evidence of the t-shirt as part of his
    defense at trial. 12/6/84 RT 4602-06, 4608; 1/15/85 RT 6508-
    11.1
    Long after trial, at Cooper’s insistence, the State performed
    a DNA test on some of the blood on the t-shirt. Cooper main-
    tained that the test would prove his innocence. Instead, the
    blood tested positive for Cooper’s DNA. Cooper maintained
    (and continues to maintain) that his blood was planted on the
    t-shirt. If the blood was planted, the only possible source was
    blood taken from Cooper by law enforcement authorities. A
    vial of blood was taken from Cooper by San Bernardino
    County Sheriff’s Department (SBCSD) personnel on August
    1, 1983, two days after his arrest. That blood contained an
    added preservative called EDTA. We wrote, “The presence of
    such a preservative would show that [Cooper’s] blood was not
    on the t-shirt at the time of the killings[.]” Cooper, 
    358 F.3d at 1124
    . We directed the district court to test Cooper’s blood
    on the t-shirt for the presence of EDTA.
    Second, Jessica Ryen, one of the murder victims, was
    1
    Citations to the reporter’s transcripts of the trial and hearings appear
    with the date of the proceeding, the letters “RT,” followed by the page
    number. Citations to the Excerpts of Record, compiled for purposes of this
    appeal, use the abbreviation “ER” followed by the page number. “FER”
    refers to the Further Excerpts of Record.
    COOPER v. BROWN                             5433
    clutching blond or light brown hair in her hand. We directed
    the district court to subject the hair to mitochondrial DNA
    testing.
    The district court held hearings on Cooper’s application for
    habeas corpus in 2004 and 2005. It denied all relief. Cooper
    v. Brown (“Dist. Ct.”), No. 04-656 (S.D. Cal. May 27, 2005).2
    A three-judge panel of our court affirmed, with one judge
    concurring specially. Cooper v. Brown, 
    510 F.3d 870
     (9th Cir.
    2007).
    There is no way to say this politely. The district court failed
    to provide Cooper a fair hearing and flouted our direction to
    perform the two tests.
    As will be described in greater detail below, the district
    court impeded and obstructed Cooper’s attorneys at every turn
    as they sought to develop the record. The court imposed
    unreasonable conditions on the testing the en banc court
    directed; refused discovery that should have been available as
    a matter of course; limited testimony that should not have
    been limited; and found facts unreasonably, based on a trun-
    cated and distorted record.
    The most egregious, but by no means the only, example is
    the testing of Cooper’s blood on the t-shirt for the presence
    of EDTA. As will be described in greater detail below, the
    district court so interfered with the design of the testing proto-
    col that one of Cooper’s scientific experts refused to partici-
    pate in the testing. The district court allowed the state-
    designated representative to help choose the samples to be
    tested from the t-shirt. The court refused to allow Cooper’s
    scientific experts to participate in the choice of samples.
    2
    The district court’s order in this case is published as an appendix to the
    panel’s opinion, Cooper v. Brown, 
    510 F.3d 870
    , 887-1004 (9th Cir.
    2007). I refer to the pagination of the district court’s order as it appears
    in that appendix.
    5434                   COOPER v. BROWN
    Indeed, the court refused to allow Cooper’s experts even to
    see the t-shirt. The state-designated lab obtained a test result
    showing an extremely high level of EDTA in the sample that
    was supposed to contain Cooper’s blood. If that test result
    was valid, it showed that Cooper’s blood had been planted on
    the t-shirt, just as Cooper has maintained.
    A careful analysis of the evidence before the district court
    strongly suggests that the result obtained by the state-
    designated lab was valid. However, the court allowed the
    state-designated lab to withdraw the test result on the ground
    of claimed contamination in the lab. The court refused to
    allow any inquiry into the alleged contamination. The court
    refused to allow Cooper’s experts to review the bench notes
    of the state-designated lab. The court then refused to allow
    further testing of the t-shirt, even though such testing was fea-
    sible.
    The district court placed two photographs of the murder
    victims at the end of its 159-page order denying relief to Coo-
    per. One is a photograph of the photogenic Ryen family —
    two beautiful children, ten-year-old brown-haired Jessica and
    eight-year-old blond-haired Josh, and their attractive parents.
    The other is a photograph of eleven-year-old Chris Hughes,
    a handsome blond-haired boy. The district court had no ana-
    lytic reason to include these photographs at the end of its
    order.
    I.   Background
    Late at night on June 4, 1983, Doug and Peggy Ryen, their
    daughter Jessica, and their houseguest Chris Hughes, were
    brutally murdered in the Ryen home in Chino Hills, Califor-
    nia. Their son, eight-year-old Josh, suffered extensive injuries
    but survived. The victims had numerous chopping, cutting
    and stabbing injuries, caused by several different kinds of
    weapons. A number of cutting and stabbing wounds were
    inflicted after the victims were already dead. Josh was found
    COOPER v. BROWN                    5435
    in the late morning of June 5 in his parents’ bedroom, near the
    bodies of his mother and Chris. His father’s body was also in
    the bedroom. His sister’s body was in an adjoining hallway.
    Two days earlier, Kevin Cooper, a black man, had escaped
    from the minimum security section of the nearby California
    Institute for Men (CIM) at Chino by walking across an open
    field. He had been serving a four-year sentence for burglary.
    Cooper hid out for those two days in a vacant house owned
    by Larry Lease (the “Lease house”), 125 yards from the Ryen
    house. The State’s theory, at trial and now, is that Cooper
    acted alone in killing the four victims and in nearly killing
    Josh, and that Cooper committed the murders to facilitate his
    escape.
    Cooper made several telephone calls from the Lease house,
    seeking money and other assistance from two women friends.
    The women refused to provide money or other help. Money
    in plain view on the kitchen counter of the Ryen house was
    untouched. Credit cards and money in Peggy Ryen’s purse
    were also untouched.
    The last of Cooper’s telephone calls from the Lease house
    was completed at 8:30 p.m. on June 4, the night of the mur-
    ders. Cooper testified that he waited until dark, finished his
    telephone call, left the house on foot immediately after the
    call, and hitchhiked to Mexico. There is undisputed evidence
    that Cooper registered at a hotel in Tijuana at 4:30 p.m. on
    June 5. When Cooper checked into that hotel, his hair was
    braided, as it had been when he escaped from prison. ER
    3252, 3260. The Ryens’ white wood-paneled station wagon
    had been parked in the driveway of the Ryen house, with the
    keys in the ignition, throughout the entire evening of June 4.
    On June 11, the Ryens’ station wagon was discovered in the
    parking lot of a Long Beach church, where it had been for
    several days. Tijuana is 125 miles south of the Ryen house.
    Long Beach is 45 miles west of the Ryen house.
    5436                   COOPER v. BROWN
    A.   Evidence Pointing to Other Killers
    There is substantial evidence that three white men, rather
    than Cooper, were the killers. Some of the evidence was intro-
    duced at trial. Some of the evidence, even though exculpatory,
    was deliberately destroyed by the SBCSD and was therefore
    not available for use at trial. Some of the evidence, even
    though exculpatory, was concealed from Cooper and therefore
    not available for use at trial.
    Josh Ryen, the only survivor of the attack, first communi-
    cated to SBCSD Deputy Sharp that the murderers were three
    white men. ER 1278-79, 1386-87. This statement was the
    likely source of an entry in the police log during the afternoon
    of June 5, stating that the suspects were “three young males”
    driving the Ryens’ white station wagon. ER 3688.
    On the night of the murders, a couple drove to Chino Hills
    from Los Alamitos (about 37 miles away) after the night
    horse races to drop off a horse trainer. As they were leaving
    the trainer’s driveway, they had to wait for a car driving rap-
    idly down Carbon Canyon Road from the direction of the
    Ryen house. The husband, who was driving, testified at trial
    that he stopped to wait for the car to pass. He stated that the
    car was a “light color” station wagon with a luggage rack. ER
    1375-76. With some coaxing from Cooper’s attorney, he
    admitted that shortly after the episode he had described the
    driver to SBCSD Lieutenant Knadler as a “young white
    male.” The wife, who was a passenger, described the car as
    “tan” or “cream” colored, with “wood grain” paneling. She
    thought the car had a luggage rack. She “remembered think-
    ing” that she saw three or four people in the car. ER 1379.
    The Ryens’ station wagon was white with wood paneling and
    a luggage rack.
    The injuries to the victims were consistent with the use of
    multiple weapons. The number of victims, and the number
    and nature of the wounds, led the coroner initially to conclude
    COOPER v. BROWN                    5437
    that there was more than one killer. ER 1367-69; see also ER
    3148-56.
    When Josh was in the hospital after the murders, he twice
    saw a picture of Cooper on television. Both times he indicated
    that Cooper was not one of the killers.
    On June 7, SBCSD Deputy Field recovered a blood-stained
    tan, medium-size, Fruit of the Loom t-shirt with a front pocket
    beside Peyton Road, not far from the Ryen house and the
    Canyon Corral Bar. ER 1575. The characteristics of the blood
    on the t-shirt were consistent with Doug Ryen’s profile and
    inconsistent with Cooper’s profile. ER 1757-61; 1/15/85 RT
    6508-13. Post-trial DNA testing, conducted at Cooper’s insis-
    tence, revealed that the shirt contained DNA consistent with
    Doug Ryen and possibly Peggy Ryen. ER 800. Post-trial
    DNA testing also revealed that at least one blood stain on the
    t-shirt contained Cooper’s DNA. I discuss the presence of
    Cooper’s DNA on the t-shirt in greater detail below.
    As will be described in more detail below, a woman named
    Diana Roper provided a statement in 1983 and again in 1998
    describing an identical tan Fruit of the Loom t-shirt with a
    front pocket that she had purchased for her then-boyfriend,
    Lee Furrow. On both occasions, Roper stated that Furrow was
    wearing the t-shirt on the evening of June 4. ER 1571, 1573.
    At the time of the murders, Lee Furrow was living with
    Roper in Mentone, approximately 45 miles east of Chino
    Hills. ER 1570, 1592, 3142; 10/22/84 RT 2266. Furrow had
    previously been convicted of strangling Mary Sue Kitts in
    1974, on the orders of Clarence Ray Allen. ER 1594, 1789-
    90, 1802; 10/22/84 RT 2266. Furrow had been a member of
    the “Allen gang,” and he testified for the prosecution at
    Allen’s capital murder trial. See Allen v. Woodford, 
    395 F.3d 979
    , 986, 991 (9th Cir. 2005); People v. Allen, 
    42 Cal. 3d 1222
    , 1236-38 (1987). Allen was executed in 2006. In return
    5438                   COOPER v. BROWN
    for his testimony, Furrow was allowed to plead guilty to sec-
    ond degree murder. He served four and a half years in prison.
    Furrow was released from prison on June 12, 1982, a year
    before the Ryen-Hughes murders. In a 1998 sworn affidavit,
    Roper stated that Furrow had “confided” in her that he “dis-
    membered [Kitts’] body and threw the body parts in the Kern
    River.” ER 1570. Kenneth Koon, another acquaintance of
    Furrow’s, confirmed that he learned that Furrow “killed a girl,
    cut her up, threw her in the Kern River.” Kenneth Koon Inter-
    view, Apr. 7, 2004, Docket No. 31, Ex. 55, at 4.
    On June 4, the day of the murders, Furrow, Roper, Roper’s
    sister Karee Kellison, and neighbors Michael and Rebecca
    Darnell, attended the U.S. Festival Country Show at the Glen
    Helen Amphitheater approximately 30 miles northeast of
    Chino Hills and 24 miles northwest of Mentone. See Answer,
    Docket No. 30, Ex. 36; see also ER 1786. That afternoon,
    Furrow picked up Michael Darnell from the county jail, where
    he had been held overnight. Answer, Docket No. 30, Ex. 39.
    ER 1786. According to Darnell, he and Furrow arrived at the
    festival after dark. Answer, Docket No. 30, Ex. 40 at 9, 16.
    Debbie Glasgow, who was having an affair with Furrow, told
    police that Furrow arrived at the festival at approximately 11
    p.m. Answer, Docket No. 31, Ex. 53. Rebecca Darnell
    recalled that she left the festival with her husband, and with
    Roper and Roper’s sister, but that Furrow did not leave with
    them. Answer, Docket No. 31, Ex. 41 at 1.
    In her 1998 affidavit, Roper stated:
    On Saturday afternoon June 4, 1983, Lee [Furrow]
    and I were at our home getting ready to go to the US
    Festival. . . . He was wearing a medium size Fruit of
    the Loom beige T-shirt which had a pocket on the
    front. I bought Lee the T-shirt at K-Mart before June
    4, 1983. I also recall telling the sheriff’s department
    COOPER v. BROWN                         5439
    fifteen years ago virtually the same thing regarding
    Lee’s T-shirt. . . .
    During the early morning hours of June 5, 1983,
    my sister Karee Kellison, and I had returned from
    the US Festival and were in the living room at my
    house in Mentone. I received a telephone call from
    Lee about 1:30 a.m. to 2:00 a.m. in the morning of
    June 5, 1983, asking me to come get him and Debbie
    Glasgow at the US Festival. I refused and hung up
    the telephone. A few hours later a car pulled into our
    driveway. My sister, Karee, walked to the French
    doors to look out. I did not look out. A few moments
    later Lee and Debbie walked through the front door.
    They were in a hurry. I heard the car depart. Lee was
    wearing long sleeve coveralls which had a zipper in
    the front. The coveralls were splattered with blood
    and there was horse hair and dried horse sweat on
    the lower leg area. He did not have the beige T-shirt
    or Levis on that he was wearing earlier in the day.
    Lee walked to the back of the house and seemed to
    be in a hurry. Lee took the coveralls off and left
    them on the floor of the closet. After he changed his
    clothes, Lee and Debbie left immediately on his
    motorcycle. Lee and Debbie were not in the house
    for more than five minutes and were obviously in a
    big hurry to leave.
    After I learned of the Ryen/Hughes murders in
    Chino I turned the coveralls over to the San Bernar-
    dino County Sheriff’s Department. I told the deputy
    the facts about how I found the coveralls and that
    Lee Furrow may be the murderer.
    A few days after the murders        I saw Lee at my
    neighbors house. . . . I told Lee I   turned the bloody
    coveralls over to the sheriff’s        department. Lee
    became furious that I had turned      them in. Lee had
    5440                  COOPER v. BROWN
    changed his appearance by cutting most of his hair
    off and trimmed his sideburns and his “Fu-Manchu”
    moustache. . . .
    Also, a few days after the murders I heard on the
    news that a hatchet was found near the crime scene
    in Chino. I immediately walked to the washer area
    of our house. Lee’s hatchet was missing. All of his
    other tools were still hanging on the wall. [Referring
    to a photograph of the hatchet found near the crime
    scene]: The hatchet in this picture looks like the
    hatchet . . . which I found missing after the Ryen/
    Hughes murders. I cannot say for sure it is the same
    hatchet that Lee owned but the curvature of the han-
    dle is the same. Even more striking in similarity than
    the curvature of the handle is the style of the handle,
    which has sort of an American Indian pattern to it.
    [Referring to a photograph of the bloody tan Fruit
    of the Loom t-shirt recovered near the Ryen house]:
    The T-shirt in this photograph looks exactly like the
    T-shirt Lee was wearing on June 4, 1983 including
    the manufacturer, the size, the color and the pocket.
    I am absolutely positive the photograph of this T-
    shirt matches the T-shirt that Lee was wearing at our
    house on the afternoon of June 4, 1983.
    ER 1571-73. The substance of Roper’s 1998 affidavit matches
    the substance of a May 1984 interview she gave to SBCSD
    Detective Stalnaker. In the interview, Roper told Detective
    Stalnaker that she had been under the influence of drugs on
    the night of June 4. ER 1588-97.
    In a separate November 1998 sworn affidavit, Roper’s sis-
    ter Karee Kellison confirmed many of the details of Roper’s
    story. She added a detail about the car in which Furrow and
    Glasgow arrived. She stated:
    COOPER v. BROWN                         5441
    . . . During the early morning hours of June 5,
    1983, [Diana and I] were in the living room talking
    when we heard a car pull in the driveway. I would
    estimate the time after 3:00 a.m. but before sunrise.
    I looked out the window in the French doors and I
    saw Lee and Debbie get out of a car. There was not
    sufficient light to identify who the other occupants in
    the car were. However, there was enough light to see
    that it was a station wagon, kind of brown in color.
    ...
    Lee and Debbie came in the house. Lee was wear-
    ing long sleeve coveralls, which were splattered with
    blood. Lee walked to the back of the house and
    changed his clothes. When he came out of the bed-
    room he was no longer wearing the coveralls and
    apparently left them in the back of the house. Lee
    and Debbie left immediately on his motorcycle. . . .
    A week or so after the Ryen/Hughes murders I
    was interviewed by the police. I never told them the
    above information because I am terrified of Lee Fur-
    row and Debbie Glasgow. I don’t recall what I told
    the police but I know that I did not tell them what I
    saw for fear of what Lee and Debbie would do to
    me. While I am still scared to death of Lee, I under-
    stand he is in Pennsylvania and Debbie is dead.
    ER 3116-17.
    On June 9, Roper examined the coveralls that Furrow had
    left at her house on the night of the murders. She called her
    father, who then called the SBCSD. ER 1591, 4895. Accord-
    ing to SBCSD Deputy Eckley, who was dispatched to Roper’s
    house, the coveralls were “[h]eavily splattered” with blood.
    ER 1578. Deputy Eckley took the coveralls into evidence and
    gave a report to his supervisor. ER 1582. Deputy Eckley’s
    report, dated June 10, 1983, stated that Roper “suspects that
    5442                   COOPER v. BROWN
    the bloody coveralls are from the Chino murders and has fur-
    ther information regarding that incident and/or possible
    suspect/s.” ER 3105. As will be recounted in more detail
    below, Deputy Eckley discarded the coveralls in a dumpster
    during Cooper’s preliminary hearing in his capital case. Dep-
    uty Eckley claimed at trial that he acted alone in destruction
    of the bloody coveralls. However, the initials “KS” on the
    Disposition Report for the coveralls indicate that Senior
    SBCSD Deputy Ken Schneckengast approved the destruction.
    The State did not provide the Disposition Report to Cooper’s
    attorneys until long after trial. There is no evidence that any-
    one in the SBCSD ever performed tests on the blood on the
    coveralls.
    Deputy Eckley had had prior experience with Diana Roper
    and Karee Kellison, as well as with their father, Bill Kellison.
    Deputy Eckley’s prior experience caused him to take seri-
    ously the coveralls and Roper’s story. In an audiotaped inter-
    view with Cooper’s investigator, Deputy Eckley stated,
    “[W]ith my relationship with the Kellison’s/Roper family I
    know their involvement in crime, as far as committing mur-
    ders as well as giving up murderers. . . . [T]hey’ve given very
    good information on a murder before.” ER 5000-01.
    Roper and Furrow separated immediately after the murders.
    Roper then began living with Kenneth Koon. ER 1592-93. In
    November 1984, Koon was incarcerated at the California
    Medical Facility at Vacaville. According to Anthony Wisely,
    another inmate at Vacaville, Koon confided in Wisely while
    they were in lockdown smoking marijuana. SBCSD Detective
    Woods interviewed Wisely on December 19, 1984. Detective
    Woods stated in his report that Wisely was in prison “for a
    two-year period of time under the psychotic and remission
    program.” ER 1601. Part of Wisely’s story directly recounts
    the murders. Another part of the story matches Roper’s story
    if Furrow’s name is substituted for Koon’s. According to
    Detective Woods’ report:
    COOPER v. BROWN                        5443
    [Wisely said that Koon told him] that he was with
    two other guys that were in the BRAND or Arian
    [sic] Brotherhood and they [had] driven to the Chino
    area to collect a debt. He also stated that they had
    driven to a residence in Chino and that the two guys
    got out and that they were in for about ten or fifteen
    minutes and that one of the guys was carrying two
    axes or hatchets. That he also had gloves on, and that
    one of them made the statement that the debt was
    officially collected and that the first guy that came
    out turned around and said who was that, and then
    again stated “Who the fuck is the nigger?” He said
    that the man that made the statement was looking in
    the direction of the window and he saw a black sub-
    ject through the window and the one subject told him
    to get out of there. He states that KOON was
    dropped off in San Bernardino somewhere, he does
    not know where. He stated that KOON went to his
    old lady’s house and changed his overalls and that
    KOON also made the statement that one of the guys
    that came out with the axes was very upset because
    they apparently had left one kid alive.
    ....
    He stated that these persons that apparently col-
    lected the debt at the Ryen residence are debt collec-
    tors for the BRAND, the Arian [sic] Brotherhood,
    and that they have been to the pen for murder before
    and it also includes or involves someone that is cur-
    rently on death row.
    He also stated that KOON thinks that they hit the
    wrong house for the collection and that after all this
    blows over that the BRAND will take care of busi-
    ness in the right way.
    5444                   COOPER v. BROWN
    I asked WISELY about any details when they saw
    the black subject there in Chino. He stated he could
    not tell me any other details.
    ....
    . . . I asked [Wisely] if the female’s name, which
    KOON referred to as his old lady, was DIANNA
    [sic] ROPER. He said that it was.
    ER 1599-1600.
    Detective Woods interviewed Koon immediately after his
    interview with Wisely. Koon confirmed Wisely’s account
    about the coveralls. However, he clarified that it had been
    Furrow who had changed his coveralls at Roper’s house.
    Woods wrote:
    I asked [Koon] if he knew a subject by the name of
    DIANNA [sic] ROPER. He stated that he did. . . . I
    asked him at that time does he recall an incident
    when she turned over some coveralls to Yucaipa
    authorities. He stated that he did. He stated the best
    of his recollection was that she found some bloody
    coveralls in a house apparently belonging to LEE
    FARRELL [sic] and this was directly after the mur-
    ders were discovered in the Chino Hills. He stated
    that apparently the cops destroyed them or lost the
    coveralls and nothing else was ever heard of it.
    In the interview with Detective Woods, Koon provided him-
    self an alibi. He claimed that he had not gone to the music fes-
    tival with Furrow and Roper. Woods wrote, “He stated that he
    was aware of that particular weekend with the country music
    but, however, he was in the Gorman, CA area, returning back
    to the San Bernardino area shortly thereafter.” Koon told
    Detective Woods that he was not affiliated with the “Arian
    COOPER v. BROWN                        5445
    Brotherhood.” He refused to answer further questions “about
    that Arian Brotherhood situation.” ER 3114.
    Even though Woods interviewed Wisely and Koon on
    December 19, the State did not provide the information to
    Cooper’s attorney until the morning of January 2, the day
    Cooper was scheduled to take the stand at his capital trial.
    Cooper’s investigator interviewed Wisely ten days later, on
    January 12. According to the investigator, Wisely was
    “wary.” He was initially unwilling to talk to the investigator,
    but finally said a little. The investigator reported:
    Wisely stated that the contact by the Det[ective] was
    the beginning of his troubles, that he has been “in the
    hole” ever since. He reported that he had not com-
    mitted a violation, and he is in the hole without
    cause. He further stated that the state is worried
    about him[.] . . . He then stated that he knew that
    Kevin Cooper did not do it, that “if you had been
    there & listened to him (Koon), you would know
    who did it” (or something close to those exact
    words).
    ER 3119.
    In 2004, when the district court was making its initial plans
    for mitochondrial DNA testing of the hairs in Jessica’s hand,
    the California Deputy Attorney General informed the court
    that Furrow had retained an attorney. Through his attorney,
    Furrow declined to provide a hair sample. 6/2/04 RT 5. The
    Deputy Attorney General informed the court that Koon, on
    the other hand, “indicated he will provide whatever the Court
    would request, voluntarily.” 
    Id. at 5
    . According to Koon’s
    story, as recounted to Wisely and recorded by Detective
    Woods, Koon stayed in the car while the other two men went
    inside with their axes. If this is true, Koon would of course
    have nothing to fear from providing a hair sample.
    5446                    COOPER v. BROWN
    On the night of June 4, three men came into the Canyon
    Corral Bar, which is located near the Ryen house. Several
    employees testified at Cooper’s trial that the men came into
    the bar sometime around 8 or 9 p.m., left, and returned later.
    Witnesses’ estimates of the time of their return range between
    11 p.m. and 1:00 a.m. See ER 3648. When the men returned,
    at least one of them was extremely inebriated. The bouncer,
    Ralph Land, did not testify at trial, but stated in a tape-
    recorded interview with Cooper’s investigator in January
    1984 that “two of them were really close-cut hairs and all that
    and then the other one had long, straggly like dirty, like I
    thought maybe they just a couple of Marines and they ran into
    an old buddy or something.” ER 3616. The witnesses who tes-
    tified at trial stated that all three of the men had close-cut mil-
    itary haircuts.
    In 2004, the district court heard testimony from two women
    — Christine Slonaker and Mary Wolfe — who were in the
    bar on the evening of June 4 and were harassed by two or
    three men. These women, who had not previously provided
    testimony, remembered the evening in some detail. They testi-
    fied in the district court that they were at the bar with a third
    friend, and that the friend was receiving unwanted attention
    from some men due to her low-cut blouse. Slonaker, a phle-
    botomist (a person who draws blood), noticed that the man
    who was most aggressive had blood all over him. 6/28/04 RT
    7, 24-25. She stated that from a distance it looked like he was
    just dirty, but that as he got closer “it was clearly apparent
    that it was blood.” 
    Id. at 70
    . Slonaker only noticed two men,
    while Wolfe noticed three, but Wolfe also noted that the third
    man was “pretty quiet and standoffish.” 
    Id. at 121-22
    . She
    also noticed that they were “definitely [wearing] tennis
    shoes[.]” 
    Id.
     Wolfe stated that the man who was not wearing
    coveralls had “medium brown or dirty blond hair that was
    longer than average and tucked behind his ears.” ER 3219.
    Wolfe also noticed that the man in the tan shirt had spots of
    blood on his shirt and a small bit of blood on his face. 6/28/04
    RT 121-23. Wolfe recalled that at least one of the others was
    COOPER v. BROWN                      5447
    wearing coveralls partly zipped down. 
    Id. at 121-22
    . The
    louder man in the coveralls also had blood on him. ER 3220.
    Slonaker noted that “they were saying really weird and gib-
    berish kind of things. . . . They were kind of like their eyes
    were rolling in their head.” 6/28/04 RT 23. When Slonaker
    told the man that he was covered in blood, he acted surprised
    and then his behavior changed. 
    Id. at 24-25
    . Both women
    recalled that the men were asked to leave the bar. 
    Id. at 72, 76, 79, 125
    .
    Lance Stark, a regular at the bar, also testified for the first
    time in 2004. He described “a couple of young loud mouths”
    being rude to some women at the bar. 7/23/04 RT 20-21, 59.
    He also commented that the third man in the group was very
    quiet and not noticeable. 
    Id. at 40
    . He described the men as
    scruffy looking or dirty looking, and he observed that one of
    the men looked like he had grease or mud on him. 
    Id.
     at 22-
    24, 60, 62, 63. He recalls one of the women telling that man
    that he had something on him. 
    Id. at 108-09
    .
    Stark also testified that in early 2004 he was visited by
    someone he believed was associated with law enforcement.
    He testified that a man in a white Crown Victoria (which is
    often a law enforcement vehicle) pulled in front of Stark’s
    trailer, asked whether he was Lance Stark, told him it would
    be in his best interest not to talk about the Kevin Cooper case,
    and then drove off. 
    Id. at 30-32
    . Stark testified that the car
    was distinctive because it had a computer sticking out from
    the dashboard. 
    Id. at 32-34
    . This visit occurred soon after one
    of Cooper’s investigators first visited Stark. 
    Id. at 90
    . When
    Cooper’s attorney met Stark after his encounter with the man
    in the white Crown Victoria, Stark said, “Well, I’m not sure
    if I should talk to you because I was told not to.” 
    Id. at 101
    .
    The district court denied Cooper’s efforts to investigate what
    may have been witness intimidation. ER 4037-39, 4663-65.
    B.   Evidence Against Cooper at Trial
    It was undisputed at trial that Cooper walked away from the
    minimum security section of CIM in Chino on June 2. It was
    5448                  COOPER v. BROWN
    also undisputed that Cooper spent two days hiding in the
    Lease house, located about 125 yards from the Ryen house.
    Telephone records show that Cooper’s last telephone call
    from the Lease house ended at about 8:30 p.m. on June 4, the
    night of the murders.
    In addition to the above, the most important evidence
    against Cooper at trial was the following:
    (1)   Eyewitness testimony of Josh Ryen.
    (2)   A spot of blood on the hallway wall of the
    Ryen house that was consistent with Cooper’s
    blood profile.
    (3)   A bloody shoeprint made by a Pro-Ked Dude
    shoe on a sheet in the master bedroom of the
    Ryen house, a matching shoeprint on a spa
    cover outside the Ryen house, and another in
    the pool room at the Lease house.
    (4)   Role-Rite prison-issue cigarettes and tobacco
    found in the Lease house and in the Ryens’
    abandoned station wagon.
    (5)   A missing hatchet from the Lease house, and a
    hatchet sheath found on the floor of a bedroom
    in the Lease house.
    (6)   A button found in the Lease house that matched
    a prison-issue jacket.
    (7)   An empty beer can in the field between the
    Ryen house and the Lease house.
    (8)   Burrs on Jessica Ryen’s nightgown.
    (9)   Positive Luminol tests in a shower in the Lease
    house.
    COOPER v. BROWN                 5449
    I discuss each piece of evidence in turn.
    1.     Eyewitness Testimony of Josh Ryen
    At trial, the jury heard two recorded statements by Josh,
    one stating and one suggesting that he saw only one man on
    the night of the attack. However, when Josh first arrived at the
    hospital, he was able to communicate to a clinical social
    worker that the assailants were three or four young white
    males. Deputies misrepresented his recollections and gradu-
    ally shaped his testimony so that it was consistent with the
    prosecution’s theory that there was only one killer.
    2.        A-41: The Spot of Blood in the Hallway
    A single drop of blood in the hallway outside the Ryen
    master bathroom — several feet away from any of the victims
    — had characteristics consistent with Cooper’s genetic profile
    and inconsistent with the victims’. The crime lab conducted
    serological testing of this blood drop (entered into evidence as
    A-41) under suspicious circumstances. The criminologist who
    conducted the testing arrived at one result, and then altered
    his records to show a different result that conformed to Coo-
    per’s known blood characteristics. The drop of blood has a
    history of being “consumed” during testing and then inexpli-
    cably reappearing in different form for further testing when
    such testing would prove useful to the prosecution.
    3.   Pro-Ked Dude Shoeprints
    Within the first few days after the murders, deputies dis-
    covered two distinctive matching shoeprints tying the crime
    scene to the Lease house. Later, a deputy in the crime lab dis-
    covered a bloody shoeprint on a bedsheet that had been col-
    lected from the Ryen master bedroom. At trial, the
    prosecution presented evidence that the shoe that likely made
    those shoeprints — a Pro-Ked Dude tennis shoe — was
    nowhere available for retail sale and was only available
    5450                   COOPER v. BROWN
    through institutions such as CIM. The prosecution also pre-
    sented evidence that Cooper had been issued such shoes at
    CIM.
    The shoeprints on the sheet in the master bedroom of the
    Ryen house, on the spa cover outside the Ryen house, and in
    the Lease house were discovered after a suspicious delay. The
    shoeprint on the sheet was not discovered at the Ryen house,
    but rather in the SBCSD Crime Laboratory. Deputy Stockwell
    testified that he discovered the print after re-folding the sheet
    in the lab to match the way it supposedly had been folded, or
    crumpled, on the floor of the bedroom. 11/19/84 RT 3506-07.
    Deputy William Baird was the manager of the lab where the
    sheet was kept when the shoeprint was discovered. Baird pro-
    vided critical testimony at trial connecting the shoeprint on
    the sheet to the shoeprint in the Lease house. He also testified
    that the shoeprints were made by Pro-Ked Dude shoes. ER
    1676-77, 3195-3201. He testified that he already had a Pro-
    Ked Dude shoe in his lab, which he matched to the print on
    the sheet.
    Pro-Ked Dude shoes were manufactured and distributed by
    Stride-Rite Corporation. Deputy Baird admitted at trial that he
    might have told the Stride-Rite official who testified at trial
    that the SBCSD wanted information from him so they could
    “shut down certain defenses.” ER 3200. Soon after Cooper’s
    trial, Baird was caught stealing heroin from the evidence
    locker at the Crime Laboratory. He stole the heroin both for
    his personal use and to sell to drug dealers. ER 1714-16.
    Two additional facts discovered after trial render the shoe-
    print evidence particularly dubious. First, Pro-Ked Dudes
    were, contrary to the testimony at trial, available (though not
    in large quantities) at retail stores in the United States. Sec-
    ond, an inmate who testified at trial that he had issued Pro-
    Ked Dudes to Cooper shortly before his escape recanted his
    trial testimony in a sworn declaration supporting Cooper’s
    application to file his second habeas application.
    COOPER v. BROWN                     5451
    4.   Cigarettes and Tobacco
    Cooper admitted to smoking “Role-Rite” prison-issue
    tobacco while he was in the Lease house. Tobacco consistent
    with Role-Rite was found on the floor between the front pas-
    senger seat and the front passenger door of the Ryens’ station
    wagon. Two cigarette butts were also found in the station
    wagon, and blood typing tests could not exclude Cooper as
    the donor of the saliva on the butts. One of the butts contained
    tobacco that was consistent with the characteristics of the
    Role-Rite brand.
    The station wagon was discovered in a parking lot in Long
    Beach, 45 miles west of the Ryen house, on June 11. When
    the station wagon was discovered, dust prints indicated that
    someone had recently closed the hood. ER 808-09. Cooper
    arrived in Tijuana, 125 miles south of the Ryen house, at 4:30
    p.m. on June 5, the day after the murders, and stayed continu-
    ously at the same Tijuana hotel until June 8. 1/7/85 RT 5874.
    Cooper then went to Ensenada, Mexico, where he found work
    on a private boat. He worked on the boat from June 8 until the
    day of his arrest. 1/3/85 RT 5468-75. The station wagon was
    processed by the police under suspicious circumstances. Some
    cigarette butts from the Lease house were never processed
    into evidence. Some of those cigarette butts could have easily
    been planted in the car. Moreover, after initial forensic test-
    ing, paper from a hand-rolled cigarette butt supposedly found
    in the station wagon was described as consumed. That same
    paper later “reappeared” and was offered into evidence. When
    the paper “reappeared,” it was significantly larger than the
    paper in the cigarette butt that had been tested.
    5.    The Missing Hatchet and the Hatchet Sheath
    On the day the bodies were discovered, detectives recov-
    ered a bloody hatchet beside the road not far from the Ryen
    house. People who had previously used the Lease house testi-
    fied that a similar hatchet was now missing from the house.
    5452                    COOPER v. BROWN
    Investigators found a sheath for the hatchet in the Lease house
    near the closet in the bedroom previously used by Kathleen
    Bilbia (“the Bilbia bedroom”), where Cooper had slept on
    June 3. Fingerprint evidence strongly suggests that the hatchet
    sheath was planted in the bedroom soon after the hatchet was
    discovered. Further, the owners of the hatchet provided incon-
    sistent testimony about the location of the hatchet before it
    disappeared.
    6.     The Camp Jacket Button
    Deputies discovered a green, blood-stained button near the
    closet in the Bilbia bedroom. It resembled buttons found on
    certain “camp jackets” issued at CIM. The blood on the but-
    ton was type A, consistent with Cooper and Doug Ryen. The
    green button was discovered under the same suspicious cir-
    cumstances as the hatchet sheath, strongly suggesting it was
    planted in the Bilbia bedroom after Cooper had become a sus-
    pect. Further, its color showed that it came from a green
    prison-issued jacket. Uncontradicted evidence at trial showed
    that Cooper was wearing a brown or tan prison-issued jacket
    when he escaped.
    7.    The Empty Beer Can
    In the refrigerator in the Ryen house, there was a six-pack
    of Olympia Gold beer with one can missing. Another can in
    the refrigerator, as well as the wall of the refrigerator, were
    smudged with reddish stains. Deputies found a stained, nearly
    empty can of Olympia Gold in the field between the Ryen
    house and the Lease house. The stain on the can in the field
    and on the wall of the refrigerator tested positive for blood.
    The blood stains on the beer cans were so degraded that the
    lab could not conduct any further tests, and deputies failed to
    collect the stain on the wall of the refrigerator as evidence. No
    one analyzed the contents of the nearly empty beer can.
    COOPER v. BROWN                     5453
    8.   Burrs on Jessica Ryen’s Nightgown
    Two burrs adhered to the inside of Jessica Ryen’s night-
    gown approximately ten inches up from the bottom hem. The
    prosecution argued to the jury that because the top of Jessica’s
    nightgown did not have holes corresponding with some of
    Jessica’s post mortem chest wounds, at some point an assail-
    ant must have raised Jessica’s nightgown, and, in the process
    of inflicting those chest wounds, deposited the burrs. The
    prosecution also presented evidence that similar burrs were
    found on the inside of the Ryen station wagon and on a blan-
    ket found in the closet where Cooper slept on June 3. Plants
    producing the burrs grew in the field between the Ryen house
    and the Lease house.
    The plant that produces the burrs is common in Chino
    Hills. It is a relative of alfalfa and is a common ingredient in
    horse and cattle feed. It is unusual for such burrs to transfer
    from one fabric surface to another. Once the burrs have
    adhered to a surface, typically they must be physically
    plucked in order to be removed. 2/4/85 RT 7483-85, 7576-77.
    Moreover, the coroner found a small beetle in Jessica’s body
    bag. This beetle is nocturnal, suggesting that Jessica may have
    been outside during or soon before the murders, when she
    could have picked up the burrs.
    9.   Positive Luminol Tests in the Shower in the Lease
    House
    Detectives tested the shower and sink in the bathroom
    adjoining the Bilbia bedroom in the Lease house for traces of
    blood. Luminol testing revealed the possible presence of
    blood on the shower walls in a broad band from approxi-
    mately two feet to five feet above the floor of the shower.
    When Kathleen Bilbia moved out of the Lease house a
    short time before the murders, she had cleaned her bathroom
    with bleach. Bleach reacts with Luminol in the same way that
    5454                   COOPER v. BROWN
    blood does. In order to exclude the possibility that a Luminol
    reaction is caused by bleach, rather than blood, a two-stage
    test is required. The evidence suggests that the detectives only
    conducted a one-stage Luminol test. Moreover, the staining
    pattern in the shower is not consistent with a person cleaning
    up after being covered in blood. The Luminol test did not
    indicate blood in the bottom portion of the shower, and it did
    not reveal patterns of drainage in the shower. One would
    expect the blood rinsed from a person’s body to travel down-
    ward in the shower, rather than moving horizontally in a
    broad, uniform, 3-foot horizontal band. Hence, the Luminol
    reaction in the shower is probably attributable to Bilbia’s
    cleaning materials, not to the presence of blood. There was no
    indication that any bloody clothing was placed anywhere in
    the vicinity of the shower.
    II.   Discussion
    Cooper advances several claims in his current habeas appli-
    cation. In my view, two of them are meritorious. I mean “mer-
    itorious” in a special sense. In part, I mean that they appear
    to be meritorious on the current record. In part, I also mean
    that if the district court had done its job — including perform-
    ing the EDTA test on Cooper’s blood on the tan t-shirt as we
    directed it to do — the likelihood of their being meritorious
    would be much higher.
    First, Cooper claims that the State presented false evidence
    at trial, in violation of Mooney v. Holohan, 
    294 U.S. 103
    (1935), and Napue v. Illinois, 
    360 U.S. 264
     (1959). Second,
    Cooper claims that the State failed to reveal exculpatory evi-
    dence, in violation of Brady v. Maryland, 
    373 U.S. 83
     (1963).
    Under both claims, Cooper claims actual innocence under
    Schlup v. Delo, 
    513 U.S. 298
     (1995). I discuss the two claims
    in the following sections.
    A.   Presentation of False Evidence
    When we granted Cooper permission to file his second or
    successive habeas application in 2004, we specifically
    COOPER v. BROWN                     5455
    directed the district court to test Cooper’s blood on the tan t-
    shirt “for the presence of the preservative EDTA.” Cooper,
    
    358 F.3d at 1124
    . The purpose of the test was to determine
    whether Cooper’s blood had been planted on the t-shirt. Judge
    Silverman stated the importance of the test: “Cooper is either
    guilty as sin or he was framed by the police. There is no mid-
    dle ground.” 
    Id.
     (Silverman, J., concurring in part and dissent-
    ing in part).
    I begin my discussion of Cooper’s claim that the State pre-
    sented false evidence with discussion of the EDTA testing. I
    do not do so because such testing will reveal directly whether
    the State presented false evidence during trial. The State did
    not put the t-shirt into evidence at trial. Rather, Cooper intro-
    duced it into evidence because the State could not show that
    any of Cooper’s blood was on it. Thus, even if state actors did
    plant Cooper’s blood on the t-shirt, the State did not thereby
    present false evidence at Cooper’s trial. If state actors did
    plant Cooper’s blood on the t-shirt, they likely did so long
    after trial — after DNA technology became available and
    after Cooper requested that DNA testing be performed on the
    t-shirt.
    Nonetheless, for two reasons, I begin with a discussion of
    the EDTA testing. First, if state actors planted Cooper’s blood
    on the t-shirt, this raises a very powerful inference that
    numerous pieces of evidence presented at trial were also
    planted by state actors. Second, if state actors planted this evi-
    dence, its presentation at trial violated Cooper’s due process
    rights under Mooney and Napue and is the basis for a claim
    of innocence under either Schlup or 
    28 U.S.C. § 2244
    (b)(2)(B).
    1.   Testing for EDTA
    The district court was hostile to our direction to perform
    EDTA testing on the t-shirt. In its order, the district court
    wrote, “Based on the last-minute representations made by
    5456                   COOPER v. BROWN
    Petitioner regarding scientific testing capabilities, the en banc
    panel of the Ninth Circuit concluded [that the district court
    should order EDTA testing].” Dist. Ct., 
    510 F.3d at 932
    . In
    discussing the EDTA tests during a hearing, the district court
    stated orally, “But the en banc panel didn’t get it.” 6/3/04 RT
    74. Whether as a result of its hostility to the EDTA testing or
    for some other reason, the district court failed to comply with
    our direction to perform the testing.
    a.   Daubert
    As a threshold matter, the district court concluded, incor-
    rectly, that EDTA testing failed the test of admissibility under
    Daubert v. Merrell Dow Pharmaceuticals, 
    509 U.S. 579
    (1993). The court wrote that “the ubiquity of EDTA in the
    environment prevents any meaningful interpretation of the
    significance of an ‘elevated’ level of EDTA within a forensic
    sample.” Dist. Ct., 
    510 F.3d at 941
     (emphasis added). The
    court found that EDTA testing was therefore not reliable. 
    Id. at 943-46
    . The court’s exclusion of EDTA evidence under
    Daubert was based on an error of law, and therefore consti-
    tuted an abuse of discretion. See United States v. Morales,
    
    108 F.3d 1031
    , 1035 & n.1 (9th Cir. 1997) (en banc).
    The district court confused the reliability of EDTA testing,
    the first prong of the Daubert analysis, with the issue of what
    that testing can prove, the second prong of the Daubert analy-
    sis. See Daubert v. Merrell Dow Pharms., Inc., 
    43 F.3d 1311
    ,
    1315 (9th Cir. 1995) (describing the two prongs of the Dau-
    bert analysis). It was uncontested at the 2004-2005 hearing in
    the district court that laboratory testing can reliably determine
    the amount of EDTA in any given sample. The district court
    found that “the levels of EDTA in the T-shirt were accurately
    measured” by Cooper’s expert, Dr. Ballard. Dist. Ct., 
    510 F.3d at 941
    . The first prong of Daubert was thus satisfied.
    The only contested issue was the second prong — what the
    presence of EDTA proved or tended to prove. The district
    COOPER v. BROWN                      5457
    court rejected Dr. Ballard’s testimony because it erroneously
    concluded that his measurements, even though accurate, did
    not satisfy the second prong of Daubert. See Dist. Ct., 
    510 F.3d at 941
     (“Lacking any evidence to show that EDTA test-
    ing is a reliable means of determining whether a blood sample
    has been planted, the Court concludes that Petitioner’s EDTA
    evidence fails the Daubert test.” (emphasis added)).
    It is hornbook law that evidence is admissible under Dau-
    bert if there is an accepted scientific method for making a
    reliable measurement, even if the evidentiary significance of
    the measurement can be disputed. See Daubert, 
    509 U.S. at 595
     (“The focus, of course, must be solely on principles and
    methodology, not on the conclusions that they generate.”).
    For example, courts (including the trial court in this case) reg-
    ularly admit evidence of Luminol testing for the presence of
    blood, even though Luminol does not indicate the origins of
    the detected blood. In fact, Luminol does not even definitively
    prove the presence of blood. As noted above, other common
    substances, including bleach, also react with Luminol. 11/
    14/84 RT 3142, 3169-70. The district court in this case repeat-
    edly referred to the results of Luminol testing as probative of
    Cooper’s guilt. Dist. Ct., 
    510 F.3d at 907-08, 909, 910, 957
    .
    EDTA test results, like Luminol test results, satisfy the second
    prong of Daubert because they “logically advance[ ] a mate-
    rial aspect of [Cooper’s] case.” See Daubert, 
    43 F.3d at 1315
    .
    The district court concluded that even if EDTA testing
    were admissible under Daubert, the results of the tests in this
    case do not support Cooper’s contention that his blood was
    planted on the t-shirt. Dist. Ct., 
    510 F.3d at 948
    . I disagree.
    In the following two sections, I discuss errors committed by
    the district court in conducting the EDTA tests and in refusing
    to continue that testing. I then discuss the information
    revealed by the testing so far conducted, and show that even
    the truncated EDTA testing strongly suggests that Cooper’s
    blood was planted on the t-shirt.
    5458                   COOPER v. BROWN
    b.   Errors by the District Court
    In conducting the EDTA tests, the district court made six
    fundamental errors.
    First, the district court refused to allow any of Cooper’s
    experts to assist in choosing the portions of the t-shirt to be
    tested. Indeed, the court refused to allow Cooper’s experts
    even to see the t-shirt.
    Second, the district court refused to allow any testing of the
    samples chosen in order to determine whether the stains on
    the samples were actually blood stains.
    Third, the district court refused to permit discovery into
    why blood from vial VV-2 — the blood taken from Cooper
    two days after his arrest — contained the DNA of two differ-
    ent people. One obvious explanation is that someone removed
    some of Cooper’s blood from the vial and then added some-
    one else’s blood to conceal the fact that he or she had
    removed Cooper’s blood.
    Fourth, when the state-designated laboratory came back
    with a result of an extremely high level of EDTA in the sam-
    ple supposed to contain Cooper’s blood, thereby indicating
    that the blood had been planted, the lab director withdrew his
    results because of unspecified “contamination” in his lab. The
    district court refused to permit Cooper’s attorneys to see the
    lab’s raw data or notes, thereby preventing an inquiry into
    whether, in fact, there had been contamination, and into the
    significance of such contamination.
    Fifth, the district court erroneously concluded that several
    samples from the t-shirt were proper “controls” when, in fact,
    they were not. These samples contained both human DNA
    and EDTA, which indicated that they likely contained blood
    that had been planted on the t-shirt. Because the district court
    erroneously concluded that these samples were proper con-
    COOPER v. BROWN                      5459
    trols, it erroneously disregarded the test results obtained by
    the two laboratories.
    Sixth, when the district court concluded that the results of
    the EDTA testing were inconclusive, it refused to permit more
    testing, even though such testing was feasible.
    I discuss each of these errors in turn.
    i.   Refusal to Allow Cooper’s Experts to Assist in
    Choosing Portions of the T-shirt to Be Tested
    One of Cooper’s experts, Dr. Peter DeForest, repeatedly
    sought to participate in the process of examining the tan t-
    shirt and determining which parts of the t-shirt might be suit-
    able for sampling. On August 4, 2004, Dr. DeForest filed a
    declaration noting that “new cuttings” of the t-shirt would be
    needed to perform the testing. He wrote, “[I]t is essential that
    the T shirt be inspected in person to select the areas to be
    sampled by cutting as well as those to serve as suitable con-
    trol areas.” ER 4021. On September 4, he wrote a letter to the
    district court, stating that he was concerned about the testing
    protocol that had been developed. He wrote:
    This protocol was generated without significant
    input from me. I feel this protocol is flawed. I will
    not agree [to] do any sampling according to this pro-
    tocol. It will not be possible to obtain any meaning-
    ful quantitative results using it. . . .
    . . . Once I have examined the shirt, I am willing
    to design an approach for review by another scientist
    or develop one in conjunction with a criminalist rep-
    resenting the prosecution. As I have been trying to
    explain for some time now, there needs to be a care-
    ful assessment of the shirt followed by a scientific
    consensus on the pre-extraction sampling. . . .
    5460                   COOPER v. BROWN
    The samples need to be taken in such a way that
    it is possible to relate any quantitative findings [of
    EDTA levels] to a specific amount of bloodstain.
    This is not possible with the protocol specified in the
    order. In addition, there is the possibility that a sci-
    entific consensus might be that sampling that would
    allow a meaningful result is simply not possible. In
    such a case, samples should not be taken.
    ER 4128-29.
    The response of the district court was to exclude Dr.
    DeForest. On September 7, three days after the date of Dr.
    DeForest’s letter, the district court ordered that the t-shirt be
    sent to Dr. Lewis Maddox of the Orchid Cellmark laboratory.
    Dr. Maddox’s laboratory is not associated with either Cooper
    or the State. The court directed Dr. Maddox and Gary Sims
    (or Sims’ designee) to prepare “Area 6G” of the t-shirt for
    testing, and to select other portions of the t-shirt for use as
    controls. ER 4151. Mr. Sims is Director of the California
    Department of Justice Laboratory. The court’s order did not
    allow a representative of Cooper to be present during the
    selection process.
    Dr. DeForest was prescient. Area 6G of the t-shirt (the area
    specified in the court’s order) was the area that had been orig-
    inally tested for DNA. That earlier testing had confirmed
    Cooper’s blood was present in Area 6G. Upon close inspec-
    tion by Dr. Maddox, Area 6G turned out to be unsuitable for
    further testing because there was no blood remaining in that
    area. The State notified the district court of this fact on Sep-
    tember 13.
    That same day, Cooper’s lawyer wrote to the district court,
    “[M]ost importantly, Petitioner vigorously requests that an
    expert of his be allowed to inspect the T-shirt and be part of
    the selection and preparation process for the anti-coagulant
    [EDTA] testing.” ER 4205. At 5:30 p.m. that same day, the
    COOPER v. BROWN                      5461
    district court denied Cooper’s request to have a representative
    present during the selection and preparation process. The
    court wrote, “The court denies petitioner’s request to have his
    own observer present at the preparation of the T-shirt for the
    EDTA testing. The court acknowledges that the 6-G stain is
    not suitable for testing. The court orders Dr. Maddox, in con-
    sultation with Dr. Myers [Mr. Sims’ designee, the State’s rep-
    resentative], to select an appropriate stain area and prepare it
    for EDTA testing[.]” ER 4207.
    As a matter of due process, a court is required to allow both
    sides to participate when important decisions are made.
    Where, as here, serious objections were made to the manner
    of choosing and processing samples to be tested, the district
    court failed in its duty to provide a fundamentally fair pro-
    cess.
    ii.   Refusal to Allow Testing of the Newly Chosen Sample
    for the Presence of Blood
    Dr. Maddox and Mr. Sims’ designee, Mr. Myers (the dis-
    trict court erroneously referred to him as Dr. Myers), chose an
    area of the t-shirt between stains labeled 6J and 6K as a
    replacement for Area 6G. In their view, this area was likely
    to contain Cooper’s blood. They therefore took their sample
    from this area. They then divided the sample into three pieces.
    They sent one of the pieces to the state-designated lab, sent
    one to Cooper’s designated lab, and retained the third piece
    at Dr. Maddox’s lab. However, no one tested the new sample
    (or any part of it) to determine if any of the three pieces actu-
    ally contained blood.
    Cooper objected to the failure to test the newly chosen sam-
    ple for blood. He specifically requested that it be tested to
    determine if it contained blood. The district court denied the
    request on the ground that Cooper had not previously
    requested testing of the sample for the presence of blood. See
    4/22/05 RT 10-11, 57-58, 171-72; see also Dist. Ct., 
    510 F.3d 5462
                        COOPER v. BROWN
    at 935 n.16. The district court’s refusal to test the newly cho-
    sen sample for the presence of blood was wrong on two
    counts.
    First, the district court was wrong procedurally. It was
    unfair to fault Cooper for not having previously requested
    testing of the sample for blood. Up until September 13, Coo-
    per reasonably assumed that the sample to be chosen for test-
    ing would come from the stain in Area 6G, which everyone
    believed contained Cooper’s blood. It was therefore unneces-
    sary to request testing of a sample from Area 6G for the pres-
    ence of blood. Such testing only became necessary when a
    new sample was chosen. After Area 6G was deemed unsuit-
    able, even the State told the district court that additional test-
    ing of the new subject sample “would be required to
    determine whether Cooper’s blood is actually present in the
    stain.” ER 4194. Cooper had no reason specifically to request
    this additional testing after the State said that it was “re-
    quired.”
    Second, and more important, the district court was wrong
    substantively. Because of the failure to test the new sample
    for the presence of blood, it was possible that, as intended by
    the protocol, all of the pieces of the new sample sent for
    EDTA testing had Cooper’s blood. But it was also possible
    that one or more of the pieces had none of his blood. This sec-
    ond possibility was greatly enhanced for the new sample, as
    compared to the old one from Area 6G. The new sample was
    adjacent to Area 6G, and therefore Dr. Maddox and Mr.
    Myers assumed that it contained Cooper’s blood. But it was
    unclear how far into the new sample Cooper’s blood extended
    (if indeed his blood extended into the new sample at all). Fur-
    ther, the new sample was both larger and more irregularly
    shaped than the old sample, making it even more likely that
    any blood on the sample was not evenly distributed through-
    out the entire sample.
    As I discuss below, these characteristics of the new sample
    may well account for the radical difference obtained by the
    COOPER v. BROWN                            5463
    two labs in testing their pieces of the sample. The State-
    designated lab found an extremely high level of EDTA in its
    piece. The Cooper-designated lab found an elevated but lower
    level of EDTA in its piece. This disparity could well have
    happened because the state-designated lab tested a piece that
    contained Cooper’s blood, and the Cooper-designated lab
    tested a piece that did not contain his blood, or contained con-
    siderably less of it. If the district court had permitted testing
    of the new sample for blood — and, specifically, Cooper’s
    blood — we would know the answer.
    iii. Refusal to Permit Inquiry into Why Vial VV-2, Which
    Should Have Contained Only Cooper’s Blood, Contained
    the DNA of Two or More People
    On August 1, 1983, two days after Cooper’s arrest, two
    SBCSD criminalists drew Cooper’s blood. They put that
    blood into a vial labeled VV-2. The vial contained the preser-
    vative EDTA. In 2004, during the court-ordered testing of the
    hairs Jessica clutched in her hand,3 the State made a mistake.
    It inadvertently sent a card containing blood from vial VV-2
    to Dr. Terry Melton, the expert charged with testing the hairs.
    ER 3187. This was the first time since 1983 that any non-
    State personnel had been permitted to see or test blood from
    vial VV-2.
    Dr. Melton tested the blood from VV-2, unaware of the fact
    that the State had not intended to send it to her. Dr. Melton
    found that the blood from VV-2 contained the DNA of two
    or more people. This was a truly startling finding. On August
    2, 2004, Dr. Melton informed the court of her finding. ER
    5645.
    3
    I do not pursue the argument here, but there is ample reason to con-
    clude that the district court also disregarded our directive by unduly limit-
    ing the mitochondrial DNA testing of the hairs in Jessica Ryen’s hand and
    by prematurely foreclosing the opportunity for further testing.
    5464                   COOPER v. BROWN
    Vial VV-2 originally contained only Cooper’s blood, and
    should have continued to contain only Cooper’s blood. The
    most logical explanation for the finding is that someone added
    another person’s blood to the vial. Why might that have been
    done? One explanation is that someone took some of Coo-
    per’s blood out of the vial for some purpose (planting it on the
    t-shirt?), and wanted the vial to appear as full as it previously
    had been. In order to accomplish that, he or she had to add
    someone else’s blood to the vial to bring it back up to the
    proper level.
    On August 4, Cooper’s lawyer raised Dr. Melton’s discov-
    ery with the district court. Perhaps the court thought Cooper’s
    lawyer was speaking of DNA from the hairs. The court stated,
    “[W]e never expected that it was going to be Cooper.” 8/6/04
    RT 138. Counsel replied, “[I]t is not the hairs that were sent
    that we’re talking about. It is the known sample that was sent,
    and that’s been contaminated. And there is a very serious
    issue about that.” 
    Id. at 139
    .
    On September 10, Cooper moved for an evidentiary hear-
    ing, inter alia, “to determine the cause for the appearance of
    a ‘mixture’ of DNA in Petitioner’s blood sample also submit-
    ted to Dr. Melton.” He wrote:
    VV-2 is the blood sample collected from Petitioner
    at the time of his arrest[.] . . . [The] [b]lood sample
    should only have contained Petitioner’s DNA[.] . . .
    Dr. Melton’s report reveals that a mixture of DNA
    sources was detected in VV-2. . . . In light of prior
    evidence presented by Petitioner regarding tamper-
    ing or contamination of biological evidence in this
    case, Dr. Melton’s findings regarding VV-2 are
    extremely alarming and mandate further inquiry.
    ER 4168. On February 3, 2005, the district court denied Coo-
    per’s motion. It did not mention vial VV-2 in its order.
    COOPER v. BROWN                     5465
    On April 22, 2005, in final oral argument to the district
    court, Cooper’s counsel returned to the subject of the blood
    in vial VV-2. He said, “[W]ith regard to VV-2, I just want to
    be — make this clear. There seems to be a possible misunder-
    standing. VV-2, which is the sample that Doctor Melton
    tested and found a mixture in, it’s Petitioner’s blood sample.
    It is not a hair sample. I wasn’t sure if the Court was clear on
    that.” 4/22/05 RT 153. The court immediately interrupted,
    “And it’s consumed.” 
    Id.
     Cooper’s counsel agreed that Dr.
    Melton had consumed the sample on the card she had been
    sent, but stated, “[T]hat doesn’t necessarily mean that there
    isn’t more VV-2 in San Bernardino or at the DOJ that could
    be tested.” 
    Id.
     The State’s counsel then responded, “I can rep-
    resent to the Court that VV-2 was completely consumed by
    Doctor Melton. Doctor Melton was shipped the remainder of
    this particular blood sample, and she consumed it, and it’s
    reflected in her report. So we don’t have any more of that par-
    ticular reference sample.” 
    Id. at 156
    .
    The State’s counsel seems to have meant to say (or at least
    to have meant the court to understand) that there was no blood
    remaining in vial VV-2. If that is what counsel meant to say,
    it was a startling statement. The State had never before said
    or even suggested such a thing. For example, when Cooper
    moved in September 2004 for an evidentiary hearing on how
    the DNA of two people came to be in vial VV-2, the State did
    not say or even suggest that vial VV-2 was empty. Nor had
    the State ever presented evidence to support such a statement.
    Nor had the district court relied on the fact that vial VV-2 was
    empty in denying Cooper’s motion for an evidentiary hearing.
    Cooper’s counsel told the district court that Dr. Melton’s
    finding that the DNA of two or more people was in the blood
    that came from vial VV-2 was “extremely alarming and man-
    date[d] further inquiry.” The district court refused to allow
    any investigation into the issue, even though the presence of
    additional DNA in vial VV-2 clearly pointed to evidence tam-
    pering by the State.
    5466                   COOPER v. BROWN
    iv. Refusal to Allow Access to the State-Designated Lab’s
    Raw Data and Notes Concerning Asserted Contamination
    Dr. Maddox of the Orchid Cellmark laboratory sent a total
    of ten samples to each of the two designated testing labs. Dr.
    Siuzdak was the tester designated by the State. Dr. Ballard
    was the tester designated by Cooper.
    Sample 1 was a piece of the t-shirt that had been chosen by
    Dr. Maddox and Mr. Myers as likely to contain Cooper’s
    blood. Samples 2-6 were taken from other parts of the t-shirt
    and were intended to serve as controls. Samples 7-10 were not
    taken from the t-shirt; they were also intended to serve as con-
    trols.
    As I will explain below, the EDTA results obtained by Dr.
    Siuzdak and Dr. Ballard are remarkably consistent for all of
    the samples except Sample 1 and Sample 8. Dr. Siuzdak
    found that his piece of Sample 1 (the sample supposedly con-
    taining Cooper’s blood) contained an extremely high level of
    EDTA, more than twice as high as any other sample. If Dr.
    Siuzdak’s piece of Sample 1 contained Cooper’s blood, and
    if his EDTA result is valid, this indicates that Cooper’s blood
    was planted on the t-shirt. By contrast, Dr. Ballard found that
    his piece of Sample 1 contained a somewhat elevated, but
    fairly low, level of EDTA.
    Dr. Siuzdak submitted his report, containing the high
    EDTA reading for Sample 1, to the district court on October
    5, 2004. On October 27, without prior warning, Dr. Siuzdak
    withdrew his report. His fax to the court stated in its entirety:
    On Monday, October 5th I submitted a report on
    the Cooper samples tested for the presence of
    EDTA. I now believe that the samples tested were
    contaminated with EDTA in my laboratory and
    therefore must retract the report submitted. I deeply
    COOPER v. BROWN                      5467
    apologize for the inconvenience and confusion this
    report may have caused.
    ER 4464.
    Cooper moved to be allowed access to Dr. Siuzdak’s raw
    data and bench notes relevant to his testing of all the samples.
    ER 4465-82. The district court denied this access. ER 4751.
    Cooper has never been permitted to see Dr. Siuzdak’s raw
    data and bench notes, and has never been permitted to investi-
    gate the nature and possible significance of the purported con-
    tamination. Dr. Siuzdak has never been asked to provide an
    explanation for his conclusion that there was contamination in
    his lab.
    v.    District Court Reliance on Faulty Controls
    As discussed in greater detail below, five supposed “con-
    trol” samples were taken from the tan t-shirt. When the two
    laboratories tested these samples, everyone assumed that they
    contained no human material (and therefore no human DNA)
    and that they contained only a background level of EDTA.
    However, at least three of the five purported control samples
    taken from the t-shirt (Samples 2, 3, and 4) actually contained
    human DNA. ER 4659, 4669. The amount of DNA in these
    samples corresponds closely with an elevated level of EDTA
    in these same samples. The combined presence of DNA and
    elevated levels of EDTA strongly suggest that these samples
    contained preserved blood that had been planted on the t-shirt.
    Therefore, these “control” samples from the t-shirt were not,
    in fact, controls at all. Nonetheless, the district court assumed
    that these samples were valid controls.
    Because the district court assumed that Samples 2, 3 and 4
    were valid controls, it concluded that the results of the DNA
    tests were invalid. See Dist. Ct., 
    510 F.3d at 939
     (“The EDTA
    level in the subject stain is not elevated, but is instead lower
    than that of most of the control areas. As a result, the test
    5468                   COOPER v. BROWN
    refutes Petitioner’s tampering theory.” (emphasis added)); 
    id. at 941
     (“From the test data, the Court concludes that the level
    of EDTA in the subject stain is 110 nanograms. Comparing
    the EDTA level of the subject stain to that for the control
    specimens, the Court concludes that there is no reliable evi-
    dence of tampering.” (emphasis added)). But because the dis-
    trict court erred in assuming that these samples were valid
    controls, it erred in concluding that the results of the DNA
    tests were invalid.
    The only sample that could provide a baseline level of
    EDTA in the t-shirt, and could thus serve as a valid control,
    is Sample 6. It is the only sample from the t-shirt that con-
    tained no human DNA. Both Dr. Ballard and Dr. Siuzdak
    found only background levels of EDTA in Sample 6. The
    levels of EDTA found in Sample 6 by both Dr. Ballard and
    Dr. Siuzdak were far lower than the levels of EDTA found in
    any other sample taken from the t-shirt. (Recall that the dis-
    trict court found that “the levels of EDTA in the T-shirt were
    accurately measured” by Dr. Ballard. Dist. Ct., 
    510 F.3d at 941
    .) The virtual absence of EDTA in Sample 6 when com-
    pared with the elevated EDTA levels in Sample 1, particularly
    the portion of Sample 1 tested by Dr. Siuzdak, strongly sug-
    gests that Cooper’s blood was planted on the shirt.
    vi.   Refusal to Allow Further Testing
    The district court refused to allow further EDTA testing
    after Dr. Ballard and Dr. Siuzdak made their reports, and after
    Dr. Siuzdak withdrew his report on the ground of asserted
    contamination. If the piece of Sample 1 tested by Dr. Siuzdak
    contained Cooper’s blood, as Dr. Maddox and Mr. Myers
    thought it did, and if Dr. Siuzdak’s finding of an extremely
    high level of EDTA in his piece of Sample 1 is valid, there
    is no question that Cooper’s blood was planted on the t-shirt.
    The district court refused to allow further testing despite
    the fact that Dr. Siuzdak and Dr. Ballard still have sufficient
    COOPER v. BROWN                      5469
    amounts of the samples provided to them to perform further
    testing. In light of the high level of EDTA detected by Dr.
    Siuzdak on his piece of Sample 1, and in light of the unex-
    plored nature of Dr. Siuzdak’s belief that there was contami-
    nation in his laboratory, it was unreasonable not to pursue
    further testing.
    c.   What the EDTA Test Results Tell Us When Combined
    with DNA Tests on the Same Samples
    After Dr. Ballard and Dr. Siuzdak submitted their reports
    to the district court, the Orchid Cellmark laboratory sent other
    pieces of the samples to the State of California’s DNA labora-
    tory to determine how much DNA was on those samples. ER
    4659, 4669. When Dr. Ballard and Dr. Siuzdak submitted
    their reports, they did not know the results of this DNA test-
    ing.
    The results of the EDTA testing (performed by Drs. Ballard
    and Siuzdak) and the DNA testing (performed by the State
    lab) can be put into a single table. In reading this table, recall
    that Sample 1 is the sample thought to contain Cooper’s
    blood.
    5470                      COOPER v. BROWN
    Table 1: EDTA and DNA Data (nanograms)4
    We can draw two conclusions from the table.
    First, we can conclude that Dr. Siuzdak’s EDTA results are
    very likely valid based on a comparison of his results with
    those of Dr. Ballard. A comparison of the results for Samples
    2-7 and 9-10 (that is, excluding Samples 1 and 8) shows that
    the results obtained by Dr. Siuzdak and Dr. Ballard are con-
    sistent. Dr. Siuzdak’s results are expressed in higher absolute
    numbers than Dr. Ballard’s, but the relative numbers of the
    two are remarkably consistent. This may be seen in the fol-
    lowing graph:
    4
    The data are from Lee Report (Siuzdak EDTA); ER 5414 (Ballard
    EDTA); ER 4659, 4669 (DNA). The district court did not order DNA test-
    ing of control Samples 7 through 9.
    5
    Dr. Siuzdak reported his values as concentration of EDTA. Dr. Lee’s
    analysis of the EDTA results converted Dr. Siuzdak’s “values to the
    equivalent amount of EDTA in the original cloth cuttings.” Lee Report at
    2.
    COOPER v. BROWN                            5471
    Graph 1: Comparison of Ballard and Siuzdak EDTA
    Measurements6 (Excluding Samples 1 and 8)
    The consistency in results for Dr. Siuzdak and Dr. Ballard
    strongly suggests that Dr. Siuzdak’s results are valid. That is,
    if Dr. Siuzdak had EDTA contamination in his lab or if his
    instruments were miscalibrated, the contamination or miscali-
    bration affected all of his measurements equally. If this is
    true, Dr. Siuzdak’s results — as measurements of relative
    rather than absolute amounts of EDTA — are valid.
    6
    The best fit linear regression equation for these data is: Siuzdak =
    288.078 + 3.031 * Ballard. The R2 value for these data is 0.892; the
    regression coefficient for the independent variable is statistically signifi-
    cant at p < 0.001.
    5472                   COOPER v. BROWN
    For purposes of this comparison (and of the above graph),
    I excluded Samples 1 and 8. Sample 8 is a control not taken
    from the t-shirt. It is a piece of cloth onto which Dr. Maddox
    at Orchid Cellmark had placed a known amount of EDTA-
    preserved blood, with a known concentration of EDTA. Both
    Dr. Siuzdak and Dr. Ballard found greatly elevated levels of
    EDTA in Sample 8, though Dr. Siuzdak found a relatively
    higher EDTA level. That difference may be partially
    explained by Dr. Siuzdak’s relatively higher amounts of
    EDTA for all of his samples. The remaining disparity may not
    be significant. Dr. Terry Lee, an expert hired by the State to
    evaluate the results of Dr. Siuzdak and Dr. Ballard, concluded
    that the disparity in results for Sample 8, considered by itself,
    was not significant. Respondent’s Analysis of EDTA Test
    Results, Oct. 28, 2004, Docket No. 264, Ex. A at 2.
    That leaves Sample 1. Sample 1 is the new sample thought
    to contain Cooper’s blood. It was chosen by Dr. Maddox and
    Mr. Myers to replace the old sample from Area 6G. Though
    Sample 1 supposedly contained Cooper’s blood, we cannot be
    sure that it did, as I explained above. Dr. Siuzdak measured
    an extremely high level of EDTA for Sample 1. Dr. Ballard
    measured only a somewhat elevated level of EDTA for Sam-
    ple 1. A possible, perhaps likely, reason for the disparity in
    their results is that Dr. Siuzdak’s piece of Sample 1 contained
    a great deal of Cooper’s blood, while Dr. Ballard’s piece con-
    tained significantly less, perhaps none. Sample 1 was an
    unusually large sample by comparison to the other samples.
    It was also irregularly shaped, unlike the other samples which
    were simple squares. ER 5202-03. Finally, Dr. Maddox and
    Mr. Myers assumed that all of Sample 1 contained Cooper’s
    blood. But, as noted above, they did not perform any tests on
    Sample 1, or its individual pieces, to confirm this assumption.
    Second, we can conclude that Dr. Siuzdak’s EDTA results
    are very likely valid, based on analysis of Samples 2, 3 and
    4. Samples 2, 3 and 4 were supposed to be control samples
    that contained no blood. Instead, they almost certainly contain
    COOPER v. BROWN                      5473
    significant amounts of blood. Further, it is likely that this
    blood was planted on the t-shirt. This may be seen if we look
    at both the DNA and the EDTA results for Samples 2, 3, 4
    and 6. All four of these samples were taken from the t-shirt,
    and were supposed to have been control samples. (Sample 5
    has inconclusive DNA results, so I put it to one side. Samples
    7, 8, 9 and 10 are control samples that were not from the t-
    shirt, so I put them to one side also.)
    For Samples 2, 3, 4 and 6, we have a known amount of
    DNA and a known amount of EDTA. Samples 2, 3 and 4 all
    have significant amounts of DNA. Only Sample 6 has no
    DNA. For all four of these samples, there is a remarkably
    strong correlation between the amount of DNA in the sample
    and the amount of EDTA in the sample. The greater the
    amount of DNA in a sample, the greater the amount of EDTA
    in that sample. This holds true for the results of both Dr. Siuz-
    dak and Dr. Ballard. This may be seen in two graphs, one for
    Dr. Siuzdak’s results and one for Dr. Ballard’s results:
    5474                       COOPER v. BROWN
    Graph 2: Correlation Between DNA and Siuzdak’s EDTA
    Measures7 (Samples 2, 3, 4, and 6)
    7
    The best fit linear regression equation for these data is: DNA = -0.185
    + 0.001 * Siuzdak. The R2 value for these data is 0.974; the regression
    coefficient for the independent variable is statistically significant at p =
    0.013.
    COOPER v. BROWN                            5475
    Graph 3: Correlation Between DNA and Ballard’s EDTA
    Measures8 (Samples 2, 3, 4, and 6)
    These graphs, showing the presence and amount of DNA
    in Samples 2, 3, 4, and 6, tell us a number of things. First,
    they tell us that Samples 2, 3 and 4 are not valid controls.
    Thus, the district court erred in treating them as controls when
    it interpreted (and disregarded) Dr. Ballard’s results. See, e.g.,
    Dist. Ct., 
    510 F.3d at 939
    . The EDTA levels in Sample 6, the
    8
    The best fit linear regression equation for these data is: DNA = -0.093
    + 0.002 * Ballard. The R2 value for these data is 0.955; the regression
    coefficient for the independent variable is statistically significant at p =
    0.023.
    5476                   COOPER v. BROWN
    only true control, are significantly lower than the EDTA
    levels in Samples 2, 3 and 4, and in Sample 1, the sample
    thought to contain Cooper’s blood.
    Second, the variation in EDTA levels in three supposed
    control samples — Samples 2, 3 and 4 — directly corre-
    sponds with the variation in human DNA levels in those sam-
    ples. EDTA is present in many substances, such as laundry
    detergent, salad dressing, and household cleaners. See Dist.
    Ct., 
    510 F.3d at
    947 & n.34. Those substances, however, do
    not contain human DNA. Human blood stored in purple-
    topped EDTA tubes, such as those used by the SBCSD, has
    both EDTA and human DNA. Samples 2, 3 and 4 had ele-
    vated levels of EDTA and correspondingly elevated levels of
    human DNA. The most plausible explanation is that the
    EDTA was deposited on the t-shirt along with human DNA.
    The most logical source of such a deposit is human blood
    stored in a purple-topped EDTA tube.
    We know that Cooper’s EDTA-preserved blood sample
    (VV-2) and other EDTA-preserved blood samples from the
    victims were kept in an evidence refrigerator to which up to
    twenty people had ready access. 6/23/03 RT 106, 134. We
    also know that blood could be withdrawn from those tubes
    without breaking any seals. 6/25/03 RT 330. Finally, we know
    that “anyone with access to that refrigerator . . . could take
    items in and out of it without there being any record of it at
    all, without any written record.” 6/23/03 RT 134. This infor-
    mation, coupled with the EDTA and DNA findings described
    above, strongly suggest that someone in the SBCSD took
    blood from one or more purple-topped EDTA tubes and
    placed that blood on the t-shirt.
    Third, the district court did not account for the correlation
    between human DNA and EDTA in the control samples when
    it stated: “While the extraction and measurement of EDTA in
    a sample may theoretically be accomplished, the ubiquity of
    EDTA in the environment prevents any meaningful interpre-
    COOPER v. BROWN                     5477
    tation of the significance of an ‘elevated’ level of EDTA
    within a forensic sample.” Dist. Ct., 
    510 F.3d at 941
    . The data
    demonstrate that, contrary to the view of the district court,
    meaningful interpretation of the significance of an elevated
    level of EDTA is possible. They demonstrate, in fact, that Dr.
    Siuzdak’s test results were probably valid.
    d.   What Dr. Siuzdak Did Not Know
    Neither Dr. Siuzdak nor Dr. Ballard, nor indeed any expert,
    was given the opportunity by the district court to look at both
    the EDTA and the DNA data.
    When Dr. Siuzdak withdrew his results, he was unaware of
    Dr. Ballard’s results. He was also unaware of the results of
    the DNA testing by the State. We are now able to see what
    Dr. Siuzdak could not see when he withdrew his test results.
    First, Dr. Siuzdak was unaware of the remarkable congru-
    ence of his supposedly contaminated results with those of Dr.
    Ballard. Second, Dr. Siuzdak was unaware that Samples 2, 3
    and 4 each contained significant amounts of DNA, which
    closely correlated with the amount of EDTA in those samples.
    Dr. Siuzdak mistakenly thought those samples contained no
    blood, and therefore should have contained no EDTA. He
    thus thought that Samples 2, 3 and 4 were valid controls, and
    that he should have found little or no EDTA in them. Had Dr.
    Siuzdak known that Dr. Ballard had obtained remarkably con-
    gruent results, and that Samples 2, 3 and 4 contained human
    DNA and were not valid controls, it is entirely possible, per-
    haps likely, that he would not have withdrawn his results
    based on supposed contamination in his lab. It should have
    been an easy matter to ask Dr. Siuzdak whether, in light of
    what the district court later learned, he would still withdraw
    his test results. The district court, however, did not do so.
    e.   Summary
    It is clear from the foregoing that the district court did not
    properly carry out our directive to perform EDTA testing on
    5478                   COOPER v. BROWN
    Cooper’s blood on the t-shirt. The court did not permit Coo-
    per’s experts to participate in or even see the selection of the
    t-shirt samples chosen for testing. It refused to allow testing
    of the newly chosen sample, which was supposed to contain
    only Cooper’s blood, to determine if, in fact, it contained his
    blood, or any blood at all. It refused to permit an inquiry into
    why vial VV-2, which was supposed to contain Cooper’s
    blood, contained the DNA of two or more people. It refused
    to permit discovery of Dr. Siuzdak’s raw data and bench notes
    after he withdrew his test results — results that strongly sug-
    gested that Cooper’s blood had been planted. It incorrectly
    concluded that Samples 2, 3 and 4 were proper controls, and
    therefore incorrectly concluded that the testing laboratories’
    results could not be correct. And it refused to allow further
    testing after it allowed Dr. Siuzdak to withdraw test results
    that had been favorable to Cooper.
    The test results that we do have strongly suggest that Dr.
    Siuzdak may not have had contamination in his laboratory
    that invalidated his results, and strongly suggest that Dr. Siuz-
    dak’s sample, supposedly containing Cooper’s blood, did in
    fact contain an extremely high level of EDTA. In other words,
    the test results we already have strongly suggest that Cooper’s
    blood was planted on the t-shirt.
    2.   False Evidence Presented at Trial
    If the EDTA testing already performed shows that Cooper’s
    blood was planted on the t-shirt, or if further EDTA testing
    does the same thing, that showing greatly increases the likeli-
    hood that much of the evidence introduced at trial was false.
    If the State introduced false evidence at trial, it violated Coo-
    per’s constitutional right to due process under Mooney and
    Napue. If Cooper’s constitutional right to due process was
    violated, he has satisfied the threshold test for showing actual
    innocence under Schlup. In assessing whether a federal
    habeas petitioner has made a showing of actual innocence
    under Schlup, a district court is not limited to evidence intro-
    COOPER v. BROWN                      5479
    duced at trial. Rather, it may consider all relevant evidence.
    See Carriger v. Stewart, 
    132 F.3d 463
    , 478 (9th Cir. 1997) (en
    banc).
    a.    The False Evidence
    If Cooper’s blood was planted, that would greatly increase
    the likelihood that the following evidence at trial was false.
    i.        Josh Ryen’s Testimony
    Josh Ryen was eight years old at the time of the murders.
    He was left for dead in his parents’ bedroom with his throat
    cut. He did not take the stand, but his audio- and videotaped
    testimony was introduced at trial. That testimony was that he
    saw either a single man or a single shadow in the house dur-
    ing the murders. This testimony was flatly inconsistent with
    the information he provided immediately after the murders.
    (a).    Josh Ryen’s Testimony at Trial
    Josh Ryen did not testify in person at trial. Instead, two
    recorded interviews with Josh were played for the jury. One
    was an audiotaped interview conducted in the office of a psy-
    chiatrist, Dr. Lorna Forbes, on December 1, 1983, six months
    after the murders. The other was a videotaped interview con-
    ducted on December 9, 1984, in the midst of Cooper’s trial.
    A transcript of the audiotaped interview with Dr. Forbes
    includes the following statements:
    Q [Dr. Forbes]:           Who went to bed first? You guys
    or everybody all together or what?
    A [Josh]:      I think my Mom, I, my sister and Chris
    went to bed, but my Dad stayed up.
    Q:   You don’t remember, for sure?
    5480                  COOPER v. BROWN
    A:   That’s what I think. Then I heard my Mom
    scream, and I walked in there with Chris, and I
    saw someone.
    Q:   Well, did, did you mean? Where did you see
    him?
    A:   Like over by the bed, my Mom’s bed.
    ***
    Q:   Okay, was the person standing up or what?
    ***
    A:   And then I went back, and . . .
    Q:   Well, why did you go back; what was happen-
    ing with the person standing by the bed; was he
    standing facing you, or what did you see?
    A:   He was turning his back against me.
    ***
    Q:   Do you remember seeing him doing anything to
    your mother when she was screaming?
    A:   No, I just saw his back and his hair.
    Q:   And his hair? Was he bending over or standing
    up straight or what?
    A:   Uh, he was like . . . His head was down, he was
    down.
    Final Stipulation Regarding Correction of Transcripts on
    Appeal, July 31, 1986, Interview of Josh Ryen by Lorna For-
    COOPER v. BROWN                     5481
    bes, Ex. D at 14-16 (ellipses in original; asterisks indicate
    omitted material). Later in the interview, Dr. Forbes asked
    Josh, “you went in there and saw this thing; did you think it
    was a man or a woman?” Josh replied, “I didn’t . . . probably
    a man because women usually don’t do that sort of thing.” Id.
    at 19.
    We know that in this audiotaped interview Josh had
    described seeing a “puff of hair” on the killer. We also know
    that the State excised the “puff of hair” reference from the
    interview, and that this part of the interview was not played
    to the jury. 12/13/84 RT 4991-92. (I discuss below the signifi-
    cance of the “puff of hair” reference, and the significance of
    the prosecution not playing it for the jury.)
    During the videotaped second interview, conducted in
    December 1984, counsel for both sides questioned Josh.
    Josh’s grandmother, Dr. Mary Howell, was also present dur-
    ing the interview. The videotaped interview was played for
    the jury on December 13, 1984, immediately before the jury
    heard the audiotaped interview with Dr. Forbes. The prosecu-
    tor asked, “Did you ever see anybody in the house that didn’t
    belong there?” The examination proceeded:
    Q [prosecutor]:    And when you say it could have
    been someone like your mom,
    who are you talking about, or
    what did you see that —
    A [Josh]:   I don’t really — saw like a shadow or
    something.
    Q:   Where did you see the shadow?
    A:   Like the bathroom was here. I saw it right like
    the bathroom.
    Q:   By the bathroom?
    5482                  COOPER v. BROWN
    A:   Yeah.
    Q:   When did you see the shadow?
    A:   When I saw my sister.
    Q:   And how many shadows did you see?
    A:   Just one.
    12/13/84 RT 4968-69 (emphasis added).
    In closing argument, the last evidence the prosecutor dis-
    cussed in rebuttal was Josh’s recorded interviews. He empha-
    sized the consistency of Josh’s story in the two interviews,
    and argued that consistency showed that Cooper was the per-
    petrator:
    We have from Josh the same basic story to differ-
    ent people at different times that have been both tape
    recorded and videotaped.
    . . . During the entire time that Josh was with his
    grandmother, a time of protection and love, Josh told
    the same basic story, a story that you ladies and gen-
    tlemen got a chance to hear, the story that shows that
    there was just . . . one attacker: Kevin Cooper with
    a hatchet in one hand and a knife in the other.
    2/7/85 RT 7823-24. On the second day of deliberations, the
    jury requested to rehear the entire audiotaped interview with
    Dr. Forbes. 2/8/85 RT 7863-A. This was the first of only two
    read-backs requested by the jury.
    The story told in the two recorded interviews is incompati-
    ble with Josh’s statements made immediately after the mur-
    ders.
    COOPER v. BROWN                    5483
    (b).   Josh Ryen’s Statements Immediately After the
    Murders
    When Josh first arrived at the emergency room on the after-
    noon of June 5, he was unable to speak because his throat had
    been cut. Josh’s handwriting was illegible. A clinical social
    worker named Don Gamundoy devised an effective way for
    Josh to communicate basic information. Gamundoy wrote out
    the letters of the alphabet, numerals 0 through 9, and the
    words “yes” and “no” on a piece of paper on a clipboard. Josh
    was able to point to the letters, numbers, and the two words
    in responding to Gamundoy’s questions.
    Josh accurately indicated his full name, his date of birth,
    and his telephone number. Then Gamundoy asked Josh how
    many people attacked him. Josh pointed to “3” and “4” on the
    sheet. Gamundoy asked whether the attackers were male. Josh
    pointed to “yes.” Gamundoy asked whether they were black.
    Josh pointed to “no.” Gamundoy asked whether the attackers
    looked like Gamundoy, who is Hawaiian but is frequently
    mistaken for Hispanic. Josh pointed to “no.” Gamundoy asked
    if the attackers were white. Josh pointed to “yes.” ER 1272-
    79. Medical personnel treating Josh in the emergency room
    described Josh as conscious, alert, and immediately respon-
    sive during the time Gamundoy was questioning him. 1/10/85
    RT 6229-30; 1/17/85 RT 6626, 6630, 6636.
    SBCSD Sergeant Arthur, who was in charge of the investi-
    gation, dispatched Deputy Sharp to the emergency room. 1/
    8/85 RT 5980-82. Deputy Sharp was the first member of law
    enforcement to question Josh. Deputy Sharp entered the emer-
    gency room as Gamundoy was ending his own questioning,
    at about 2:23 p.m. 1/8/85 RT 5933, 5966-67. Deputy Sharp
    testified at trial that he questioned Josh in the emergency
    room from about 2:30 to 2:45, and then resumed questioning
    again in the CAT scan room at about 3:45. Gamundoy did not
    accompany Deputy Sharp and Josh into the CAT scan room.
    1/8/85 RT 6007-08, 6018-19. Periodically during the inter-
    5484                   COOPER v. BROWN
    view, Deputy Sharp communicated with Sergeant Arthur who
    was at the scene of the crime.
    Deputy Sharp’s report and trial testimony suggest that even
    at this early stage of the investigation, the SBCSD did not
    want to be tied down to Josh’s recollection of events. Deputy
    Sharp testified at trial that he dictated his report at 6 p.m. on
    June 5, but that he never verified that the typed version corre-
    sponded to his two or three pages of handwritten notes. He
    subsequently destroyed those notes. Deputy Sharp testified
    that he reviewed the typed report two days later, after having
    been out in the field and working on the case, and after inter-
    viewing people about the Lease house. Over the lunch recess
    during his trial testimony, Deputy Sharp altered the date on
    the report from June 6 to June 7, to make the date on his
    report correspond to the date he gave in his trial testimony.
    ER 1403; 1/8/85 RT 5991, 5994, 5998-6001.
    COOPER v. BROWN                    5485
    Volume 2 of 2
    Deputy Sharp testified at trial that he used a hand-squeeze
    method of communication with Josh. The hand-squeeze
    method, as described by Deputy Sharp, failed to provide Josh
    with an effective means of distinguishing between “no” and
    “I don’t know.” Deputy Sharp testified that a squeeze meant
    “yes.” The absence of a squeeze meant “no.” ER 1291-93.
    Deputy Sharp’s report states that Josh “wr[o]te his name and
    point[ed] to numbers for his home phone number, date of
    5486                     COOPER v. BROWN
    birth, etc. Then, all other information was done by the suspect
    [sic] squeezing my hand in order to answer yes to any ques-
    tions that I might ask.” ER 1403-04. At trial, Deputy Sharp
    could not explain how he had used the hand-squeeze method
    to elicit the first piece of information in his report that he attri-
    buted to Josh: “The victim first advised me that there were
    three white male adult subjects in the residence and he had
    been asleep.” ER 1403, 1293-99. In all likelihood, Gamundoy
    was the source of this information. But according to Deputy
    Sharp’s report, he obtained all of this information himself.
    At trial, the prosecutor dismissed the information Josh pro-
    vided to Gamundoy — that the murderers were three or four
    white men — as a mistake. The prosecutor contended that
    Josh had been confused, and that he had described the killers
    as three Mexican men who had come to the Ryen house dur-
    ing the day of June 4. Deputy Sharp’s report indicates that he
    received information about the Mexicans in the CAT scan
    room, where, according to Deputy Sharp, Josh “seemed to be
    more alert.” See also ER 1299. This part of the interview took
    place after a one hour break, during which Deputy Sharp had
    communicated with Sergeant Arthur at the crime scene.
    Deputy Sharp initiated discussion of the three Mexicans by
    asking “if there was anyone around [Josh’s] house yesterday
    that didn’t belong there,” to which Josh responded “yes.”
    Then, through yes-or-no questioning and hand squeezes, Dep-
    uty Sharp purportedly learned that “at approx. dusk there was
    an older model Chevy Impala, blue in color, four-door, no
    primer, that had pulled into the driveway of his residence.”
    The vehicle was “definitely a low rider.” Through the hand-
    squeeze method, Josh purportedly provided a detailed descrip-
    tion of the three men in the car, including precise height,
    build, hair color and length, style and color of clothing, age
    ranges, and nationalities.
    After this detailed description, Josh gave the first response
    that we know is false. Deputy Sharp wrote, “I asked the vic-
    COOPER v. BROWN                    5487
    tim if his family owned a white station wagon. The reply was
    in the negative.” ER 1404. (In fact, we know that the family
    owned a white station wagon.) The question that allegedly
    followed formed the basis for the prosecutor’s claim at trial
    that Josh was merely “confused” and had no actual recollec-
    tion of his assailants: “I then asked him if he felt these [the
    Mexicans] were the people that were in his house this morn-
    ing when everything went crazy, to which he advised in the
    affirmative.” When Deputy Sharp “again asked the victim if
    he was sure it was three Mexicans rather than three white
    male adults[,] [t]he victim advised of the three Mexicans. The
    victim then moved his shoulders and appeared to be con-
    fused.” At that point, the interview terminated as doctors took
    Josh into surgery. Deputy Sharp conceded at trial that he did
    nothing in asking his questions to try to distinguish for Josh
    between the three Mexican males and the three or four white
    males he had initially identified as his assailants. 1/8/85 RT
    6038-40.
    Over the course of the interview, Deputy Sharp spoke with
    Sergeant Arthur (who was at the crime scene) three times on
    the telephone. Despite what Deputy Sharp later wrote in his
    report, officers at the crime scene on that day — including
    Detective Hector O’Campo — learned that Josh had described
    his assailants as three white men. ER 1315. The communica-
    tions are confirmed by the police log of the attack, which was
    entered sometime between 2:06 p.m. and 2:29 p.m., describ-
    ing as suspects “three young males” driving the Ryens’ sta-
    tion wagon. ER 3688; 8/13/04 RT 73. Given the times in the
    police log, Deputy Sharp must have communicated this infor-
    mation to Arthur either before he started questioning Josh or
    within a few minutes of initiating contact. 1/4/85 RT 5973-75.
    On June 6, Sergeant Arthur assigned SBCSD Detective
    O’Campo to make contact with Josh. At trial, Detective
    O’Campo denied talking to Josh about suspects or what hap-
    pened during the attack at any point prior to a formal inter-
    view with Josh on June 14, 1983. ER 1322. Yet hospital
    5488                   COOPER v. BROWN
    personnel confirmed that Detective O’Campo did speak with
    Josh about a Mexican in a red shirt prior to that date. ER
    1302, 1307-09. Further, both Josh’s grandmother and a nurse
    on duty on June 6 testified that Detective O’Campo asked
    Josh substantive questions on that date, and that Detective
    O’Campo took notes about what he was learning. ER 1306-
    09; 1/14/85 RT 6303-11, 6326-27; 2/7/85 RT 7795-96. At
    trial, Detective O’Campo initially lied, denying that he took
    any notes on June 6. He then admitted that he destroyed his
    June 6 notes on June 8, after typing up a report. The docu-
    mentation that Detective O’Campo had submitted to the
    SBCSD as of October 1983 gives no indication he had even
    spoken with Josh on any date other than June 14. Yet Detec-
    tive O’Campo eventually admitted at trial that he had spoken
    with Josh approximately twenty times during Josh’s stay at
    the hospital, meeting with him almost every day during that
    period. ER 1316-17; 1/9/85 RT 6076-84.
    Detective O’Campo admitted at trial that sometime
    between June 6 and June 9, he became convinced in his own
    mind that Cooper was responsible for the murders. He admit-
    ted that he had a very strong desire for Cooper to be con-
    victed, and that he had more than a mere belief that Cooper
    was responsible — he knew that Cooper was responsible. 1/
    9/85 RT 6095-96. SBCSD Captain Schuyler, commander of
    the public affairs division, made clear at trial that by June 6
    or 7, after Cooper had become a suspect, the SBCSD’s pub-
    licly stated position was that Josh could not provide any use-
    ful information. 1/15/85 RT 6423-24. Captain Schuyler
    testified on cross examination that he had “probably” been the
    source of an Associated Press story reporting that Josh had
    told investigators that three Mexicans in a blue car had come
    to the Ryen house during the day before the killings. Id. at
    6422-23. No public mention was made of the fact that Josh
    had told Gamundoy, before any SBCSD detectives had
    arrived at the hospital, that the killers had been three or four
    white men.
    COOPER v. BROWN                     5489
    On June 14, 1983, Detective O’Campo interviewed Josh
    about what he remembered about the attacks. Josh was now
    able to speak. Breaking with his standard policy of tape-
    recording formal interviews, Detective O’Campo only took
    notes on what Josh said. He testified at trial that he destroyed
    his notes after writing his report the following day. 1/9/85 RT
    6097-98. At trial, Detective O’Campo conceded that he could
    think of no reason not to tape record the interview other than
    to “shape the context that the interview is presented [and]
    shape the information that is presented in the report.” 1/9/85
    RT 6103.
    Dr. Jerry Hoyle, a psychologist at the hospital, observed the
    entire June 14 interview between Josh and Detective
    O’Campo. During the interview, Dr. Hoyle took notes to
    assist with Josh’s treatment. Dr. Hoyle’s testimony at trial
    showed that Detective O’Campo’s report of the interview had
    distorted what Josh had said. For example, Josh had used the
    words “they” and “them” to refer to his assailants, while
    Detective O’Campo’s report referred to only one attacker. ER
    1434, 1330-32, 1422-28.
    SBCSD Reserve Deputy Simo guarded Josh for part of the
    time he was in the hospital. Deputy Simo testified that on the
    evening of June 15 he was playing Uno with Josh when the
    television news came on. The program displayed a photo-
    graph of Kevin Cooper, with his hair combed out in a natural
    “Afro” hairstyle. Josh stated without prompting that Cooper
    was not the person “who did it.” Within fifteen minutes, Dep-
    uty Simo called Detective O’Campo to report what Josh had
    said. ER 1338.
    As a volunteer deputy, Deputy Simo was not responsible
    for preparing reports. Rather, Detective O’Campo had that
    duty. 1/10/85 RT 6183-84. Detective O’Campo returned Dep-
    uty Simo’s call that evening. Deputy Simo testified that Dep-
    uty O’Campo told him that “he knew about that information,
    that he had talked to — there was three Mexicans that had
    5490                   COOPER v. BROWN
    come previously, apparently looking for a job, and it could
    have been, either because of the trauma it was some flashback
    on that,” and that Deputy O’Campo suggested to Deputy
    Simo that the reason Josh made the statement that it wasn’t
    Kevin Cooper was because of Josh’s trauma. 1/15/85 RT
    6404-05. No one from SBCSD took an official report from
    Deputy Simo. In May 1984, Deputy Simo saw newspaper
    articles about Josh and contacted Detective O’Campo again.
    Only then did Detective O’Campo ask Deputy Simo to come
    down to the station so that Detective O’Campo could make a
    report. 1/15/85 RT 6405-06.
    Soon after Josh made this statement to Deputy Simo, he
    made a similar statement to his grandmother, Dr. Mary How-
    ell. Cooper’s picture appeared on television. Dr. Howell asked
    Josh whether he had ever seen Cooper before. Josh responded
    no. ER 1344-45.
    Dr. Lorna Forbes is a psychiatrist specializing in treating
    children who have survived the murder of a family member.
    She began treating Josh in October 1983. On December 1,
    1983, six months after the attacks, Dr. Forbes audiotaped an
    interview with Josh using questions submitted by Cooper’s
    defense counsel. This was one of the two interviews played
    for the jury during Cooper’s trial.
    Dr. Forbes questioned whether Cooper committed the mur-
    ders. In a sworn declaration dated October 30, 1999, she
    stated that she had “kept in touch with Josh and Dr. Howell
    from our first contact and to this day.” Cooper v. Calderon,
    No. 99-71430, Exhibits to Petition for Writ of Habeas Corpus,
    Ex. 74 (S.D. Cal. 1999). While professionally and ethically
    bound to maintain any confidences Josh shared with her dur-
    ing the course of treatment, Dr. Forbes urged the court in
    1999 to pursue all opportunities to exculpate Cooper. She
    stated, “I have maintained the opinion that it would have been
    extremely difficult for one person to have committed the four
    murders and the assault on Josh. . . . It is in my mind that it
    COOPER v. BROWN                     5491
    is imperative that a DNA test is performed to assure as much
    as possible that an innocent man is not executed.” Id.
    In his tape-recorded interview with Dr. Forbes, Josh said
    that he saw a person with a “puff of hair” standing over his
    mother during the attack. He made a similar statement to his
    grandmother in May of 1984. 1/10/85 RT 6173 (in camera
    statement of Mr. Negus). It is almost certain that Cooper did
    not have a “puff of hair” at the time of the murders. It is
    undisputed that Cooper’s hair was braided when he escaped
    from prison on June 2. His hair was also braided on June 5,
    the day after the murders. A hotel clerk in Tijuana, where
    Cooper arrived at 4:30 p.m. on June 5, testified that Cooper
    had braids when he arrived. Another hotel clerk, who did not
    testify, provided a sworn statement that Cooper had his hair
    in braids. However, pictures of Cooper that appeared on tele-
    vision after the murders show him wearing a large Afro —
    that is, they show him with a “puff of hair.”
    The prosecutor excised the “puff of hair” statement from
    the audiotape that was played to the jury. The excision of the
    reference to the “puff of hair” meant that Cooper’s attorney
    did not argue to the jury that Josh could not have seen Cooper
    with a “puff of hair” during the killings, and that Josh was
    confusing what he saw on the night of the killings with what
    he saw later on television. It also meant that the prosecutor
    could plausibly represent to the jury that “Josh told the same
    basic story.”
    (c).     Josh Ryen’s Testimony at the District Court Hearing
    At the invitation of the district court, Josh Ryen delivered
    an in-person unsworn statement at the end of the final day of
    the 2004-2005 hearing on Cooper’s second habeas applica-
    tion. Josh testified on April 22, 2005:
    The first time I met Kevin Cooper, I was eight
    years old and he slit my throat. He hit me with a
    5492                   COOPER v. BROWN
    hatchet and put a hole in my skull. He stabbed me
    twice, which broke my ribs and collapsed one lung.
    ....
    Every time Kevin Cooper opens his mouth, every-
    one wants to know what I think, what I have to say,
    how I’m feeling, and the whole nightmare floods all
    over me again. The barbecue, me begging to let
    Chris spend the night, me in my bed, Chris on the
    floor beside me. My mother screams, Chris gone,
    dark house, hallway, bushy hair, everything black,
    mom cut to pieces, saturated in blood, the nauseating
    smell of blood.
    . . . . Helicopters give me flashbacks of the life
    flight and my Incredible Hulks being cut off by
    paramedics. Bushy hair reminds me of the killer.
    Silence reminds me of the quiet before the screams.
    Cooper is everywhere. There is no escaping him.
    4/22/05 RT 129-133 (emphases added).
    The district court denied Cooper’s counsel’s repeated
    requests to allow discovery concerning Josh’s memory and
    his testimony at trial. The court also denied Cooper’s coun-
    sel’s request to cross examine Josh at the habeas hearing.
    Immediately after Josh’s statement at the hearing, the district
    court required Cooper’s attorneys to make their closing argu-
    ments.
    The district court denied Cooper’s habeas application from
    the bench the same day Josh testified. The court stated:
    I do recognize that the victims who came today are
    technically under the federal law — it’s not man-
    dated or it’s analogous they’re not required to speak.
    But I did think for completion of the record, it is
    COOPER v. BROWN                       5493
    helpful for any reviewing court to at least hear their
    statements about their views of the matter. And so
    the Court recognizes your objections, but believes
    that because it’s a case of major importance, that it’s
    appropriate to give the victims a say. And it was not
    a significant period of time, but it does complete the
    record for the reviewing court.
    4/22/05 RT 182 (emphases added). Given the district court’s
    refusal to permit discovery and cross examination, Josh’s
    unsworn testimony at the end of the habeas hearing hardly
    “completed the record.”
    Instead, Josh’s testimony at the 2004-2005 hearing raised
    new questions. In its order denying Cooper’s habeas petition,
    the district court wrote that the videotape and audiotape inter-
    views benefitted Cooper because “[t]he defense also avoided
    the drama and sympathy that would have undoubtedly
    occurred had the defense called victim Josh Ryen to the stand
    in the trial and heard his firsthand recollection about a man
    with bushy hair.” Dist. Ct., 
    510 F.3d at
    1000 (citing 4/22/05
    RT 133) (emphasis added). Josh’s testimony in 2005 was the
    first time Josh made an in-court statement that he had seen
    someone with “bushy hair” during the attack. If Cooper’s
    attorneys had been permitted to question Josh in the 2005
    hearing, they could have asked him about his recollection of
    the attacker’s “bushy hair” (or “puff of hair”), and could have
    pointed out that it was extremely unlikely that the man with
    “bushy hair” on the night of June 4 — if, indeed, there was
    such a man — was Cooper.
    ii.   Drop of Blood A-41
    A drop of blood was found by Deputy Baird on the wall in
    the hall across from the master bedroom in the Ryen house.
    Chips of paint with the blood on them were taken from the
    wall and labeled “A-41.” The State introduced evidence at
    trial through SBCSD Criminologist Daniel Gregonis that his
    5494                   COOPER v. BROWN
    testing of A-41 showed that its blood characteristics matched
    those of Cooper. During deliberations the jury requested that
    Gregonis’s testimony about A-41 be read back. 2/11/85 RT
    7864-65. This was the second of only two read-backs
    requested by the jury. (The other was Josh’s audiotaped inter-
    view with Dr. Forbes.)
    There is a strong likelihood that the results of the blood
    tests performed on A-41, presented at trial, were false evi-
    dence. There is also a strong likelihood that state actors tam-
    pered with A-41 to ensure that it would generate inculpatory
    results when Cooper’s post-conviction DNA testing was con-
    ducted in 2002.
    Gregonis delayed most of his testing of A-41 until he had
    information about Cooper’s genetic profile — that is, until he
    knew what he had to match. He then delayed doing the most
    sensitive and discriminating tests of A-41 until after Cooper
    had been arrested and had the vial of Cooper’s blood (vial
    VV-2) in his lab. Without contacting Cooper’s counsel, Gre-
    gonis re-ran tests that consumed more of the limited sample
    that constituted A-41. Gregonis then tested a known sample
    of Cooper’s blood side by side on the same testing plate with
    A-41, but initially denied doing so and represented under oath
    that he had tested the samples blind. ER 761.
    When the results of Gregonis’s tests on A-41 were initially
    inconsistent with Cooper’s expert’s results for a known sam-
    ple of Cooper’s blood, Gregonis altered his lab notes and
    claimed that he had misinterpreted his results. ER 747. Here
    is some of Gregonis’s trial testimony:
    Q [by Cooper’s attorney]: Did you change your
    mind about A-41 after you learned that if your origi-
    nal call was accurate, A-41 couldn’t have come from
    Mr. Cooper?
    A [by Gregonis]:    Not immediately, no. But it was
    after. Yes.
    COOPER v. BROWN                         5495
    Q:    Prior to your learning that if your original call
    about A-41 was correct, then it couldn’t have
    come from Mr. Cooper, how many times did
    you testify on the witness stand, under oath, that
    A-41 was a B and nothing else but a B?
    A:    It is probably about three times.
    Q:    And your explanation was that it was a techni-
    cal fault on your part, you made a mistake?
    A:    Essentially, yes.
    ER 746.
    The blood sample in A-41 has had a disturbing pattern of
    being entirely “consumed” in the testing, and then reappear-
    ing in a form that could be subjected to further testing. Gre-
    gonis had initially used so much of the limited sample (or so
    he said) that when the parties finally did joint testing, they
    were forced to place the remaining tiny flakes of white paint
    in a liquid solvent to dissolve any remaining blood. The sam-
    ple was so small that their results were largely inconclusive.
    All of the chips that had had any traces of blood on them were
    discarded. ER 722-23; 12/6/84 RT 4543-47; 1/30/85 RT
    7380-87. Then, in early July 1984, “just out of curiousity [sic]
    sake, [Gregonis] . . . open[ed] the [A-41] pillbox and saw a
    very small quantity of blood remaining.” ER 722-24; 12/6/84
    RT 4548. The parties tested those remaining small specks of
    blood in October 1984, and again the results were inconclu-
    sive. 12/5/84 RT 4442-45.
    In August 1999, Gregonis checked A-41 out of the evi-
    dence storage room for one day. ER 1629, 2650-54. When
    Cooper’s post-conviction DNA testing took place in 2002, a
    “bloodstained paint chip” and “blood dust” had inexplicably,
    and conveniently, appeared in the A-41 canister. ER 790. The
    blood on that chip was tested, and Cooper’s DNA was found.
    5496                      COOPER v. BROWN
    The appearance of a blood-stained chip in 2002 is, to say the
    least, surprising, given that Gregonis had testified at trial that
    in the October 1984 testing of A-41 they had processed and
    discarded all of the paint chips with blood on them. ER 722-
    23.
    In the 2004-2005 habeas proceeding, Cooper requested
    EDTA testing of A-41. Without taking evidence on the feasi-
    bility of testing any remaining A-41 for the presence of
    EDTA, the district court rejected Cooper’s request. Dist. Ct.,
    
    510 F.3d at 948-50
    ; ER 3467; 6/29/04 RT 209.
    iii.   Pro-Ked Shoeprints
    Three matching shoeprints made by a Pro-Ked Dude shoe
    were critical evidence against Cooper at trial. According to
    testimony at trial, one was a bloody print found on a crumpled
    sheet in the master bedroom of the Ryen house. Another print
    was found on a spa cover at the Ryen house outside the mas-
    ter bedroom. The third print was found in dust on the floor
    inside the Lease house near a pool table.
    I discuss below difficulties with the evidence about whether
    Cooper was, or could have been, wearing Pro-Ked Dude
    shoes. For the moment, I am concerned only with the suspi-
    cious circumstances under which the shoeprints were purport-
    edly found.
    The most suspicious of the shoeprints are the two found at
    the Ryen house. The most incriminating was the bloody print
    purportedly found on the crumpled sheet in the master bed-
    room. At trial, only one person testified that he saw the
    bloody print while the sheet was still in the bedroom. That
    person was SBCSD Deputy Duffy. No one else claimed to
    have seen the bloody print while the sheet was still in the bed-
    room. However, Deputy Duffy had testified under oath at
    Cooper’s preliminary hearing that he had not seen the print in
    the master bedroom. ER 706, 1557. If Deputy Duffy was tell-
    COOPER v. BROWN                      5497
    ing the truth at the preliminary hearing, no one saw the bloody
    print on the sheet while it was in the bedroom. If Deputy
    Duffy was telling the truth at the preliminary hearing, he lied
    at trial.
    SBCSD Deputies Stockwell and Schecter picked up the
    crumpled sheet from the master bedroom after 5:20 p.m. on
    June 5. Neither Deputy Stockwell nor Deputy Schecter
    noticed the bloody print (if it was, indeed, there) when they
    picked up the sheet. They folded the sheet and packaged it to
    take it back to the crime lab for processing. Deputy Stockwell
    testified at trial that sometime later, in the crime lab, he
    refolded the sheet so it matched the way it had been crumpled
    on the floor of the master bedroom. He testified that this
    brought together the separated parts of a bloody Pro-Ked
    shoeprint on the sheet, which he was then able to see for the
    first time. Deputy Stockwell was unable to specify when he
    made this discovery. ER 708; 11/19/84 RT 3506-07; 11/21/84
    RT 3698. Deputy Baird, the manager of the Crime Laboratory
    where the sheet and the other critical physical evidence was
    kept, testified at trial that he had matched the newly discov-
    ered shoeprint to a Pro-Ked Dude shoe that he already had in
    the lab. (Recall that it was also Deputy Baird who found the
    drop of blood, A-41, in the hall in the Ryen house.) Soon after
    the trial, Deputy Baird was caught stealing heroin from the
    SBCSD evidence locker for his own use and for distribution
    to others. ER 1714-16.
    The other shoeprint at the Ryen house was on the spa cover
    outside the master bedroom. That print was discovered after
    some delay. Several people testified that, even after the print
    was purportedly discovered, that they had looked at the spa
    cover and had not seen the print. SBCSD Sergeant Arthur,
    head of the investigation, testified that he pointed the print out
    to Deputy Duffy at approximately 2:07 p.m. on June 5, but
    Deputy Duffy testified that he did not know about the shoe-
    print until Detective O’Campo pointed it out to him at sunset
    that day. SBCSD Sergeant Gilmore testified that he did not
    5498                    COOPER v. BROWN
    see any footwear impressions on the cover on the afternoon
    of June 5. 12/18/84 RT 5160-63; 11/15/84 RT 3363-64, 3396,
    3299. Deputy Smith, who was ordered to sketch all of the
    shoeprints on the spa cover on the morning of June 8, three
    days later, testified that she sketched all of the prints that she
    saw on the spa cover. However, none of her sketches matched
    the print of a Pro-Ked Dude shoe. An unidentified officer
    directed her back to the spa cover later on June 8. She then
    saw a Pro-Ked Dude print that she had not previously seen.
    1/22/85 RT 6869, 6871-73, 6878-79.
    The spa cover was left out in the open, moved several
    times, and stepped on by at least one deputy before it was
    taken into evidence. Then the shoeprint was destroyed, pur-
    portedly in a failed effort to lift and preserve the image. 11/
    15/84 RT 3297-98; 11/19/84 RT 3447-48.
    If the shoeprints on the sheet and the spa were planted, it
    does not matter much whether a matching Pro-Ked Dude
    shoeprint near the pool table in the Lease house was also
    planted. It is possible that Cooper was, in fact, wearing Pro-
    Ked Dude shoes while he was staying at the Lease house. If
    the print in the Lease house was not planted, this would only
    mean that the prints in the Ryen house were put on the sheet
    and the spa cover to match the print that had been found in
    the Lease house.
    But the shoeprint in the Lease house could also have been
    planted. The day before the Lease house search began and the
    Pro-Ked Dude shoeprint was discovered, the front door to the
    house had been left unlocked and other evidence was likely
    planted in the house. (See the discussion below of the hatchet
    sheath and jacket button.) 1/29/85 RT 7312. The print in the
    Lease house was discovered after some delay, after approxi-
    mately twelve people had walked through the house. 1/23/85
    RT 6956-58. Deputy Baird testified that the shoeprint had
    been “marked off so that it would not be stepped on” by mid-
    afternoon of June 7. 11/12/84 RT 4760. Deputy Smith, who
    COOPER v. BROWN                      5499
    arrived at 8 p.m. that day to draw sketches of the prints, testi-
    fied that nobody had instructed her to sketch the print that
    turned out to be the Pro-Ked Dude print. Instead, she testified
    that she discovered that print on her own after first sketching
    several other prints she was instructed to sketch and then
    walking around the house to look for more suspicious prints.
    Contrary to Deputy Baird’s testimony, Deputy Smith gave no
    indication that the Pro-Ked Dude print had been marked off
    at the time she sketched it. 11/22/85 RT 6864-72.
    iv.    Cigarettes and Tobacco
    Cigarette and tobacco evidence presented at trial connected
    Cooper to the Ryens’ white station wagon. Recall that the sta-
    tion wagon was found in a church parking lot in Long Beach,
    45 miles west of the Ryen house, seven days after the crime.
    Recall also that Cooper arrived at a hotel in Tijuana, Mexico,
    125 miles south of the Ryen house, at 4:30 p.m. the day after
    the murders.
    Cooper had left evidence in the Lease house that he was
    rolling his own cigarettes, using prison-issued “Role-Rite”
    tobacco. Tobacco consistent with Role-Rite was purportedly
    found on the floor between the front passenger seat and the
    front passenger door of the Ryens’ station wagon. Two ciga-
    rette butts were also purportedly found in the car. One
    appeared to be made with Role-Rite tobacco. The other was
    a commercial filter cigarette. One of the cigarettes contained
    evidence that it had been smoked by someone with blood type
    A (Cooper’s blood type, and, according to Gregonis, the
    blood type of about 40% of the population), but testing before
    trial did not reveal any further information. ER 748-49.
    SBCSD deputies had ample opportunity to plant incrimi-
    nating cigarette and tobacco evidence in the Ryens’ station
    wagon. Most of the cigarette butts that Cooper left behind in
    the Lease house were never processed into evidence. ER 894,
    834-35. Their whereabouts is unknown. Cooper’s own car had
    5500                   COOPER v. BROWN
    been impounded in Los Angeles after he was imprisoned at
    CIM on the burglary charge. A hand-rolled cigarette butt
    recovered from Cooper’s car had been taken into evidence,
    but it disappeared prior to DNA testing. ER 895, 897; 6/23/03
    RT 109-10, 113, 134-35. Role-Rite tobacco in a white box
    was recovered from the closet of the bedroom in the Lease
    house previously used by Katherine Bilbia (“the Bilbia bed-
    room”), but it was never weighed or otherwise measured.
    SBCSD Deputy Mascetti obtained a paper bag of assorted
    smoking tobaccos from CIM. ER 3185.
    When police formally took into custody the station wagon
    in the church parking lot in Long Beach, the rear door was
    already unlocked. The cigarette butts and tobacco in the car
    were not noted in the initial visual inspection of the car. Nor
    were they noted when the interior of the car was first pro-
    cessed for evidence. They first appear on a handwritten,
    undated, unsigned list of vehicle contents. ER 806-11, 836-37.
    Finally, the cigarette and tobacco evidence from the car has
    a perplexing history much like that of A-41. The paper from
    the hand-rolled butt supposedly found in the car had suppos-
    edly been entirely consumed in testing in 1984. At the time
    of the testing, the paper was 4 millimeters long. The paper
    “reappeared” in 2001, just in time for DNA testing. When it
    reappeared, it had grown to 7 millimeters by 7 millimeters
    square. ER 978-79, 4947, 3170. Perhaps not coincidentally,
    the hand-rolled cigarette butt from Cooper’s own car had dis-
    appeared prior to the DNA testing.
    v.    The Hatchet Sheath and Button
    On June 6, the day after the bodies were discovered, two
    deputies did a sweep through the Lease house, purportedly to
    check for possible assailants. During this sweep they were
    aware that the department had recovered a bloody hatchet
    from beside the road near the Ryen house, and that it was a
    suspected murder weapon. However, they did not report find-
    ing any evidence in the Lease house. Other deputies returned
    COOPER v. BROWN                     5501
    the following day to conduct another search. Those deputies
    discovered a hatchet sheath and a green, blood-stained button
    in the Bilbia bedroom, in plain view on the floor near the
    closet. 10/29/84 RT 2732-33; 10/30/84 RT 2838-40, 2846;
    10/31/84 RT 2905-08. (The district court erroneously wrote
    that the deputies found the sheath “near the bedroom.” Dist.
    Ct., 
    510 F.3d at 898
     (emphasis added).) The bedroom was vir-
    tually empty, which would have made the sheath (and, indeed,
    the button) very easy to see. The bedroom had no furnishings
    other than a headboard.
    One of the deputies who conducted the initial sweep of the
    Lease house denied ever setting foot in the Bilbia bedroom,
    where the other deputies later found the hatchet sheath. How-
    ever, processing of physical evidence in the Lease house later
    showed that this was a lie. Several of that deputy’s finger-
    prints were found on the inside of the closet of the Bilbia bed-
    room where the hatchet sheath and button were found. The
    deputy’s insistence that he had never set foot in the bedroom,
    the presence of the deputy’s fingerprints inside the closet of
    the bedroom, the failure of the deputy to find the sheath and
    button in plain view on the floor near the closet, and their
    “discovery” the next day strongly suggest that the hatchet
    sheath and button were planted in the Bilbia bedroom.
    Further, the button found on the floor of the Bilbia button
    was green. That means that it came from a green prison-
    issued jacket. However, Cooper testified that he was wearing
    a brown or tan prison jacket. 1/2/85 RT 5331, 5355-57; 1/3/85
    RT 5567-68. Cooper’s testimony was confirmed by Lieuten-
    ant Cornelius Shephard, who worked at the prison. On June
    2, Lieutenant Shephard was driving near the prison and saw
    Cooper. He testified that Cooper was wearing “what appeared
    to be prison jacket and blues.” 10/24/84 RT 2388. He was
    “jogging toward me . . . [and] [h]is hair was in braids. Id. at
    2389. Lieutenant Shephard’s written report stated that Cooper
    was wearing a jacket that was “brown in color.” Id. at 2408.
    5502                   COOPER v. BROWN
    In his testimony at trial, he described Cooper’s jacket as
    “brownish.” Id.
    b.    Summary
    In sum, if EDTA testing shows that Cooper’s blood was
    planted on the tan t-shirt, that showing would greatly increase
    the likelihood that the State introduced false evidence at trial.
    Indeed, in my view, the likelihood would then be so strong
    that the district court would be obliged to find that the State
    did so.
    That evidence at trial was Josh’s two recorded interviews
    in which he stated that he saw a single man or single shadow
    in the house during the murders; the drop of blood labeled A-
    41 that Gregonis testified was consistent with Cooper’s blood;
    the bloody Pro-Ked Dude shoeprints found on the sheet in the
    master bedroom and on the spa cover at the Ryen house (and
    probably the print found in the Lease house as well); the Roll-
    Rite cigarettes and tobacco found in the Ryen station wagon
    in Long Beach; and the hatchet sheath and button found in the
    Bilbia bedroom in the Lease house.
    Even with this evidence, the State’s case against Cooper
    was weak. (I will elaborate on this point below.) Without this
    evidence, the State would not have been able to obtain a con-
    viction against Cooper. That is, if this evidence was false,
    Cooper would be able to make a showing of actual innocence
    under either the standard of Schlup or of 
    28 U.S.C. § 2244
    (b)(2)(B)(ii).
    B.        Brady Claims
    Cooper contends that the State violated Brady in several
    respects. The most important of those are (1) the failure to
    provide Warden Midge Carroll’s information that Pro-Ked
    Dude shoes at the prison were not special prison-issue shoes;
    (2) the failure to provide the Disposition Report showing that
    COOPER v. BROWN                      5503
    a Senior Deputy approved the destruction of Lee Furrow’s
    bloody coveralls, and the failure to preserve and make the
    coveralls available; and (3) the failure to provide SBCSD
    daily logs showing that an apparently bloody blue shirt had
    been found, and the failure to preserve and make the shirt
    available. I consider these contentions in turn.
    1.    The Pro-Ked Dude Shoes
    At trial, the prosecution presented evidence that prints
    made by a Pro-Ked Dude tennis shoe were found on a bloody
    sheet in the master bedroom in the Ryen house, on the spa
    cover at the Ryen house, and in the Lease house. To tie those
    shoeprints to Cooper, the prosecution presented testimony by
    Michael Newberry, General Merchandise Manager for Pro-
    Keds Division of Stride Rite Corporation, that the Pro-Ked
    Dude is “a tennis shoe that is manufactured only for institu-
    tions,” is “supplied strictly for prison use within the State of
    California,” is “unavailable for retail stores within Califor-
    nia,” and is not “sold on the street.” 10/23/84 RT 2281; 2/7/85
    RT 7749, 7752. Newberry testified at trial that to the best of
    his knowledge the Pro-Ked tennis shoe is not “sold across the
    counter on a retail basis in any state in this country,” and he
    confirmed that “the only place those shoes arrive, if those
    arrived in California, is in some type of state facility.” 10/
    29/84 RT 2624, 2621-22.
    Shortly before Cooper’s scheduled execution in 2004, now-
    retired Warden Midge Carroll provided a sworn declaration in
    support of Cooper’s motion to file a second or successive fed-
    eral habeas application. She stated in her declaration, dated
    January 30, 2004, that after reading newspaper reports during
    the investigation into the Ryen/Hughes murders in 1983, she
    had conducted “a personal inquiry of the appropriate staff,
    including the deputy warden, the business manager responsi-
    ble for procurement, and the personnel responsible for ware-
    housing,” and as a result of that inquiry she had concluded
    that the investigators’ assertion that the shoeprints tied to the
    5504                    COOPER v. BROWN
    crime scene “likely came only from a prison-issue tennis shoe
    was inaccurate.” Id.; see ER 1680-82. She stated that she had
    conveyed this information to the SBCSD, but that the infor-
    mation was not provided to Cooper’s trial counsel.
    Inmate James Taylor, who worked in the recreation equip-
    ment room at the prison, testified at trial that he personally
    gave Cooper a pair of P.F. Flyer shoes, and then allowed Coo-
    per to trade those shoes in for a pair of Pro-Ked Dude tennis
    shoes just days before Cooper’s escape. ER 1689; 10/25/84
    RT 2547-48. However, in a sworn declaration, dated January
    8, 2004, submitted in support of Cooper’s motion to file a sec-
    ond or successive application, Taylor stated that he had not
    told the truth at trial. Contrary to his trial testimony, Taylor
    now stated that he had issued Cooper only one pair of shoes
    — P.F. Flyers — and that Cooper had not traded the P.F. Fly-
    ers in for a pair of Keds or any other shoes. ER 1678-79.
    An en banc panel of this court held in 2004 that Warden
    Carroll’s declaration, in combination with inmate Taylor’s
    declaration, was enough to make out a prima facie showing
    of a Brady violation. Cooper, 
    358 F.3d at 1121
    . Based on this
    prima facie showing, the en banc panel granted Cooper per-
    mission to file a second or successive federal habeas applica-
    tion.
    During the district court hearing in 2004-2005, now-retired
    Warden Midge Carroll, Stride Rite executive Don Luck, and
    inmate James Taylor all testified. I discuss their testimony in
    turn.
    a.   Warden Midge Carroll
    Warden Midge Carroll testified that in 1983 she was the
    Warden of CIM, the prison in which Cooper had been incar-
    cerated. At the district court hearing, Carroll testified that dur-
    ing the investigation of the murders, she became concerned
    that the SBCSD was describing shoeprints found at the scene
    COOPER v. BROWN                        5505
    as being from “a prison-made tennis shoe, a special shoe that
    had to come from a prison.” She testified that she knew CIM
    did not have prison-manufactured tennis shoes, and she
    directed her staff to “look into this special shoe and get back
    to me.” 6/2/04 RT 102. Her staff reported back that “there
    was no special shoe, that the shoes we had and that we issued
    were common, ordinary shoes that were commonly manufac-
    tured and sold in retail stores.” Id. at 103-04. Carroll testified
    that she then called the SBCSD and shared this information
    with one of two lead detectives, whose name she could not
    now recall. Id. at 104-05. Carroll testified that she “felt a little
    put off by” the detective’s response. Id. at 107-08. She testi-
    fied that she made a couple of additional attempts to contact
    the SBCSD about the issue, but her calls were not returned.
    Id. at 108-09.
    Carroll also testified that there were a few inaccuracies in
    her January 30, 2004 declaration. For example, she testified
    that she did not keep records of all of her contacts with the
    SBCSD about the Cooper case, and that she had not “kept
    meticulous records of all matters stemming from Mr. Coo-
    per’s case, including [her] contacts with detectives.” Id. at
    147, 172. She also testified that she did not keep records of
    her telephone conversations. Id. at 198.
    SBCSD Detective Derek Pacifico testified during the dis-
    trict court hearing that he could find no record that Carroll
    had communicated with the SBCSD. He testified that he had
    reviewed “all of the files and materials in the custody of the
    Sheriff’s Department relating to the Kevin Cooper case” and
    could not find “any indication of any contact from Ms. Car-
    roll.” 6/3/04 RT 46-47. Detective Pacifico had been with the
    department only since 1990. On cross-examination, Detective
    Pacifico conceded that he had not spoken with anyone who
    had worked on the case in 1983, and did not know more than
    two names of officers who had been involved with the investi-
    gation. Id. at 49-50. He also admitted that he was not familiar
    with the SBCSD paperwork for processing telephone calls in
    5506                   COOPER v. BROWN
    1983, and he conceded that although the files included the
    standard pink “missed your call, please call” slips, “I would
    guess that you wouldn’t take a message . . . if someone is live
    to answer the phone.” Id. at 53-54.
    b.   Stride Rite Executive Don Luck
    Stride Rite executive Don Luck also testified at the district
    court hearing. Luck had managed the Keds shoe line for forty-
    one years before his retirement. Luck’s testimony directly
    contradicted Newberry’s trial testimony that the Pro-Ked
    Dude was not “sold across the counter on a retail basis in any
    state in this country.” 10/29/84 RT 2624. Luck stated that no
    large chains on the West Coast sold Pro-Ked Dudes in the
    early 1980s, and he stated that the market for Pro-Keds was
    generally weak in the West. But he testified that Pro-Ked
    Dudes “enjoyed a good business” in retail stores in New
    York; that he “did lots of thousands of pairs in the five bor-
    oughs of New York”; and that the demand “splash[ed] on
    down to Atlanta.” 6/2/04 RT 239-40. Luck testified that smal-
    ler chains and individual stores, including those on the West
    Coast, could order Pro-Ked Dudes through the wholesale cat-
    alog. He also left open the possibility that major chains east
    of Chicago may have sold Pro-Ked Dudes. He conceded that
    the Pro Ked Dude sold to institutions was the “[s]ame shoe,
    same storage, same shipment point” as the Pro Ked Dude sold
    to retail outlets through the catalog. Id. at 239.
    Michael Newberry, who had testified as the Stride Rite rep-
    resentative at the 1983 trial, now lives and works in China.
    The district court initially proposed that the parties conduct a
    deposition on written interrogatories. 4/2/04 RT 3-4. On May
    7, 2004, Cooper’s counsel submitted to the court a list of
    twenty proposed questions for Newberry. ER 127-30. The dis-
    trict court, however, never authorized the deposition on writ-
    ten interrogatories.
    COOPER v. BROWN                   5507
    c.   Inmate James Taylor
    Inmate James Taylor testified at trial that he had issued
    Pro-Ked Dudes to Cooper just before his escape. At the 2004-
    2005 hearing in the district court, Taylor could not provide a
    version of events contradicting his trial testimony without
    risking perjury charges.
    In early May 1983, Taylor began working in the equipment
    room of the CIM gym, where he issued athletic shoes and
    other equipment to inmates, and where he had regular access
    to television. 10/24/84 RT 2466; 1/28/85 RT 7189-90. Five
    days after the murders, on June 9, 1983, two SBCSD deputies
    went to CIM and met with Taylor’s supervisor in the gym,
    correctional officer Alfred Hill. That same day, Taylor had
    heard news reports that the SBCSD was trying to connect ten-
    nis shoes with Kevin Cooper, and that Cooper was the pri-
    mary suspect. After the SBCSD deputies left, Hill told Taylor
    that the deputies had been looking into shoe impressions, and
    that the deputies had specifically mentioned a diamond-
    shaped pattern. 10/25/84 RT 2507-08, 2524-26; 1/28/85 RT
    7199-7201. Taylor went that day to Recreation Supervisor
    Skip Arjo to report that he had information about Cooper. 10/
    24/84 RT 2477-78.
    On June 10, CIM Investigators Murray and Hernandez con-
    ducted a tape-recorded interview with Taylor. In the interview
    Taylor stated that he knew Cooper from the basketball team
    and that he had given Cooper both Pro-Ked and P.F. Flyer
    tennis shoes. He stated that “[s]omehow, somewhere or
    another [Cooper] got them, come up with some Pro Keds.”
    10/25/84 RT 2555. Murray wrote a two-paragraph report of
    the interview dated June 12, 1983. The report did not mention
    that the interview had been tape recorded, nor did the report
    mention P.F. Flyers. The tape was not transcribed until early
    December 1983 and was not made available to Cooper’s
    attorneys until well after the beginning of the preliminary
    hearing. 1/28/85 RT 7183-88.
    5508                   COOPER v. BROWN
    Taylor’s testimony was the only testimony at trial that
    definitively placed Pro-Keds on Cooper’s feet. His testimony
    was not consistent with that of the correctional officer who
    provided Cooper with his tennis shoes. Cooper had a medical
    problem with one of his feet and therefore when he arrived at
    CIM he obtained a “chrono” which entitled him to wear soft-
    soled shoes. On May 3, Sidney Mason, a search and escort
    correctional officer, assisted Cooper with filling the chrono.
    Officer Mason testified that he got the shoes from “a Latin
    officer that worked in . . . Palm Hall.” 1/28/85 RT 7175-77,
    7181. Officer Mason testified that the officer came out with
    “the standard issue Keds, hi-top black type tennis shoes” size
    “either 9’s or 10’s. I remember asking two sizes in case one
    size was out.” Id. at 7175-77. Officer Mason originally told
    Cooper’s investigator, in a tape-recorded statement, that the
    shoes “had appeared to me to be Keds, a black and white high
    top type” and that “I do believe they were used because that
    would be all that they issued at that time that they had on
    hand.” Id. at 7178-79; 1/29/85 RT 7318; 2/7/85 RT 7805. At
    trial, however, Officer Mason testified that the shoes were
    new because they were in a box. 1/28/85 RT 7178-80.
    Cooper’s testimony was consistent with Officer Mason’s,
    except that Cooper consistently maintained that the shoes he
    was given were old. 1/2/85 RT 5332-41. Cooper testified that
    he did not know what type of tennis shoes they were. Cooper
    testified that it was “safe to say” that he was “not paying
    attention to the shoes at that particular time.” 1/3/85 RT 5553.
    Taylor’s testimony was critical to Cooper’s conviction. On
    March 7, 1985, a CIM internal investigator recommended to
    Warden Carroll that Taylor be given a reduction in his sen-
    tence based on his “testimony [that] implicated Mr. Cooper in
    the Chino Hills murders.” ER 376-77. On March 11, District
    Attorney Kottmeier wrote a supporting letter to Daniel
    McCarthy of the California Department of Corrections stating
    that “Mr. Taylor’s testimony [that he had given Cooper a pair
    of shoes that were available only at institutions such as CIM]
    COOPER v. BROWN                      5509
    was of critical importance both at the preliminary hearing and
    at the jury trial.” ER 5605-06. In affirming Cooper’s convic-
    tion on direct appeal, the California Supreme Court specifi-
    cally referred to Taylor’s testimony that he gave Cooper a pair
    of Pro-Ked Dude shoes shortly before his escape from prison.
    Cooper, 
    53 Cal. 3d at 797-98
    .
    Taylor’s statements during the district court habeas pro-
    ceedings have been incoherent. State investigators Derek
    Pacifico and Don Mahoney conducted a tape-recorded inter-
    view of Taylor on April 1, 2004, at Ironwood State Prison.
    FER 22-48. At various times during that interview, Taylor
    stated that Pro-Keds and P.F. Flyers are the same shoe, that
    P.F. Flyers are a particular model of Pro-Keds, that P.F. Fly-
    ers and Pro-Keds are different shoes, that P.F. Flyers are
    prison-made shoes, and that prison-made shoes are not called
    P.F. Flyers. By the end of the interview, Taylor’s story was
    that in the morning he gave Cooper a pair of P.F. Flyers,
    which are also known as standard prison-issue shoes, and that
    later in the day he allowed Cooper to exchange those shoes
    for Pro-Keds. He stated that if he could change anything in his
    sworn affidavit of January 8, 2004 (I quote this so that the
    reader can get the full flavor of Taylor’s incoherence), he
    would “distinguish the fact of what I mean right here . . . that
    I gave him a pair of pf flyers the pro keds pf flyer pro keds
    not the pf flyer not the pro keds ah the pf flyer prison issue.”
    FER 47.
    Taylor testified in the district court on June 2, 2004. He tes-
    tified that P.F. Flyer shoes are an “everyday, recreational situ-
    ation shoe” that was “an institutional-made shoe” also known
    as “PIA tennis shoes.” 6/2/04 RT 3-4. He testified that Pro-
    Keds, by contrast, were “specifically for the [basketball]
    team,” and were “a better shoe.” Id. at 3. He stated that he did
    not “understand that the [January 8, 2004] declaration stated
    that the only shoes that [he] ever provided to Kevin Cooper
    were P.F. Flyers.” Id. at 8. He testified that he signed the dec-
    laration that described the shoes as P.F. Flyers because, “The
    5510                   COOPER v. BROWN
    lango — that’s what I was trying to — that’s the only discrep-
    ancy that — from what I said 20 years ago to what I’m saying
    now to what I told her. These are also called P.F. Flyers.
    They’re — the Pro Ked P.F. Flyers — they’re — that’s just
    what they’re called. It’s a lango.” Id. at 12-13. As he was say-
    ing these words, he was pointing to a photograph of actual
    P.F. Flyers, not prison-manufactured shoes and not Pro-Keds.
    Id. at 13; ER 427.
    Taylor testified in the district court that the statements in
    his January 8, 2004, declaration had not been written on the
    paper when he signed it, and when he initialed each line. 6/
    2/04 RT at 34-35, 56-58, 69. The following day, Sandra Coke,
    the private investigator who took Taylor’s declaration, testi-
    fied that she had been very careful in ensuring that the con-
    tents of the declaration accurately reflected what Taylor
    intended to say. She testified that she wrote the declaration by
    hand in Taylor’s presence, reviewing each line with Taylor as
    she wrote it. She denied that Taylor had ever claimed that he
    sometimes referred to Pro-Keds as P.F. Flyers or that the term
    P.F. Flyers referred to a prison-made tennis shoe. She further
    testified that she again met with Taylor approximately two
    months after the initial interview, when she showed him the
    declaration and informed him that he would likely be called
    upon to testify. She stated that Taylor reread the declaration
    and stated that “everything was accurate and that he had no
    changes, that he was satisfied with the declaration as it was
    written.” 6/3/04 RT 92-99, 111.
    d.   District Court Conclusion
    The district court concluded that the failure to provide War-
    den Carroll’s information to Cooper was not a material Brady
    violation. First, the court wrote that Don Luck’s testimony
    (which contradicted Newberry’s testimony) “would not
    change the inculpatory nature of the shoeprint evidence. . . .
    [I]t was never assumed that the distribution of the shoes was
    limited to prison inmates; what mattered was that the shoes
    COOPER v. BROWN                        5511
    were linked to Petitioner.” Dist. Ct., 
    510 F.3d at 979
    . Second,
    the court wrote that “Mr. Taylor’s credible testimony at the
    evidentiary hearing corroborates his trial testimony that he
    gave Petitioner a pair of Pro-Keds Dude tennis shoes, which
    Mr. Taylor correctly identified by sight, and that Petitioner
    never returned those shoes. Mr. Taylor was simply confused
    by his interview with [Cooper’s] investigator[.] . . . [T]he
    Court concludes therefore that there was no recantation by
    Mr. Taylor and that Petitioner had a pair of Pro-Keds Dude
    tennis shoes when he escaped from CIM.” 
    Id. at 978
    .
    The district court was wrong on both counts. First, the dis-
    trict misstates the evidence at trial. That the distribution of
    Pro-Ked Dudes was “limited to prison inmates” was not
    “never assumed.” Quite the contrary — it was “assumed”;
    indeed, it was shown by uncontradicted testimony of Stride
    Rite executive Michael Newberry. Newberry’s testimony
    effectively eliminated the possibility that the shoeprints on the
    bloody sheet in the master bedroom and on the spa cover
    could have come from anyone other than an escaped prisoner.
    The California Supreme Court specifically relied on Newber-
    ry’s testimony in affirming Cooper’s conviction and death
    sentence. It wrote:
    The Stride Rite Corporation sells Pro Ked tennis
    shoes to the state for use in institutions such as CIM.
    All “Dude” tennis shoes contain the same sole pat-
    tern. The general merchandise manager for Stride
    Rite testified that this pattern is not found on any
    other shoe that the company manufactures nor, to his
    knowledge (which was extensive), on any other
    shoe. The shoes are not sold retail, but only to states
    and the federal government.
    Cooper, 
    53 Cal. 3d at 798
    .
    Second, Taylor did not give “credible testimony corro-
    borat[ing] his trial testimony.” Rather, at the district court
    5512                   COOPER v. BROWN
    hearing Taylor gave a combination of incoherent and flatly
    incredible testimony. Nor did Taylor “correctly identif[y]”
    Pro Ked Dudes “by sight.” Quite the contrary — during his
    testimony, he pointed at P.F. Flyers and said that they were
    Pro Ked Dudes.
    It is still possible that Cooper received Pro-Ked Dudes
    from Officer Mason. But neither Officer Mason nor Cooper
    can say for certain whether those shoes were, or were not,
    Pro-Ked Dudes. The Brady evidence from Warden Carroll,
    the testimony of Don Luck, as well as Taylor’s obvious lack
    of credibility, substantially increase the probability that Coo-
    per was not wearing Pro Ked Dudes during the time he was
    in the Lease house.
    2.   The Bloody Coveralls
    As described in detail above, Lee Furrow’s bloody cover-
    alls were important evidence indicating that Furrow and two
    or three others may have been the killers. Diana Roper turned
    over the coveralls to SBCSD Deputy Eckley on June 9, 1983.
    At that time, she told him that she had reason to believe that
    Furrow had committed the murders. Deputy Eckley logged
    the coveralls into evidence at the Yucaipa Station of SBCSD.
    Less than six months later, Deputy Eckley destroyed the cov-
    eralls without telling Cooper’s counsel.
    Deputy Eckley testified at trial that he acted alone in dis-
    carding the bloody coveralls, and that in so doing he failed to
    follow SBCSD procedures for disposing of property. ER
    4985-89. On December 1, 1998, Cooper’s investigator discov-
    ered a Disposition Report for the coveralls dated December 1,
    1983. This report had not previously been disclosed to Coo-
    per. ER 3010, 4934. The report stated “destroyed no value
    coveralls.” The initials “KS” were written on the bottom right
    side of the report. In February 1999, Cooper’s investigator
    learned that “KS” was likely to have been Ken Schreckengost,
    a now-retired Senior Deputy in the property division of
    COOPER v. BROWN                    5513
    SBCSD. The investigator contacted Deputy Schreckengost,
    but Schreckengost refused to speak with him. Schreckengost
    told him, “I don’t like [Cooper]. I won’t help him,” and hung
    up the phone. ER 3113. Schreckengost told a second investi-
    gator, “I know everybody who worked the case, I know all
    about it but I’m not going to say nothing. . . . They should
    have shot that son of a bitch when he escaped.” ER 4902.
    Deputy Eckley also refused to speak with Cooper’s investi-
    gators. He told one investigator that he “could not add any-
    thing to the information which was written on the disposition
    report.” ER 4917. Eckley told another investigator that he
    would not “say anything without his attorney, unless he is
    ‘drug into court.’ ” ER 4939. He added that although he was
    retired from the SBCSD, “he owed them allegiance.” ER
    4790.
    An investigator for Cooper contacted Deputy Schrecken-
    gost in February 2005. This time, perhaps because the investi-
    gator was a former police officer, Deputy Schreckengost
    confirmed that the initials “KS” on the Disposition Report
    were his. He told the investigator, “If you are trying to help
    [Cooper], you are in the wrong place.” Deputy Schreckengost
    stated that if proper procedures were followed no property
    associated with a murder would ever be destroyed, and that
    property with blood on it would have routinely been tested at
    least for blood type. ER 4787-89. We now know, based on
    Deputy Schreckengost’s 2005 statement, that, contrary to
    Eckley’s testimony at trial, Eckley did not discard the cover-
    alls on his own. Instead, his action had been approved by his
    superior officer, Deputy Schreckengost. The fact that Deputy
    Schreckengost had approved the disposal of the bloody cover-
    alls was material information that should have been revealed
    to Cooper under Brady.
    This information would have impeached Deputy Eckley’s
    testimony that he acted alone, and would have identified Dep-
    uty Schreckengost as the SBCSD officer responsible for the
    5514                   COOPER v. BROWN
    disposal. The SBCSD’s lead investigator testified at trial,
    based on Deputy Eckley’s testimony, that the reason the cov-
    eralls were never processed was “neglect.” 12/19/84 RT 5261.
    We now know, based on Deputy Schreckengost’s 2005 state-
    ment that his initials were on the Disposition Report, that a
    responsible officer of the SBCSD deliberately destroyed
    material evidence that should have been provided to Cooper.
    If the prosecution had disclosed to Cooper prior to trial that
    Deputy Schreckengost had authorized the destruction of Fur-
    row’s bloody coveralls, his counsel could have questioned
    him at that time about the circumstances under which he had
    put his initials on the Disposition Report. More than twenty
    years later, during the district court hearing in April 2005,
    Deputy Schreckengost responded “I don’t remember,” “I
    don’t know,” or “I have no idea” more than fifty times. 4/1/05
    RT 2-71. He purported not to remember anything about the
    policy manual or the rules for handling evidence. When
    asked, “Do you remember anything about signing off on this
    form?” he replied, “You know I don’t. It’s been 23 years.” Id.
    at 31.
    Deputy Schreckengost and Deputy Eckley are good friends.
    ER 4788. They rode to the district court hearing together.
    Both of them understood the value of a fading memory. In
    response to one question during the hearing in the district
    court, Deputy Schreckengost replied, “You’re talking about
    stuff that happened many years ago. I can’t remember what
    I had for lunch yesterday.” 4/1/05 RT 68. When his turn
    came, Deputy Eckley similarly testified, “I can’t remember
    what I ate two weeks ago.” Id. at 103.
    In April 2005, Eckley testified in the district court that he
    had thought the story told by Roper when she gave him the
    bloody coveralls “was a goofy story.” He testified, “So in
    other words there’s holes in the story. When I run it by the
    supervisors, we don’t think the story had much credibility.
    COOPER v. BROWN                       5515
    But I was told to take [the coveralls] as evidence and pass
    them on which is what I did.” Id. at 151.
    However, in a May 26, 1984, audiotaped interview with
    Cooper’s investigator, Deputy Eckley said that he believed
    Roper’s story: “She said that this guy had come from prison
    and that he has a rap for crimes like murder. I would believe
    that part, because the Kellison girls have a tendency or pro-
    pensity to marry, live and shack-up with guys from prison.”
    He also stated, “And then with my relationship with the
    Kellison’s/Roper family I know their involvement in crime, as
    far as committing murders as well as giving up murderers.”
    The investigator asked, “They’ve done this in the past?” Dep-
    uty Eckley replied, “Oh, yeh, they’ve given very good infor-
    mation on a murder before. One of their sons try to commit
    murder, tried to kill me once. You know, I know the family.
    So knowing that what Diana said might have validity to it I
    called the supervisor. . . . He told me to take the coveralls . . .
    and place [them] in evidence.” ER 5000-01.
    One week after this 1984 interview, Deputy Eckley was
    interviewed by SBCSD Detective Woods, who was eager to
    learn what Deputy Eckley had said to Cooper’s investigator.
    This conversation, which Deputy Eckley did not know was
    being recorded, strongly suggests that the SBCSD’s plan was
    to pass off Deputy Eckley’s actions as those of an ignorant
    deputy. Deputy Eckley told Detective Woods:
    [T]he impression that I tried to leave [Cooper’s
    investigator] Forbush with is that I am so low on the
    Totem pol[e] that homicide could come in there and
    start a fire and I wouldn’t even know it. Okay. [A]nd
    ah you know I’m just a field patrol officer took the
    stuff in, put it in evidence and notified my superiors.
    And whatever homicide and my superiors did I don’t
    really know, I’ve heard you know from hearsay I
    understand they were contacted, but I don’t know.
    Okay. And what I think they are going to do, is try
    5516                   COOPER v. BROWN
    and get me on the stand and put a shadow of doubt
    on homicide investigation that they did not use this
    lead. That’s what my impression is I don’t know if
    that’s true or not.
    ER 4932, 4967.
    Based on the recently discovered Disposition Report, and
    on Deputy Schreckengost’s 2005 admission his initials were
    on it, we now know that the prosecution committed two
    Brady violations. First, the prosecution should have turned
    over to Cooper’s attorneys a copy of the Disposition Report.
    Second, the State should have preserved and made available
    the coveralls themselves. These two failures severely ham-
    pered Cooper’s attorneys in their effort to show that Cooper
    was innocent, and that Lee Furrow and his companions were
    the true killers.
    3. Blue Shirt and SBCSD Daily Logs
    On June 6, 1983, at 2:41 p.m., a woman unconnected with
    the investigation, Laurel Epler, called the SBCSD and
    reported that she had found a blue shirt beside Peyton Road
    that possibly had blood on it. (Recall that the tan t-shirt was
    also found beside Peyton Road, which suggests that the killers
    may have decided to shed some of their bloody clothing at
    roughly the same time.) SBCSD Deputy Field was dispatched,
    and he recovered the shirt that day. ER 3703. The State’s con-
    sistent theory of the case was that Cooper had acted alone. If
    Cooper could have introduced evidence of a second bloody
    shirt found beside Peyton Road, this would have been virtu-
    ally conclusive evidence that there had been more than one
    killer.
    Cooper’s trial counsel, David Negus, maintains that he was
    never given, and never saw, the blue shirt or any document
    referring to it. At the 2004-2005 hearing in the district court,
    the State produced SBCSD daily logs that contained a five-
    COOPER v. BROWN                      5517
    line reference to the blue shirt. The log for June 6 records the
    following at 14:41 (2:41 p.m.): “On Payton and Glenridge,
    . . . Laurel Epler reports finding a blue shirt that possibly has
    blood on it. UNIT 21D14 HANDLING. . . . EVIDENCE
    PICKED UP.” ER 3703. The State contended at the district
    court hearing that these daily logs were handed over to Coo-
    per’s counsel before the 1983 trial. Negus maintains that he
    was never given the daily logs. However, the district court
    refused to allow Negus to testify at the 2004-2005 hearing to
    rebut the State’s contention that it had given the logs to him.
    It is almost certain that the State did not give the daily logs
    to Cooper’s counsel before trial. District Attorney Kottmeier
    testified at trial that documents turned over to the defense are
    stamped with a page number “relating the number of discov-
    ery when that page was given to the defense.” 1/6/85 RT
    6555. The daily logs introduced into evidence at the 2004-
    2005 hearing have no such page numbers. At the 2004-2005
    hearing, Assistant District Attorney Kochis testified that doc-
    uments obtained by means of a subpoena would not go
    through the D.A.’s office, and therefore would not have page
    numbers. 8/13/04 RT 182-84. But the State has been unable
    to produce any record indicating that the logs were turned
    over to Cooper’s counsel pursuant to subpoena.
    Other evidence also indicates that no part of the daily logs
    was disclosed to Cooper before trial. Evidence that a second
    shirt stained with blood had been found near the place where
    another blood-stained shirt had been found, would have criti-
    cally undermined the prosecution’s theory that there was a
    single assailant. If Negus had had such evidence, he obviously
    would have used it at trial. Yet there is no mention, at any
    point during the trial, of a blue shirt, possibly stained with
    blood, that had been found beside Peyton Road shortly after
    the murders.
    At the 2004-2005 hearing in the district court, the prosecu-
    tion argued, and the district court concluded, that the blue
    5518                   COOPER v. BROWN
    shirt found by Laurel Epler was the same shirt as the tan t-
    shirt recovered by Deputy Field. There is no entry in the
    SBCSD daily log mentioning the tan t-shirt. The district court
    relied on this fact to conclude that the one daily log entry
    mentioning a shirt must correspond to the tan t-shirt. There is
    absolutely no basis for the district court’s conclusion.
    The district court disregarded the procedures used by the
    SBCSD. The daily log, on which Epler’s call was recorded,
    is created based on incoming telephone calls. The call intake
    person at the SBCSD takes a call, fills out a sheet, and hands
    that sheet to a dispatcher who then sends a deputy to respond
    to the call. After the deputy has responded, the dispatcher
    records the disposition of the call, and the sheet is used to
    complete the entry in the daily log. Id. at 57-59. Laurel Epler
    called in to report the shirt, and there is therefore an entry in
    the daily log recording that call. Epler testified in the 2004-
    2005 district court hearing that she recalled that the shirt was
    blue and that the location beside Peyton Road where she
    found the shirt was different from the location where the tan
    t-shirt was found on the following day. 8/26/04 RT 140-48,
    154-55, 161, 165, 187, 201-03.
    Deputy Field reported that on the afternoon of June 7,
    1983, he participated in a search for physical evidence in the
    Peyton Road area. He testified that during this search, some-
    one — possibly Sergeant Arthur — brought a tan t-shirt and
    an orange towel to his attention. Deputy Field took them into
    custody at 5:30 p.m. that day. ER 908, 2780-81. Because the
    tan t-shirt was discovered by a SBCSD officer during the
    course of a search, there was no telephone call to the SBCSD
    dispatcher about that shirt. There was thus no way that a refer-
    ence to the tan t-shirt could appear in the daily log.
    The district court denied Cooper the opportunity to develop
    the record to learn whether the SBCSD maintained any
    records of which documents it handed over to the defense
    prior to trial. Further, the district court denied Cooper the
    COOPER v. BROWN                     5519
    opportunity to conduct discovery to learn whether any other
    documents relating to the blue shirt, such as evidence tags and
    radio logs, were not disclosed prior to trial. See 8/13/04 RT
    198-201.
    However, even with the record in its current state, it is now
    clear based on the daily logs that the state has just revealed
    that there were two shirts, a blue shirt and a tan t-shirt. The
    prosecution committed a Brady violation in not turning over
    a copy of the SBCSD daily logs that recorded the discovery
    of the blue shirt. The prosecution committed a further Brady
    violation by not making the blue shirt available to Cooper’s
    attorneys.
    4.   Summary
    The en banc panel granted permission to file a second or
    successive application for habeas corpus based on a prima
    facie showing that the State violated Brady in not providing
    to Cooper Warden Midge Carroll’s information about the ten-
    nis shoes issued by the prison. Now, after the 2004-2005 dis-
    trict court hearing, we know that there were numerous Brady
    violations. First, the State violated Brady by not providing
    Warden Carroll’s information. Second, the State violated
    Brady by not providing a copy of the Disposition Report with
    Deputy Schreckengost’s initials on it, and by not preserving
    and making available Lee Furrow’s bloody coveralls. Third,
    the State violated Brady by not providing a copy of the
    SBCSD daily logs showing Epler’s call reporting that she had
    found a blue shirt, and by not preserving and making the shirt
    available.
    Given the weakness of the evidence against Cooper, if the
    State had given Cooper’s attorneys this exculpatory evidence
    it is highly unlikely that Cooper would have been convicted.
    Thus, based on the State’s Brady violations, Cooper would be
    able to make a showing of actual innocence under either the
    standard of Schlup or of 
    28 U.S.C. § 2244
    (b)(2)(B)(ii).
    5520                    COOPER v. BROWN
    III.    Response to the Concurrence in the Denial of the
    Petition for Rehearing En Banc
    The concurrence in the denial of the petition for rehearing
    en banc misunderstands the bases for my dissent; makes
    incorrect and misleading statements about the evidence in the
    trial record; and mistakenly states that the evidence of Coo-
    per’s guilt at trial was “overwhelming.”
    A.   Misunderstanding the Bases for the Dissent
    The concurrence argues that I am merely rehashing evi-
    dence and issues that have long been known to everyone
    (“from day one”), and that I have “improperly marshal[ed] the
    facts in the light most favorable to Kevin Cooper.” Conc. at
    5534. In part, this is a legal argument, contending that Cooper
    has not satisfied the standard for a second or successive
    habeas petition under § 
    28 U.S.C. § 2244
    (b)(2). In part, this
    is an argument almost invariably made in long-running crimi-
    nal cases, contending that so many courts have gone over this
    case so many times that we can be confident that justice has
    been done.
    On the contrary, there are two important claims, based on
    newly available evidence, in Cooper’s current application for
    habeas corpus. Both claims satisfy the requirements of
    § 2244(b)(2). And both claims seriously undermine any confi-
    dence that justice has been done.
    The first claim is that the State introduced false evidence at
    trial in violation of Mooney and Napue. The en banc court
    directed the district court to test Cooper’s blood on the t-shirt
    for EDTA. The district court flouted that direction. If the
    EDTA testing had been properly performed, there would
    likely be important new evidence of evidence tampering by
    state actors. Even based on the EDTA testing that has so far
    been performed, there is strong evidence that Cooper’s blood
    was planted. Further, new evidence of evidence tampering
    COOPER v. BROWN                      5521
    appeared during the district court hearing when it was discov-
    ered that vial VV-2, which was supposed to contain only Coo-
    per’s blood, contained blood with the DNA of two or more
    people. The district court refused to acknowledge the signifi-
    cance of this new evidence, and refused to allow any investi-
    gation into the circumstances that might have led to the blood
    of a second person being placed in vial VV-2.
    Evidence available from the EDTA testing that has been
    done so far, and new evidence of evidence tampering con-
    nected with vial VV-2, substantially increases the likelihood
    that the State presented false evidence at trial in violation of
    Mooney and Napue. Evidence that would be available if fur-
    ther EDTA testing were performed may well increase that
    likelihood still further. The false evidence at trial includes
    Josh’s two recorded statements, the State’s analysis of the
    blood contained in A-41, the Pro Ked Dude shoeprints pur-
    portedly found in the Ryen and Lease houses, the cigarettes
    and tobacco purportedly found in the Ryens’ station wagon,
    and the hatchet sheath and button purportedly found in the
    Bilbia bedroom in the Lease house.
    The second claim is that the State failed to reveal exculpa-
    tory evidence in violation of Brady. Cooper’s Brady claims
    based on new evidence are the failure to reveal Warden Car-
    roll’s information that the prison issued no shoes to prisoners
    that were available only in prisons; the failure to turn over the
    Disposition Report showing that Senior Deputy Schrecken-
    gost approved Deputy Eckley’s discarding of Lee Furrow’s
    bloody coveralls; and the failure to turn over daily logs show-
    ing the SBCSD took into custody a blue shirt that possibly
    had blood on it, and that there had been a report to the
    SBCSD shortly after midnight on the night of the murders of
    a white station wagon with wood sides, carrying three young
    males. The new evidence showing the second and third Brady
    violations casts new light on what had previously been known
    — that Deputy Eckley had discarded the bloody coveralls
    without testing them and without alerting Cooper’s attorneys,
    5522                   COOPER v. BROWN
    and that the SBCSD contended that a bloody blue shirt was
    never taken into custody. Based on the new evidence, we now
    know that the failure to make available to Cooper the bloody
    coveralls and blue shirt were also Brady violations.
    If either of the above claims is valid, Cooper has a newly
    available claim of innocence under Schlup or 
    28 U.S.C. § 2244
    (b)(2)(B). Once such a claim of innocence is properly
    before the district court, all of the evidence in the record must
    be considered, including all of the evidence previously known
    to the parties. See Carriger, 
    132 F.3d at 478
    . Rather than “im-
    properly marshal[ing] that evidence in the light most favor-
    able to . . . Cooper,” I have described that evidence
    scrupulously and at length.
    B.   Incorrect and Misleading Statements
    The concurrence makes incorrect and misleading state-
    ments about the evidence in the trial record. The statements
    include the following:
    (1) The concurrence states that Cooper “admitted being
    [in the Lease house] within an hour of the murders.” Conc. at
    5541. This statement is incorrect. Sunset on June 4, 1983, was
    at 7:59 p.m. 1/22/85 RT 6874. Cooper testified that he waited
    until dark to leave the Lease house, and that he left the house
    right after finishing a telephone call to his former girlfriend,
    Diane Williams. 1/7/85 RT 5828-31. Telephone records show
    that the call to Williams ended at 8:30 p.m. It is impossible
    to fix a precise time for the murders, but it is likely that they
    took place quite late that night. The Ryens and their house
    guest, Chris Hughes, had gone to a barbeque that evening.
    According to a neighbor, they returned to the house sometime
    between 9:00 and 9:30 p.m. 10/30/84 RT 2758. Josh
    recounted that after the return to the house he and Chris
    stayed awake talking and looking at magazines, while his
    father stayed up watching television. 12/13/84 RT 4952-54.
    COOPER v. BROWN                     5523
    Josh recounted that he and Chris were wakened from their
    sleep.
    (2) The concurrence states that “a bloodstained rope simi-
    lar to the one in the Ryen’s driveway” was found in the Lease
    house. Conc. at 5541, 5542. This statement is highly mislead-
    ing. There were two different ropes. One was in the driveway
    of the Ryen house. The other was in the closet of the Bilbia
    bedroom in the Lease house. The parties stipulated that “one
    of the ropes had a center cord and the other did not.” 2/5/85
    RT 7694.
    (3) The concurrence states that there were “signs of blood
    in the bathroom.” Conc. at 5541. This statement is highly mis-
    leading. A large horizontal band on the sides of the shower in
    the bathroom of the Bilbia bedroom tested positive for
    Luminol. For reasons explained above, the likely basis for the
    positive test is the bleach used by Katherine Bilbia in cleaning
    the shower.
    (4) The concurrence states that there were “hairs consis-
    tent with the victims’ hair in the drain” in the Lease house.
    Conc. at 5541. This statement is misleading. The hairs found
    in the drain were examined only for shape and color. They
    were never subjected to mitochondrial DNA testing. Further,
    we know that the Ryen children’s maternal grandmother had
    owned the Lease house a few years before, and that the chil-
    dren had stayed with her in the house. Indeed, Josh specifi-
    cally testified that he had taken showers in the house. 12/
    13/84 RT 4934-35.
    (5) The concurrence states that “[t]he prosecution never
    suggested that Pro-Keds [sic] Dudes were only distributed to
    prisons.” Conc. at 5541. This statement is incorrect. In its
    opening statement to the jury, the prosecution emphasized
    that the shoes “were supplied strictly for prison use within the
    State of California and were unavailable for retail stores
    within California.” 10/23/84 RT 2281. Michael Newberry, a
    5524                   COOPER v. BROWN
    Stride Rite executive, then testified at trial that Pro Ked
    Dudes were sold only to prisons and institutions. In its closing
    statement, the prosecution again emphasized that Cooper
    wore a “Pro Ked tennis shoe, a tennis shoe that you can’t pur-
    chase in a store anywhere in this country; a tennis shoe that
    is manufactured only for institutions; a tennis shoe such as
    this which was sent to state prison in Chino[.]” 2/7/85 RT
    7749. It continued, “Mike Newberry knows these aren’t sold
    on the street.” 2/7/85 RT 7752.
    (6) The concurrence states that Cooper “had a prison-
    issued jacket with buttons like the button found on the rug in
    the bedroom of the Lease house where he stayed.” Conc. at
    5541. This statement is highly misleading. The problem lies
    in its use of the word “like.” As explained above, the button
    found in plain view near the closet in the Bilbia bedroom was
    probably planted by the SBCSD. One of the reasons for
    believing that the button was planted was that it was not the
    same color as the buttons on Cooper’s prison jacket. The but-
    ton found in the bedroom was green, and would have come
    from a green prison-issued jacket. There was uncontradicted
    testimony at trial from Cooper, as well as corroborating testi-
    mony from a prison official, that Cooper had a brown or tan
    prison jacket.
    (7) The concurrence states that “Cooper’s expert agreed
    that the person who deposited A-41 on the wall of the hallway
    in the Ryen house was African-American.” Conc. at 5541.
    This statement is highly misleading. Cooper’s expert, Dr.
    Edward Blake, testified at trial that test results obtained from
    blood spot A-41 indicated that blood in the spot came from
    someone of African-American descent. 1/30/85 RT 7405,
    7409. However, when he so testified Dr. Blake was relying on
    tests performed on A-41 by SBCSD criminologist Daniel Gre-
    gonis. As discussed above, there is a strong likelihood that
    Gregonis falsified the results of those tests. During trial, Coo-
    per’s counsel forced Gregonis to admit that earlier in the trial
    COOPER v. BROWN                     5525
    he had repeatedly testified untruthfully about his test results
    for A-41. ER 746.
    C.   Incorrect Statement that Evidence at Trial of Cooper’s
    Guilt Was “Overwhelming”
    The concurrence embraces the opinion of the three-judge
    panel denying Cooper’s application for habeas corpus. Conc.
    at 5533-34. The last paragraph of that opinion states that the
    “evidence of Cooper’s guilt was overwhelming.” Cooper, 
    510 F.3d at 887
    . This is incorrect. In fact, the evidence of Coo-
    per’s guilt at trial was quite weak. That weakness led to a vio-
    lation of Cooper’s constitutional right to due process under
    Beck v. Alabama, 
    447 U.S. 625
     (1980). The evidence against
    Cooper at trial was so weak that his counsel thought there was
    a good chance the jury would acquit him outright. The trial
    judge indicated that, in his view, the evidence merited a
    second-degree as well as a first-degree murder instruction. He
    said, just before the end of trial, “Just for you to chew on, it
    appears to me that we’re going to have to have instructions on
    Murder One and Two but not manslaughter.” 1/24/85 RT
    7135. He tried to persuade Cooper’s attorneys to accept a sec-
    ond degree instruction, saying that omitting a second degree
    instruction “may well inure to you[r] detriment. They could
    find first degree, possibly. Otherwise they might only find
    second degree.” 1/28/85 RT 7264.
    But because the evidence against Cooper was so weak,
    Cooper’s attorney thought it worth a gamble. In order to elim-
    inate the possibility of a compromise verdict and to increase
    the chance of acquittal, Cooper’s attorneys refused to accept
    a second degree murder instruction:
    Mr. Negus [Cooper’s attorney]: Mr. Cooper and I
    both agreed that we don’t want a second degree
    instruction. Correct
    Mr. Cooper:    That’s true.
    5526                   COOPER v. BROWN
    The Court:    What this does, among other things,
    Mr. Cooper, it prevents the jurors from
    compromising, I suppose.
    Mr. Negus:     That’s what we don’t want. . . . We
    want them to go on the testimony. It is
    first degree or it is nothing.
    1/28/85 RT 7264. The instructions that went to the jury con-
    tained only a first degree murder instruction.
    The jury deliberated for seven days before returning a
    guilty verdict, and deliberated for four more days before
    returning a death penalty verdict. Many years later, five jurors
    wrote to Governor Schwarzenegger asking that he grant clem-
    ency. One of those jurors wrote:
    There are so many unanswered questions that we
    may never know. Why did Josh not recognize Mr.
    Cooper? Why were there no fingerprints found
    where the evidence showed there should have been?
    Why wasn’t information about the station wagon fol-
    lowed up on? What weren’t the three Mexicans
    located and interviewed? Why with so much blood,
    was only one drop of Mr. Cooper’s blood found?
    Why did the prosecution cover up evidence? Why
    was the jury not shown the photograph of Jessica
    Ryen clutching hair? Why wasn’t the hair tested?
    Why wasn’t the jury told about the convicted mur-
    derer’s bloodied coveralls turned into the police?
    Why did the police destroy those coveralls? Why
    wasn’t the found beer can ever tested for saliva?
    These are just some of the many questions I have
    had over the years.
    App. in Suppt. of Mo. to File S. or S. Application, Vol. 7, tab
    145.
    COOPER v. BROWN                        5527
    In Beck v. Alabama, the Supreme Court held that due pro-
    cess requires that a lesser-included offense instruction be
    given in a capital case:
    [W]hen the evidence unquestionably establishes that
    the defendant is guilty of a serious, violent offense—
    but leaves some doubt with respect to an element
    that would justify conviction of a capital offense—
    the failure to give the jury the “third option” of con-
    victing of a lesser included offense would seem inev-
    itably to enhance the risk of an unwarranted
    conviction. Such a risk cannot be tolerated in a case
    in which the defendant’s life is at stake.
    
    447 U.S. at 637
     (emphasis added). Beck requires the trial
    court to instruct sua sponte on a lesser-included offense in a
    capital case when evidence would support such a verdict. Cla-
    bourne v. Lewis, 
    64 F.3d 1373
    , 1379-80 (9th Cir. 1995).
    “Waivers of constitutional rights not only must be volun-
    tary but must be knowing, intelligent acts done with sufficient
    awareness of the relevant circumstances and likely conse-
    quences.” Brady, 397 U.S. at 748 (emphasis added). The
    requirement of a knowing and intelligent waiver extends to
    Beck claims. Spaziano v. Florida, 
    468 U.S. 447
    , 457 n.6
    (1984).
    Cooper waived his right under Beck to a lesser-included
    second degree murder instruction only after Negus, his coun-
    sel, misinformed him about the law and about the “likely con-
    sequences” of foregoing that instruction. At the end of the
    trial, the judge stated that he was “duty bound” under the law
    to instruct on second degree murder, despite Cooper’s attor-
    ney’s desire to forgo the instruction. The following open-
    court colloquy then took place, during which Cooper waived
    his right to a second degree instruction:
    Mr. Negus (defense counsel): . . . Mr. Cooper and I]
    have discussed between ourselves that the only role
    5528                   COOPER v. BROWN
    that I could see a second degree verdict playing in
    this particular case is to give a jury a compromise
    verdict.
    I also informed you [i.e., Cooper] that as far as the
    eligibility for the penalty phase in this particular case
    is concerned, according to the present law as it is
    interpreted by—at least by most of the courts, it says
    in the statute that a verdict of second degree is the
    same as a first degree if you find two second
    degrees, then we’re into the penalty phase. In order
    not to encourage a compromise verdict, I believe that
    it is in your best interest to only give the jury two
    choices: First degree or nothing. Do you agree to do
    that?
    Mr. Cooper: Yes.
    The Court: Do you join in the waiver?
    Mr. Negus: I certainly do.
    The Court: All right, I accept it.
    2/5/85 RT 7711-12; ER 260-261 (emphasis added). Thus,
    Cooper waived his second-degree murder instruction based on
    his counsel’s statement that he was death-eligible whether the
    jury convicted him of first-degree or second-degree murder.
    If Negus had been right about California law, it made sense
    to waive the second-degree murder instruction. But Negus
    was wrong. A conviction for second-degree murder can never
    support the death penalty under California law, no matter how
    many victims there are. On Cooper’s direct appeal, the Cali-
    fornia Supreme Court acknowledged that his trial counsel’s
    statement of California law was wrong. Cooper, 
    53 Cal. 3d at 828
    .
    COOPER v. BROWN                      5529
    Because Cooper was misinformed by his attorney about the
    consequences of foregoing the second-degree murder instruc-
    tion, his waiver of his due process right under Beck was
    clearly invalid. Cooper specifically cited Beck in his first fed-
    eral application for habeas corpus. However, the district court
    failed to mention Cooper’s Beck claim in denying his first
    application. On appeal, the three-judge panel of this court
    rejected his Beck claim by a vote of two to one. Cooper, 
    255 F.3d at 1110-11, 1114-15
    . One of the two judges held that
    Cooper had failed properly to raise the claim in his applica-
    tion, even though he had specifically cited Beck. The other
    held that he had properly raised the claim, and that his due
    process rights under Beck had been violated, but that the error
    was harmless. The third judge voted to grant the writ. He
    would have held that Cooper properly raised his Beck claim,
    that his due process rights under Beck had been violated, and
    that the error was not harmless. An en banc call of the panel’s
    decision failed.
    We cannot now go back to revisit our earlier decision on
    Cooper’s due process claim under Beck, for it was raised and
    rejected in his earlier federal habeas application. 
    28 U.S.C. § 2244
    (b)(1). But the background of his Beck claim, and the
    fate of that claim, make two things clear. First, contrary to
    what the concurrence says, the evidence against Cooper at
    trial was not “overwhelming.” On the contrary, it was so weak
    that it led Cooper’s counsel to gamble on getting an outright
    acquittal. Second, contrary to what the concurrence suggests,
    there is no way to be confident that justice has been done in
    this case.
    IV.   Conclusion
    Doug, Peggy and Jessica Ryen, and Chris Hughes, were
    horribly killed. Josh Ryen, the surviving victim, has been
    traumatized for life. The other members of the Ryen family,
    and the surviving members of the Hughes family, have also
    5530                   COOPER v. BROWN
    been traumatized for life. The criminal justice system has
    made their nightmare even worse.
    San Bernardino County Sheriff’s Department investigators
    were confronted with a horrifying multiple murder, far worse
    than any that had previously occurred in the county. They had
    an obvious suspect, an escaped prisoner who had stayed for
    two days at a house 125 yards away from the murder victims.
    They were under heavy pressure from the news media from
    the very first moment of their investigation. They drew what
    seemed, at the beginning, a sensible conclusion — that Kevin
    Cooper, the escaped prisoner, was the murderer. They drew
    that conclusion by the end of the first day of their investiga-
    tion, and from that time forward they organized their investi-
    gation around it.
    Once SBCSD investigators drew that conclusion, they
    manipulated and planted evidence in order to convict Cooper.
    In the course of their investigation, they discounted, disre-
    garded, and discarded evidence pointing to other killers. Their
    decision to close their eyes early in the investigation to the
    possibility that someone other than Kevin Cooper might be
    guilty has led us to the situation in which we find ourselves
    today.
    Unfortunately, the district court made things worse. After
    our en banc panel granted Cooper permission to file a second
    habeas application, the district court obstructed and impeded
    Cooper and his lawyers in almost every way imaginable.
    Kevin Cooper has now been on death row for nearly half
    his life. In my opinion, he is probably innocent of the crimes
    for which the State of California is about to execute him. If
    he is innocent, the real killers have escaped. They may kill
    again. They may already have done so.
    We owe it to the victims of this horrible crime, to Kevin
    Cooper, and to ourselves to get this one right. We should have
    COOPER v. BROWN                      5531
    taken this case en banc and ordered the district judge to give
    Cooper the fair hearing he has never had.
    WARDLAW, Circuit Judge, dissenting from the denial of
    rehearing en banc, joined by PREGERSON, REINHARDT,
    THOMAS, and BERZON, Circuit Judges:
    Public confidence in the proper administration of the death
    penalty depends on the integrity of the process followed by
    the state. See Furman v. Georgia, 
    408 U.S. 238
    , 299 (1972)
    (Brennan, J., concurring) (“[I]t is our society that insists upon
    due process of law to the end that no person will be unjustly
    put to death, thus ensuring that many more of those sentences
    will not be carried out.”). So far as due process is concerned,
    twenty-four years of flawed proceedings are as good as no
    proceedings at all.
    We authorized Kevin Cooper to file his second or succes-
    sive habeas petition so that the district court could resolve the
    problem so ably articulated by Judge Silverman: “Cooper is
    either guilty as sin or he was framed by the police. There is
    no middle ground.” Cooper v. Woodford, 
    358 F.3d 1117
    ,
    1124 (9th Cir. 2004) (en banc) (Silverman, J., concurring in
    part and dissenting in part). Instead, through a series of errors
    accurately described by Judge Fletcher in his dissent, the dis-
    trict court precluded Cooper from having his last day in court.
    As Judge Fletcher states, the district court “imposed unrea-
    sonable conditions on the testing the en banc court directed;
    refused discovery that should have been available as a matter
    of course; limited testimony that should not have been lim-
    ited; and found facts unreasonably, based on a truncated and
    distorted record.” Fletcher Dissent at 5433.
    Because of the district court’s erroneous rulings and failure
    to follow the express direction of our en banc court, regretta-
    bly, we still do not know the answer to Judge Silverman’s
    5532                   COOPER v. BROWN
    query. See also Cooper v. Brown, 
    510 F.3d 870
    , 1004 (9th
    Cir. 2007) (McKeown, J., concurring) (“I . . . am troubled that
    we cannot, in Kevin Cooper’s words, resolve the question of
    his guilt ‘once and for all.’ ”). Granting rehearing en banc and
    remanding for a full and fair evidentiary hearing may not have
    resolved the question of guilt or innocence, but it most cer-
    tainly would have given us confidence that Cooper received
    his due opportunity to prove the innocence he has insisted
    upon since his arrest. I therefore respectfully dissent from the
    denial of rehearing en banc.
    FISHER, Circuit Judge, dissenting from denial of rehearing
    en banc, joined by KOZINSKI, Chief Judge, PREGERSON,
    GRABER and BERZON, Circuit Judges:
    I generally agree with Judge Fletcher that we should have
    taken this case en banc to require the factual inquiry the previ-
    ous en banc court expected to occur.
    REINHARDT, Circuit Judge, dissenting from the denial of
    rehearing en banc:
    I concur in Judge Fletcher’s thorough and highly persuasive
    dissent, as well as in Judge Wardlaw’s pithy summary of the
    judicial failures that infect this case. I would add, however,
    that the failures are not solely those of the district court. Our
    own handling of the matter, some of which has been made
    public and some of which has not, leaves much to be desired,
    and is a cause of considerable regret. There is no purpose,
    however, to looking backward at this point. What matters is
    that we have an obligation to afford Kevin Cooper a full and
    fair judicial hearing, and that once again we fail. By denying
    en banc review, we add to the prior systemic judicial malfunc-
    tions, and this time, we do so under a cloak of secrecy.
    COOPER v. BROWN                     5533
    I have on numerous occasions urged that we disclose the
    names of judges who vote for and against rehearing cases en
    banc. Spears v. Stewart, 
    283 F.3d 992
    , 997 (9th Cir. 2002)
    (Reinhardt, J., dissenting from denial of en banc review); In
    re Silicon Graphics Inc. Securities Litigations, 
    195 F.3d 521
    ,
    523-24 (9th Cir. 1999) (Reinhardt, J., dissenting from denial
    of en banc review); United States v. Koon, 
    45 F.3d 1303
    ,
    1308-10 (9th Cir. 1995) (Reinhardt, J., dissenting from denial
    of en banc review); Brewer v. Lewis, 
    997 F.2d 550
    , 556 (9th
    Cir. 1993) (Reinhardt, J., dissenting from denial of en banc
    review); Elder v. Holloway, 
    984 F.2d 991
    , 1001 (9th Cir.
    1993) (Reinhardt, J., dissenting from denial of en banc
    review); Harris v. Vasquez, 
    949 F.2d 1497
    , 1539-40 (9th Cir.
    1990) (Reinhardt, J., dissenting from denial of en banc
    review). Here, once again, the vote is extremely close, closer
    than the list of dissenters would suggest. I believe that as
    judges we have an obligation to let the public know how we
    vote on critical issues. The public, the legal academy, our col-
    leagues on other courts, and appointing authorities have a
    right to judge us based on our performance on the bench. In
    this case, in particular, I believe that public disclosure is
    important. Revealing how we voted would provide informa-
    tion that would be of interest to those who follow the course
    of our circuit law and who have drawn certain assumptions
    about the jurisprudence of various judges that sometimes are
    unwarranted. Most important, this is a case in which a man’s
    life is at stake. Kevin Cooper may or may not be guilty, but
    serious flaws in our legal system have been exposed. Whether
    to go en banc or not is a matter of judicial discretion. An en
    banc review by our court would surely do no harm. Nor
    would revealing the names of those who agree and disagree
    with affording this capital defendant a final protection before
    sending him on his way to execution by the state.
    RYMER, Circuit Judge, concurring:
    I concur in the court’s order declining to rehear this case en
    banc. The panel opinion explains why we affirmed denial of
    5534                     COOPER v. BROWN
    the writ, Cooper v. Brown, 
    510 F.3d 870
     (9th Cir. 2007)
    (Cooper IV), so I write now only to highlight the main reasons
    I agree with the court’s decision and not the dissents from
    failure to take the case en banc. I take particular issue with
    Judge Fletcher’s dissent because:
    1. The dissent improperly marshals the facts in the light
    most favorable to Kevin Cooper, yet the evidence was
    resolved against Cooper at trial — after he took the stand and
    testified — and at each step of post-conviction proceedings.
    The dissent also approaches the issues as if they were new,
    yet the same issues have been on the table since day one
    (except for DNA testing which didn’t exist at the time and
    which has turned out to be inculpatory). This includes Josh
    Ryen’s statements; handling the drop of blood A-41; the pres-
    ence of three strangers at the Canyon Corral Bar; that three
    white men were seen driving on the road leading away from
    the Ryen house; Cooper’s tennis shoes and whether they were
    Pro-Keds Dudes or left the impressions made at the Lease
    house as well as at and near the Ryen house; the bloody cov-
    eralls that were linked to Lee Furrow and were destroyed by
    law enforcement; where in the Lease house the hatchet sheath
    was found; whose blood was on a tan T-shirt found along the
    road; cigarette butts in the Ryen station wagon that were
    linked to Cooper; if there were a blue shirt and what happened
    to it; and police incompetence, mishandling of evidence, and
    chain of custody problems. All were known at the time of trial
    and have been litigated in one forum or another, unfavorably
    to Cooper, for the last 24 years.
    2. The dissent ignores the rules that Congress established
    for federal habeas review of state criminal judgments in the
    Antiterrorism and Effective Death Penalty Act of 1996
    (AEDPA). Cooper’s petition is a successive federal habeas
    corpus application that was preceded by appeals through the
    California courts, post-conviction petitions in state and federal
    court, and evidentiary hearings in both state and federal court.1
    (Text continued on page 5536)
    1
    Cooper’s February 19, 1985 judgment of conviction, and sentence to
    death on May 15, 1985, was affirmed by the California Supreme Court.
    COOPER v. BROWN                            5535
    People v. Cooper, 
    53 Cal.3d 771
    , 837, 
    281 Cal.Rptr. 90
    , 129 (1991). The
    United States Supreme Court denied a petition for certiorari, Cooper v.
    California, 
    502 U.S. 1016
     (1991). Cooper’s first federal application, sub-
    sequently amended and supplemented, was filed August 11, 1994, and
    denied August 25, 1997 following an evidentiary hearing. 92-CV-427,
    Doc. No. 165. We affirmed, Cooper v. Calderon, 
    255 F.3d 1104
     (9th Cir.
    2001) (Cooper I); denied Cooper’s PFR/PRFEB; and his petition for a writ
    of certiorari was denied, 
    537 U.S. 861
     (2002). Cooper filed a second fed-
    eral application on April 20, 1998 raising a claim related to the Koon
    statement, which we construed as an application for authorization to file
    a second or successive petition, and denied. Cooper v. Calderon, 
    274 F.3d 1270
     (9th Cir. 2001) (Cooper II). Cooper sought to file another successor
    application that involved DNA testing and tampering, which we denied,
    Cooper v. Calderon, No. 99-71430 (9th Cir. Feb. 14, 2003, April 7, 2003)
    (orders noting that DNA tests to which the state and Cooper agreed do not
    exculpate him, nor were there newly discovered facts establishing his
    innocence, and denying PFR based on asserted deficiencies in the testing
    and tampering). Meanwhile, Cooper filed seven petitions in the California
    Supreme Court together with a writ of mandate and various motions, a
    habeas petition in the San Diego County Superior Court, and six other
    petitions for a writ of certiorari in the United States Supreme Court as well
    as two petitions for habeas corpus, each of which was denied. On October
    22, 2002 Cooper filed a motion for mitochondrial DNA testing of hairs,
    and on June 16, 2003 for testing of the tan T-shirt, to show tampering.
    These were denied following an evidentiary hearing in the state court that
    found no tampering. Another petition in the California Supreme Court
    raised essentially the same claims as asserted in this application: actual
    innocence, tampering with evidence, failure to disclose exculpatory evi-
    dence, offering unreliable eye witness testimony of Josh Ryen, and deny-
    ing Cooper effective assistance of counsel during post-conviction DNA
    proceedings; the supreme court denied all claims on the merits on Febru-
    ary 5, 2004, and also denied as untimely those having to do with evidence
    tampering, failure to disclose exculpatory evidence, submission of false
    testimony to the jury, and offering Joshua Ryen’s unreliable testimony. On
    February 6, 2004, Cooper filed another application to file a successive
    application, which was initially denied, Cooper v. Woodford, 
    357 F.3d 1019
     (9th Cir. 2004), withdrawn, 
    357 F.3d 1054
     (Editor’s Note Feb. 8,
    2004), but was later granted when this court reheard the application en
    banc, Cooper v. Woodford, 
    358 F.3d 1054
    , 1117 (9th Cir. 2004) (Cooper
    III). As authorized by Cooper III, Cooper filed his third application in the
    district court April 2, 2004. The district court held an evidentiary hearing
    5536                       COOPER v. BROWN
    At this stage, AEDPA requires federal courts to presume that
    factual determinations by state courts are correct absent clear
    and convincing evidence to the contrary. 
    28 U.S.C. § 2254
    (e)(1).2 This applies to determinations adverse to Coo-
    per on police tampering with A-41 and cigarette butts, the
    presence of three men at the Canyon Corral, and tampering
    with the tan T-shirt. AEDPA also precludes federal courts
    from granting habeas relief when a state court has adjudicated
    the merits of a claim raised in a federal petition and the state
    court’s adjudication is neither contrary to, nor an unreason-
    able application of, federal law as determined by the United
    States Supreme Court. § 2254(d).3 This applies to state court
    determinations on Cooper’s claim of actual innocence, evi-
    dence tampering, Josh Ryen’s statements, destruction of the
    bloody coveralls, the blue shirt, and failure to disclose excul-
    patory evidence.
    at which 42 witnesses testified, and denied the writ. See Cooper IV,
    Appendix A. It is this ruling that is the subject of this appeal.
    2
    All references are to Title 28 of the U.S. Code Section 2254(e)(1) pro-
    vides:
    (e)(1) In a proceeding instituted by an application for a writ of
    habeas corpus by a person in custody pursuant to the judgment
    of a State court, a determination of a factual issue made by a
    State court shall be presumed to be correct. The applicant shall
    have the burden of rebutting the presumption of correctness by
    clear and convincing evidence.
    3
    Section 2254(d) provides:
    (d) An application for a writ of habeas corpus on behalf of a per-
    son in custody pursuant to the judgment of a State court shall not
    be granted with respect to any claim that was adjudicated on the
    merits in State court proceedings unless the adjudication of the
    claim—
    (1) resulted in a decision that was contrary to, or involved an
    unreasonable application of, clearly established Federal law,
    as determined by the Supreme Court of the United States; or
    (2) resulted in a decision that was based on an unreasonable
    determination of the facts in light of the evidence presented
    in the State court proceeding.
    COOPER v. BROWN                            5537
    AEDPA further provides that federal courts must dismiss
    claims that the petitioner has presented in a prior application.
    § 2244(b)(1).4 This includes handling of A-41; the Koon state-
    ment; destruction of the bloody coveralls; and planting evi-
    dence, or manipulating tests, on cigarette butts. AEDPA
    likewise requires dismissal of claims for which the factual
    predicate could have been discovered before and that, if
    proved and viewed in light of the evidence as a whole, would
    be sufficient to establish by clear and convincing evidence
    that, but for constitutional error, no reasonable factfinder
    would have found the petitioner guilty of the underlying
    offense. § 2244(b)(2)(B)(i)-(ii).5 This includes the blue shirt
    and daily log; cigarette butts and their testing; Josh Ryen’s
    statements; the three men at the Canyon Corral; Lee Furrow
    and the bloody coveralls; destruction of the coveralls; and the
    hatchet sheath.6
    4
    Section 2244(b)(1) provides:
    (b)(1) A claim presented in a second or successive habeas corpus
    application under section 2254 that was presented in a prior
    application shall be dismissed.
    5
    Section 2244(b)(2) provides:
    (b)(2) A claim presented in a second or successive habeas corpus
    application under section 2254 that was not presented in a prior
    application shall be dismissed unless—
    (A) the applicant shows that the claim relies on a new rule
    of constitutional law, made retroactive to cases on collateral
    review by the Supreme Court that was previously unavail-
    able; or
    (B)(i) the factual predicate for the claim could not have been
    discovered previously through the exercise of due diligence;
    and
    (ii) the facts underlying the claim, if proven and viewed in
    light of the evidence as a whole, would be sufficient to estab-
    lish by clear and convincing evidence that, but for constitu-
    tional error, no reasonable factfinder would have found the
    applicant guilty of the underlying offense.
    6
    Even if Schlup v. Delo, 
    513 U.S. 298
     (1998), rather than AEDPA,
    applies to actual innocence for purposes of gateway claims, the question
    5538                        COOPER v. BROWN
    As the dissent pays no attention, and gives no deference, to
    state court determinations, and reaches the merits of claims
    without regard to whether AEDPA mandates their dismissal,
    the picture it paints is quite different from the canvas that is
    actually before us.
    3. In lieu of AEDPA, the dissent starts with the false
    premise that the district court “flouted” this court’s “direc-
    tion” (in our en banc opinion authorizing Cooper to file his
    successive habeas application) to perform EDTA and mito-
    chondrial DNA testing.7 The district court did no such thing;
    would be whether, in light of all the evidence, including reliable new evi-
    dence, it is more likely than not that no reasonable juror would have found
    Cooper guilty beyond a reasonable doubt. Following an evidentiary hear-
    ing the district court found that, to the extent there was new evidence, it
    was incredible, unreliable or unpersuasive with respect to all claims in the
    petition, that neither EDTA nor mitochondrial DNA tests showed inno-
    cence or undermined evidence of Cooper’s guilt, and that the showing of
    actual innocence was insufficient to avoid procedural bars. The evidence
    supports these findings, therefore Cooper cannot pass through to other
    claims by means of either § 2244(b)(2)(B)(i)-(ii), or Schlup.
    7
    Cooper III, 
    358 F.3d at 1124
    . Having noted Cooper’s position that the
    question of his innocence could be answered “once and for all” by mito-
    chondrial testing of blond hairs in Jessica Ryen’s hands and testing for the
    presence of the preservative agent EDTA on the tan t-shirt, Cooper III also
    observed: “The district court may be in a position to resolve this case very
    quickly. As soon as Cooper’s application is filed, it should promptly order
    that these two tests be performed in order to evaluate Cooper’s claim of
    innocence.” 
    Id.
    Given the posture of the case as it was before the en banc panel, how-
    ever, I disagree with the dissent’s implicit assumption that this court could
    have “directed” the district court to conduct the two tests. The only issue
    before the en banc panel was whether Cooper should be allowed to file a
    successive application for habeas corpus under AEDPA. 
    28 U.S.C. § 2244
    (b)(3)(A); Cooper III, 
    358 F.3d at 1123
     (“We hold only that Cooper
    has made out a prima facie case that entitles him to file a second or suc-
    cessive application.”). Once this court authorized the application to be
    filed, it was up to the district court to decide in the first instance whether
    any of the claims — including those involving EDTA and mitochondrial
    DNA testing — could go forward, and how to deal with them. See 
    28 U.S.C. § 2244
    (b)(4). Its decision is, of course, subject to appeal and those
    issues are now properly before us. But our review is guided by AEDPA,
    not by a mistaken assumption about the scope of our prior mandate.
    COOPER v. BROWN                     5539
    it did precisely what Cooper III suggested that it do. Between
    April 2, 2004 (when Cooper filed his application pursuant to
    our authorization) and April 1, 2005, the district court held a
    tutorial on testing for the presence of EDTA,7 requested sub-
    missions on a variety of testing-related issues which the par-
    ties provided after consulting with their experts, participated
    in numerous hearings and conferences over a three-month
    period to develop a protocol, ordered the two tests to be con-
    ducted, resolved problems that came up during the process,
    analyzed the results, and issued a thoughtful 159-page ruling
    that discusses in meticulous detail all aspects of the testing as
    well as each claim in Cooper’s petition.
    4. The district court did the job it was required to do
    under Daubert v. Merrell Dow Pharmaceuticals, Inc., 
    509 U.S. 579
    , 589 (1993) — determine whether EDTA testing is
    scientifically valid. The district court concluded that the anal-
    ysis by Cooper’s expert did not pass muster. There is no evi-
    dence that the methodology Cooper’s expert espoused has
    been, or can be, tested. No other scientists regularly perform
    this kind of testing. There are no industry standards regarding
    proper testing protocol, and we have no idea how the history
    of each specimen’s exposure to environmental EDTA will
    affect the sample and, in turn, the testing. There is no “nor-
    mal” base level of EDTA. EDTA testing has only been
    offered to courts on two occasions, and in the one proceeding
    in which it was challenged, Cooper’s expert was involved and
    both his qualification, and his EDTA testing, were rejected as
    biased and flawed. Thus, the district court appropriately exer-
    cised its gate-keeping discretion to exclude EDTA testing as
    unreliable. The effect is to leave the state superior court’s fac-
    tual finding (following an evidentiary hearing) — that there
    was no tampering or contamination by law enforcement offi-
    cers — un-refuted.
    7
    As the dissent focuses only on EDTA testing, so do I.
    5540                   COOPER v. BROWN
    The Supreme Court has made it abundantly clear that this
    court may not set aside a district court’s factual findings
    unless those findings are clearly erroneous. Knowles v. Mirza-
    yance, No. 07-1315, __ S.Ct. __, __U.S. __, 
    2009 WL 746274
    (March 24, 2009), at *10; Fed. R. Civ. P. 52(a). The dissent,
    however, assesses the EDTA evidence from scratch (includ-
    ing evidence that was withdrawn), and substitutes its own
    findings for those of the district judge — based on charts and
    graphs and regression analyses that it has crafted. No court,
    let alone this court on appellate review, may consider evi-
    dence that an expert does not stand behind, nor may we reject
    a district court’s findings of fact — informed by the record
    before that court — by manipulating data in a way that
    appears better and brighter to us.
    In short, there is no reliable evidence of tampering.
    5. In asserting that “[t]he State of California may be about
    to execute an innocent man,” the dissent neglects to acknowl-
    edge the evidence tying Cooper to the murders, or the fact
    that, after all the testing that has been done post-conviction,
    no forensic evidence suggests that anyone else was at the
    scene of the crime or was the killer.
    Cooper testified at trial. He told an implausible story about
    what he did at the Lease house and how he left for Mexico.
    The jury disbelieved him. The defense theory (then as now)
    was that there were multiple assailants and police incompe-
    tence. Josh Ryen’s ambivalent recollections, the presence of
    three Caucasian men at the Canyon Corral Bar, that three His-
    panics were at the ranch before the murders, that three white
    men were seen driving down the road from the Ryen house,
    how the blood spot A-41 was handled (or mishandled), the
    bloody coveralls that were destroyed, the possible involve-
    ment of Koon or Lee Furrow, the possibility of an Aryan
    Brotherhood hit gone awry, the tan T-shirt, and a blue shirt
    that was logged in but not located, were all known at the time
    of trial. The jury rejected evidence pointing to other killers.
    COOPER v. BROWN                    5541
    No one else has ever been connected to the Lease house —
    where Cooper admitted being within an hour of the murders
    — and where, after the murders, were found a bloodstained
    rope similar to one in the Ryens’ driveway, signs of blood in
    the bathroom, and hairs consistent with the victims’ hair in
    the drain.
    Cooper admitted at trial that he had prison-issued tennis
    shoes and never said they weren’t Pro-Keds Dude, the shoe
    with a distinctive pattern that was consistent with impressions
    in both the Lease house and the Ryen house. The prosecution
    never suggested that Pro-Keds Dudes were only distributed to
    prisons, which makes both the warden’s recollections and
    inmate Taylor’s immaterial; the prosecution’s theory was that
    there was a link between the imprints found at and near the
    Ryen house and in the Lease house to Cooper, who admitted
    staying there, did not deny wearing prison-issue shoes, and
    whose shoe-size fit the prints.
    Cooper had a prison-issued jacket with buttons like the but-
    ton found on the rug in the bedroom of the Lease house where
    he stayed. Prison-issue tobacco for cigarettes, which Cooper
    smoked, was found in the Lease house and in the Ryen station
    wagon, where a hand-rolled butt (and a manufactured butt)
    consistent with his saliva were found. Hairs and hair frag-
    ments in the car were likewise consistent with Cooper’s.
    Cooper tossed incriminating evidence — including his
    prison-issued tennis shoes and other prison clothing — over-
    board into the ocean from the boat on which he got a job in
    Ensenada after leaving the Lease house. The jury could, and
    we must presume that it did, infer consciousness of guilt from
    this conduct. No “other killer” toward whom the dissent says
    that “evidence points” is African-American. Cooper’s expert
    agreed that the person who deposited A-41 on the wall of the
    hallway in the Ryen house was African-American. Post-
    conviction DNA testing, requested by Cooper, inculpates him.
    5542                   COOPER v. BROWN
    No test on any item has ever pointed to any one else as the
    killer. This includes, in addition to A-41, tests of blood on the
    hatchet taken from the Lease house that matched Doug and
    Jessica Ryen and Chris Hughes; a bloodstained rope found in
    the Lease bedroom closet that was similar to a bloodstained
    rope found on the Ryens’ driveway; cigarette butts found in
    the Ryen station wagon and headboard of the bed on which
    Cooper slept one night in the Lease house; hairs recovered
    from the Ryen’s station wagon; fingerprints in the Lease
    house; hairs in the sink and shower at the Lease house that
    were consistent with Jessica’s and Doug Ryen’s; and the tan
    T-shirt.
    The state courts found no evidence of tampering with
    respect to the shoe prints, cigarette butts, A-41, or DNA test-
    ing. Neither EDTA testing, mitochondrial DNA testing, nor
    anything else presented at the evidentiary hearing in district
    court, clearly and convincingly shows otherwise.
    For sure, no investigation or trial is perfect. Judge
    McKeown makes this point in her panel concurrence, Cooper
    IV, 
    510 F.3d at 1007
    , yet recognizes that the law demands
    affirmance. I agree with the full court’s decision that this is
    so.
    

Document Info

Docket Number: 05-99004

Filed Date: 5/11/2009

Precedential Status: Precedential

Modified Date: 10/14/2015

Authorities (22)

95-cal-daily-op-serv-7114-95-daily-journal-dar-12163-scott-d , 64 F.3d 1373 ( 1995 )

william-daubert-joyce-daubert-individually-and-as-guardians-ad-litem-for , 43 F.3d 1311 ( 1995 )

elsie-brewer-individually-and-as-next-friend-of-john-george-brewer-v , 997 F.2d 550 ( 1993 )

Mooney v. Holohan , 55 S. Ct. 340 ( 1935 )

Furman v. Georgia , 92 S. Ct. 2726 ( 1972 )

Daubert v. Merrell Dow Pharmaceuticals, Inc. , 113 S. Ct. 2786 ( 1993 )

William Lyle WORATZECK, Petitioner, v. Terry L. STEWART, ... , 118 F.3d 648 ( 1997 )

United States v. Gloria Ann Morales , 108 F.3d 1031 ( 1997 )

Kevin Cooper v. Jeanne Woodford, Warden, San Quentin State ... , 357 F.3d 1054 ( 2004 )

anthony-marshall-spears-v-terry-stewart-director-of-the-arizona , 283 F.3d 992 ( 2002 )

in-re-silicon-graphics-inc-securities-litigation-edmund-j-janas-v , 195 F.3d 521 ( 1999 )

Napue v. Illinois , 79 S. Ct. 1173 ( 1959 )

Brady v. Maryland , 83 S. Ct. 1194 ( 1963 )

Schlup v. Delo , 115 S. Ct. 851 ( 1995 )

Cooper v. Brown , 510 F.3d 870 ( 2007 )

charles-k-elder-beverly-s-elder-husband-and-wife-v-rd-holloway-other , 984 F.2d 991 ( 1993 )

Kevin Cooper v. Arthur Calderon, Warden of California State ... , 274 F.3d 1270 ( 2001 )

Kevin Cooper v. Jeanne Woodford, Warden, San Quentin State ... , 357 F.3d 1019 ( 2004 )

Clarence Ray Allen v. Jeanne S. Woodford, Warden, of the ... , 395 F.3d 979 ( 2005 )

Beck v. Alabama , 100 S. Ct. 2382 ( 1980 )

View All Authorities »