Anthony Olvera v. David Gomez ( 2021 )


Menu:
  •                                In the
    United States Court of Appeals
    For the Seventh Circuit
    ____________________
    No. 18-3435
    ANTHONY C. OLVERA,
    Petitioner-Appellant,
    v.
    DAVID GOMEZ, Warden,
    Respondent-Appellee.
    ____________________
    Appeal from the United States District Court for the
    Central District of Illinois.
    No. 4:15-cv-04042-SLD — Sara Darrow, Chief Judge.
    ____________________
    ARGUED JANUARY 14, 2021 — DECIDED JUNE 22, 2021
    ____________________
    Before RIPPLE, KANNE, and ROVNER, Circuit Judges.
    RIPPLE, Circuit Judge. Petitioner Anthony Olvera was the
    driver in a gang-related, drive-by shooting that resulted in the
    death of an innocent bystander. An Illinois jury found
    Mr. Olvera guilty of first-degree murder on the theory that he
    was accountable for the shooter’s conduct. Mr. Olvera now
    seeks postconviction review, claiming that his trial counsel
    provided ineffective assistance by conducting an inadequate
    pretrial investigation. The state courts denied Mr. Olvera’s
    2                                                  No. 18-3435
    petition. He then filed this petition for habeas corpus under
    28 U.S.C. § 2254. The district court denied relief. We now af-
    firm the district court’s judgment.
    I
    BACKGROUND
    Mr. Olvera’s conviction stems from the January 8, 2000,
    shooting death of Stephen Stropes, the victim of a gang-re-
    lated drive-by shooting in East Moline, Illinois. It is undis-
    puted that Mr. Olvera’s codefendant, Kristian Delgado, fired
    the shot that killed Stropes. It is also undisputed that
    Mr. Olvera was the driver of the vehicle and did not fire any
    shots that evening. Delgado pleaded guilty to murder, and
    the State prosecuted Mr. Olvera on the theory that he was ac-
    countable for Delgado’s actions that killed Stropes.
    The tragic events of January 8 started with a rather trivial
    argument. Mr. Olvera’s girlfriend, Guadalupe Raya, was at a
    party where she got into an argument with members of the
    Latin Kings gang. She accused them of using her camera to
    take photos of themselves flashing gang signs. The Latin
    Kings are a rival of the Bishops gang of which Mr. Olvera and
    Delgado were members. Upset over the camera incident,
    Raya called Mr. Olvera and asked him to pick her up from the
    party. Mr. Olvera drove over with Delgado in the passenger
    seat. As they approached the house where Raya was waiting,
    they passed the group of Latin Kings, who also had left the
    party and were standing on a street corner. Delgado fired one
    shot out of the car window on the first pass; then he and
    Mr. Olvera picked up Raya, turned around, sped off in the di-
    rection of the Latin Kings, and fired several more shots at the
    Latin Kings on the second pass. It was during the second pass
    No. 18-3435                                                                3
    that a bullet struck Stropes in the head; he was later pro-
    nounced dead.
    A.
    The State’s theory at trial was that Mr. Olvera was ac-
    countable for Delgado’s firing at the Latin Kings on the sec-
    ond pass and killing Stropes. 1 The State called Raya as a wit-
    ness. She testified that she was at a party, got into an argu-
    ment with Latin King gang members, and called Mr. Olvera
    at his friend Daniel Mendoza’s house to come and pick her
    up. As she waited outside for Mr. Olvera, she saw the Latin
    King members and then heard a loud noise. Immediately after
    hearing the loud noise, Mr. Olvera and Delgado pulled up in
    Mr. Olvera’s four-door Buick sedan. Raya testified that
    Mr. Olvera told her to get in the car and duck. As she hid on
    the floor of the Buick’s backseat, she heard several loud noises
    and smelled smoke. Raya testified that shortly after the shoot-
    ing, Mr. Olvera sold the Buick sedan to an out-of-town ac-
    quaintance. She also testified that she never saw a gun and
    that she never heard Mr. Olvera tell Delgado to shoot at the
    Latin Kings.
    Raya’s friend, Alma Mendoza, who had hosted the party
    on the evening of the shooting, also testified. Her testimony
    corroborated Raya’s story about her getting into an argument
    with the Latin Kings. Alma Mendoza also identified the Latin
    1 Under Illinois law, an individual can be accountable for the criminal acts
    of another if “either before or during the commission of an offense, and
    with the intent to promote or facilitate that commission, he or she solicits,
    aids, abets, agrees, or attempts to aid that other person in the planning or
    commission of the offense.” 720 ILCS 5/5-2(c).
    4                                                           No. 18-3435
    King members at the party as Michael Olvera,2 Johnny
    Rigsby, Gabino Gutierrez, Leo Reyes, and Jose Perez. Alma
    Mendoza told the jury that she did not want gang-related
    problems at her party, so she asked Raya and the Latin Kings
    to leave. She also testified that she heard something that
    sounded like “three fireworks” after Raya and the group of
    Latin Kings left the party.3
    Tara Ramos, another partygoer, testified that she saw
    Raya get into a car after leaving the party. The driver of that
    car, Ramos testified, bent over and appeared to grab some-
    thing from under his seat. She said that the driver yelled for
    Raya to get quickly into the car, which then sped down the
    street. She then heard gunshots.
    Four of the Latin King members at whom Delgado fired—
    Reyes, Rigsby, Michael Olvera, and Gutierrez—also testified.
    The fifth member, Perez, was out of the state at the time of
    Mr. Olvera’s trial.
    Gutierrez testified that he and the four other Latin Kings
    left the party and were standing on the corner of 15th Avenue
    and 12th Street. A car pulled around the corner and stopped
    next to them. The passenger then stuck a gun out the window.
    As the Latin King members fled, they heard a gunshot.
    Gutierrez, Perez, and Reyes ended up on 18th Avenue. There,
    Gutierrez saw the car speeding closer again. On this second
    pass, he testified that he saw the front-seat passenger lean out
    the window before firing two or three more shots. Reyes
    2 Michael Olvera is Mr. Olvera’s cousin.  We will include Michael Olvera’s
    first name in all references to avoid confusion with the petitioner, who we
    will continue to refer to as Mr. Olvera.
    3 R.8-16   at 53.
    No. 18-3435                                                    5
    testified along the same lines as Gutierrez. Reyes added that
    he heard someone in the car yell “King blink,” which is a dis-
    respectful term to members of the Latin Kings. 4
    Rigsby testified that although Perez, Reyes, and Gutierrez
    fled after seeing the gun on the car’s first pass, he and Michael
    Olvera remained. He also testified to hearing a single gunshot
    followed by three more a short time later. Michael Olvera tes-
    tified along the same lines as Rigsby.
    David Routien, a local pastor and disinterested witness,
    testified that he heard three gunshots as he was preparing his
    evening sermon. When he looked out his window, he saw
    three young Hispanic men running down 18th Avenue.
    Lucille Starkey, another disinterested witness, testified
    that as she was driving home from work, a four-door sedan
    pulled in front of her car. Both the sedan and her car turned
    onto 18th Avenue. She testified that she saw Stropes standing
    by a tree, saw someone else run by him, then heard gunshots
    and saw Stropes collapse. The sedan in front of her sped away
    right after the gunshots. Starkey later identified Mr. Olvera’s
    Buick as the car she saw.
    Two witnesses testified that Mr. Olvera told Delgado to
    shoot at the Latin Kings. The first was Jolene Montalvo, who
    was dating Mr. Olvera’s friend, Robert Espinoza. Montalvo
    initially testified in the grand jury proceedings. During her
    trial testimony, Montalvo claimed not to remember events to
    which she had testified during the grand jury proceeding. The
    prosecutor therefore used Montalvo’s grand jury transcripts
    to refresh her recollection throughout her trial testimony.
    4 R.8-17   at 84.
    6                                                  No. 18-3435
    Ultimately, she testified that four days after the shooting, she
    overheard Mr. Olvera tell Espinoza that the Latin Kings
    rushed the car and that he told Delgado to shoot in the air to
    scare them away. She also testified that Mr. Olvera said the
    second shooting happened when he thought the Latin Kings
    were going to shoot at the car.
    The other witness was Darrin Rhodes, an inmate housed
    on the same cellblock as Mr. Olvera during Mr. Olvera’s pre-
    trial custody. Rhodes testified that he spoke with Mr. Olvera
    about his case several times and Mr. Olvera said he instructed
    Delgado to shoot at the Latin Kings. Mr. Olvera also told
    Rhodes that after the shooting, he and Delgado scratched the
    inside of the gun barrel to prevent forensic tracing.
    Mr. Olvera’s friend, Daniel Mendoza, to whose house Del-
    gado and Mr. Olvera went following the shooting, also testi-
    fied at the trial. Police arrested Delgado at Mendoza’s house
    after the shooting and recovered evidence from the residence.
    While arresting Delgado, the police recovered a .45-caliber
    handgun that forensics later matched to shell casings found
    near both shooting scenes. Mendoza testified that he had been
    convicted for filing down the serial number on the gun and
    that he had been subpoenaed to appear at the trial.
    The jury also heard evidence that shortly after the shoot-
    ing, Mr. Olvera sold his Buick. Police spotted the Buick ten
    days after the shooting. When they stopped the car, they iden-
    tified the driver as John Teague. Upon searching the car, the
    police found Mr. Olvera’s license plates and registration, as
    No. 18-3435                                                            7
    well as paperwork purporting to transfer the title of the car
    from Mr. Olvera to “Boy Tiegue” a day earlier. 5
    Mr. Olvera called a hairdresser, Marisol Sandoval, as a de-
    fense witness. Sandoval testified that, according to her ap-
    pointment logs, Mr. Olvera was in her salon at the time when
    Montalvo said she overheard Mr. Olvera speaking to Espi-
    noza at Espinoza’s apartment. Sandoval’s testimony,
    Mr. Olvera contended, undercut Montalvo’s credibility.
    After the parties rested, the jury deliberated for approxi-
    mately one half-hour before returning a guilty verdict. The Il-
    linois courts later upheld his murder conviction on direct ap-
    peal. 6
    B.
    In May 2002, Mr. Olvera sought postconviction review of
    his conviction in the Illinois state courts. His primary allega-
    tion was that his trial counsel had failed to “contact or call”
    several witnesses “whose testimony would have been of sig-
    nificant benefit to him.” 7 In support of his petition, he at-
    tached affidavits from multiple witnesses and potential wit-
    nesses, seven of whom are relevant to his appeal here. 8
    5   Id. at 2.
    6   R.8-1 (unpublished order resolving direct appeal).
    7   R.8-11 at 85.
    8 There were two additional affidavits, one from partygoer Tara Ramos
    and another from Mr. Olvera’s friend Rose Garza. The state appellate
    court concluded that neither impacted the ineffective assistance inquiry,
    and Mr. Olvera does not challenge the decision with respect to those two
    affidavits.
    8                                                    No. 18-3435
    First, Robert Espinoza averred that his girlfriend, Mon-
    talvo, testified falsely to having overheard Mr. Olvera admit
    to Espinoza that he had told Delgado to fire at the Latin Kings.
    According to Espinoza, no such conversation occurred. Espi-
    noza posited that Montalvo had lied to gain leniency in a fed-
    eral racketeering case he faced at the time of Mr. Olvera’s trial.
    Espinoza also noted that Mr. Olvera’s attorney knew of his
    whereabouts during the trial but never contacted him.
    Second, Damian Olvera, another cousin of Mr. Olvera,
    averred that he heard Espinoza tell Montalvo on multiple oc-
    casions to testify against Mr. Olvera. Damian Olvera did not
    come forward with this information until May 2002, after
    speaking with Mr. Olvera.
    Third, Daniel Mendoza averred that he was with
    Mr. Olvera and Delgado on the night of the shooting. He
    claimed that Mr. Olvera received a call from Raya to pick her
    up from the party and that he did not hear Mr. Olvera ask
    Delgado to accompany him.
    Fourth, John Teague averred that he purchased
    Mr. Olvera’s Buick following the shooting. He also claimed
    that he was already in discussions with Mr. Olvera about buy-
    ing the car prior to the shooting. He added that Mr. Olvera’s
    counsel never contacted him about testifying. Teague has
    since died.
    Fifth, Jose Perez, the Latin Kings member who was una-
    vailable during Mr. Olvera’s trial, averred that after
    Mr. Olvera’s car turned onto 18th Avenue, he saw Reyes run
    at the car, then heard shots fired.
    Sixth, Michael Olvera averred that he overheard a conver-
    sation between Perez, Reyes, and Gutierrez discussing
    No. 18-3435                                                   9
    Reyes’s running at the Buick before the shots were fired. Mi-
    chael Olvera did not claim in his affidavit that Mr. Olvera’s
    counsel never contacted him. He did, however, state that he
    would testify to hearing the conversation if called at a future
    trial.
    Seventh, Kristian Delgado averred that he decided to join
    Mr. Olvera on the trip to pick up Raya on the night of January
    8, 2000, on his own volition, not because Mr. Olvera extended
    an invitation. Nor, according to Delgado’s affidavit, did
    Mr. Olvera know that Delgado was carrying a firearm. Del-
    gado also averred that he fired the shot in the air on the first
    pass of the Latin King members without Mr. Olvera’s direc-
    tion. He claimed that Mr. Olvera became upset at his decision
    to fire during the first pass. On the second pass, he claimed
    that one of the Latin Kings ran toward the car and he fired at
    the charging person out of fear that the Latin King was going
    to shoot at the Buick.
    Delgado added to this account in a second affidavit. There,
    he claimed that Mr. Olvera told him to get out of the car after
    the first shooting but that he pointed his gun at Mr. Olvera
    and said he would harm Mr. Olvera if he told anyone about
    the shooting. He also added that he thought he saw a gun in
    the hand of the charging Latin King. He added that his state-
    ments from his own sentencing hearing, in which he said
    Mr. Olvera told him to fire, were false. He now claims to have
    made those statements in the hope of receiving a lesser sen-
    tence. Finally, Delgado claimed that Mr. Olvera’s counsel did
    not contact him about testifying on Mr. Olvera’s behalf.
    The Illinois Circuit Court denied Mr. Olvera’s postconvic-
    tion petition. Mr. Olvera then appealed, claiming that the af-
    fidavits showed that trial defense counsel was ineffective for
    10                                                   No. 18-3435
    failing to investigate the self-defense argument raised in Del-
    gado’s affidavit. The Illinois Appellate Court reviewed his pe-
    tition de novo and affirmed the trial court’s decision. It con-
    cluded that none of the affidavits that Mr. Olvera had submit-
    ted were sufficient on their own, or in combination, to estab-
    lish ineffective assistance of counsel. For six of the seven affi-
    davits, the appellate court held that they failed to show defi-
    cient performance by Mr. Olvera’s counsel: (1) Delgado’s af-
    fidavit “relate[d] to appropriate trial strategy”; (2) Daniel
    Mendoza’s affidavit was both “speculative” and only stated
    that Mr. Olvera did not tell Delgado to join him on the trip to
    pick up Raya; (3) Teague’s affidavit was about events that
    happened after the shooting, thus irrelevant to whether
    Mr. Olvera was culpable under Illinois law for Delgado’s
    shooting Stropes; (4) Damian Olvera’s affidavit was about in-
    formation that Mr. Olvera’s counsel would not have reasona-
    bly uncovered ahead of the trial; (5) Michael Olvera’s affidavit
    was inadmissible hearsay; and (6) Perez was out of the state
    at the time of Mr. Olvera’s trial, so he could not have been
    called to testify.9 For the seventh affidavit, that from Espi-
    noza, the appellate court went straight to the prejudice prong
    and held that, even with Espinoza’s proposed testimony,
    there was still substantial evidence to convict Mr. Olvera.
    C.
    After exhausting his state review, Mr. Olvera filed this pe-
    tition under 28 U.S.C. § 2254, submitting that the state appel-
    late court’s decision was both contrary to and an unreasona-
    ble application of the Supreme Court’s holding in Strickland v.
    9   R.8-5 at ¶ 17.
    No. 18-3435                                                                11
    Washington, 
    466 U.S. 668
     (1984). 10 On the “contrary to” argu-
    ment, Mr. Olvera alleged that the state court required him to
    prove that he would have been acquitted in order to demon-
    strate prejudice. The district court rejected this argument, not-
    ing that even though the state appellate court misarticulated
    the Strickland prejudice definition at one point in its opinion,
    it had earlier articulated the correct definition.
    On the “unreasonable application” argument, the district
    court found no basis to upset the state appellate court’s deci-
    sion. The district court held that the state appellate court did
    not unreasonably apply Strickland when it disposed of Espi-
    noza’s affidavit on the prejudice prong. Even if the jury had
    heard Espinoza’s proposed testimony, they still would have
    heard Rhodes’s testimony about Mr. Olvera’s jailhouse con-
    fession and Raya’s testimony about Mr. Olvera telling her to
    duck in the Buick and then hearing the shots. The district
    court also held that the state appellate court did not act un-
    reasonably in resolving the other six affidavits on Strickland’s
    deficient performance prong.
    The district court then granted Mr. Olvera a certificate of
    appealability because it believed that reasonable jurists could
    disagree with its decision. See 28 U.S.C. § 2253(c).
    10The parties agree that the district court correctly identified the state ap-
    pellate court’s opinion as the last reasoned state court opinion, and thus
    the opinion we should look to on review. See Carrion v. Butler, 
    835 F.3d 764
    , 772 (7th Cir. 2016).
    12                                                             No. 18-3435
    II
    DISCUSSION
    We review a district court’s denial of habeas relief de
    novo. 11 Felton v. Bartow, 
    926 F.3d 451
    , 464 (7th Cir. 2019).
    When a person in custody pursuant to a judgment of a state
    court asks the district court to grant habeas relief from a crim-
    inal judgment imposed by a state court, the Antiterrorism and
    Effective Death Penalty Act (“AEDPA”) governs, and greatly
    curtails, its review. See Hicks v. Hepp, 
    871 F.3d 513
    , 524 (7th
    Cir. 2017). A federal court may set aside a state court’s legal
    determinations only when the state court’s adjudication of the
    petitioner’s claim “resulted in a decision that was contrary to,
    or involved an unreasonable application of, clearly estab-
    lished Federal law, as determined by the Supreme Court of
    the United States.” 28 U.S.C. § 2254(d)(1).
    This standard is purposefully difficult to satisfy. See Das-
    sey v. Dittmann, 
    877 F.3d 297
    , 302 (7th Cir. 2017) (en banc) (cit-
    ing Harrington v. Richter, 
    562 U.S. 86
    , 102 (2011)). Legal error
    by the state court only warrants relief when the state court’s
    decision is “unreasonably wrong under an objective stand-
    ard.” 
    Id.
     (citing Williams v. Taylor, 
    529 U.S. 362
    , 410–11 (2000)).
    A state court acts contrary to Supreme Court precedent when
    it “arrives at a conclusion opposite to that reached by [the Su-
    preme] Court on a question of law or if the state court decides
    a case differently than [the Supreme] Court has on a set of
    materially indistinguishable facts.” Williams, 
    529 U.S. at 413
    .
    A state court’s legal decision may also be objectively unrea-
    sonable if it “identifies the correct governing legal principle
    11 The district court exercised its jurisdiction under 28 U.S.C. §§   1331, 2241,
    and 2254. We exercise ours under 28 U.S.C. §§ 1291 and 2253.
    No. 18-3435                                                   13
    from [the Supreme] Court’s decisions but unreasonably ap-
    plies that principle to the facts of the prisoner’s case.” Id.
    Mr. Olvera’s petition asserts that the state appellate court’s
    decision was both contrary to and an unreasonable application
    of the Supreme Court’s clearly established law in Strickland v.
    Washington, 
    466 U.S. at 668,
     which sets out the framework for
    evaluating ineffective assistance of counsel claims.
    When a habeas petitioner challenges his conviction based
    on ineffective assistance of counsel, the Supreme Court’s case
    law imposes a “‘doubly deferential’ standard of review that
    gives both the state court and the defense attorney the benefit
    of the doubt.” Burt v. Titlow, 
    571 U.S. 12
    , 15 (2013) (quoting
    Cullen v. Pinholster, 
    563 U.S. 170
    , 190 (2011)). AEDPA provides
    the first layer of deference to the state court, and substantive
    Sixth Amendment law provides the second layer of deference
    to defense counsel. See 
    id.
    It is well established that the Sixth Amendment requires
    effective assistance of counsel. To determine whether a peti-
    tioner has been deprived of this guarantee, we employ the fa-
    miliar two-pronged test from Strickland, 
    466 U.S. at 688
    . First,
    we determine whether an attorney’s “representation fell be-
    low an objective standard of reasonableness.” 
    Id.
     When coun-
    sel makes a “thorough [pretrial] investigation of [the] law and
    facts,” counsel’s trial strategy is “virtually unchallengeable.”
    
    Id. at 690
    . In contrast, when counsel’s pretrial investigation is
    less than complete, counsel’s strategic choices are “reasonable
    precisely to the extent that reasonable professional judgments
    support the limitations on investigation.” 
    Id. at 691
    . In short,
    “counsel has a duty to make reasonable investigations or to
    make a reasonable decision that makes particular investiga-
    tions unnecessary.” 
    Id. 14
                                                                No. 18-3435
    The Supreme Court has elaborated on the Strickland stand-
    ard in ways that are pertinent to the task at hand. In Wiggins
    v. Smith, 
    539 U.S. 510
    , 527–28 (2003), the Court emphasized
    that in applying Strickland’s first prong, we may not assume
    that counsel’s decision to limit an investigation was reasona-
    ble without actually inquiring into why counsel stopped in-
    vestigating. 12 For example, if counsel never bothers to find
    out what a potential witness may say on the stand, counsel’s
    decision not to call that witness to testify cannot be passed off
    as a matter of strategy. See Mosley v. Atchison, 
    689 F.3d 838
    , 848
    (7th Cir. 2012) (“According to [two potential witnesses’] affi-
    davits, which were treated as true for purposes of the state
    courts’ summary disposition, [the petitioner’s] lawyer never
    even interviewed them to learn what they might say. On that
    limited record before the state courts, the courts had to as-
    sume the lawyer was not aware of the specifics of their poten-
    tial testimony.”). In short, assigning strategic value to coun-
    sel’s decision requires addressing “the adequacy of the pre-
    trial investigation, which was clearly established under Strick-
    land as the critical threshold question.” Campbell v. Reardon,
    
    780 F.3d 752
    , 766 (7th Cir. 2015).
    Strickland’s second prong asks whether the defendant suf-
    fered any prejudice as a result of counsel’s deficient perfor-
    mance. See Strickland, 
    466 U.S. at 693
    . As the Supreme Court
    put it in Strickland, “[w]hen a defendant challenges a convic-
    tion, the question is whether there is a reasonable probability
    12 See also Campbell v. Reardon, 
    780 F.3d 752
    , 764 (7th Cir. 2015) (“If coun-
    sel’s decision not to investigate [two potential witnesses] was itself unrea-
    sonable, then his decision not to present their testimony—and to rely on
    [an alternative] theory instead—was too ill-informed to be considered rea-
    sonable.”).
    No. 18-3435                                                              15
    that, absent the errors, the factfinder would have had a rea-
    sonable doubt respecting guilt.” 
    Id. at 695
    . A reasonable prob-
    ability, the Court said, “is a probability sufficient to under-
    mine confidence in the outcome.” 
    Id. at 694
    .
    With this doubly deferential framework in mind, we first
    will address Mr. Olvera’s claim that the state court acted con-
    trary to Strickland, then turn to his claim that the state court
    unreasonably applied Strickland.
    A.
    Mr. Olvera’s first argument relies on § 2254(d)(1)’s “con-
    trary to” clause. He focuses on the state appellate court’s
    statement that even if Espinoza’s affidavit were true, the other
    evidence against Mr. Olvera “was so substantial that we can-
    not find this new evidence would have resulted in an acquittal.
    Therefore, we find that defendant was not prejudiced and
    counsel was not ineffective.” 13 Mr. Olvera points out correctly
    that Strickland, 
    466 U.S. at 694,
     only requires “a reasonable
    probability that, but for counsel’s unprofessional errors, the
    result of the proceeding would have been different.” He sub-
    mits that, by misstating the burden for establishing prejudice,
    the state appellate court acted contrary to Strickland. 14
    13   R.8-5 at ¶ 18 (emphasis added).
    14 Mr. Olvera also likens his case to Andrus v. Texas, 
    140 S. Ct. 1875
    , 1886
    (2020), in which the Supreme Court found lacking the Strickland inquiry
    conducted by the Texas Court of Criminal Appeals. But there, the Texas
    court denied the petitioner’s Strickland claim in a single sentence, leaving
    the Supreme Court to wonder whether it “considered Strickland prejudice
    at all.” 
    Id.
     Here, there is no doubt the Illinois Appellate Court considered
    prejudice; Mr. Olvera simply contends that the court applied the wrong
    definition of prejudice.
    16                                                   No. 18-3435
    The State acknowledges that the appellate court was
    wrong to reference acquittal in its analysis but nevertheless
    contends that the court’s error was simply an inartful articu-
    lation of Strickland’s prejudice inquiry. It invites our attention
    to two other aspects of the state appellate court’s opinion as
    evidence that the state court applied the correct Strickland
    prejudice standard. First, the State notes that on the same
    page that the appellate court misstated the standard, it also
    quoted the correct Strickland prejudice standard. Second, the
    State notes that immediately following the erroneous articu-
    lation of prejudice, the state appellate court cited People v.
    Thompson, 
    835 N.E.2d 933
    , 937 (Ill. App. Ct. 2005), which con-
    tains the correct articulation of Strickland’s prejudice prong.
    Reading the state appellate court’s opinion in its totality, the
    State submits that the appellate court applied the correct
    standard, even if in one part of its opinion it mistakenly artic-
    ulated that standard.
    The State is correct. The Supreme Court has said, explic-
    itly, that “use of the unadorned word ‘probably’ is permissi-
    ble shorthand when the complete Strickland standard is else-
    where recited.” Holland v. Jackson, 
    542 U.S. 649
    , 655 (2004) (per
    curiam). When examining a state court’s articulation of the
    Strickland standard, we must keep in mind “the presumption
    that state courts know and follow the law” and give their ar-
    ticulation of that standard “the benefit of the doubt.” Woodford
    v. Visciotti, 
    537 U.S. 19
    , 24 (2002) (per curiam). A pair of our
    decisions illustrate how this deference operates in circum-
    stances like those presented here. In Stanley v. Bartley, 
    465 F.3d 810
    , 813 (7th Cir. 2006), we encountered a state court opinion
    that correctly stated the Strickland prejudice standard in one
    section, but later stated the standard as requiring a showing
    that absent counsel’s error the outcome of the trial would have
    No. 18-3435                                                                   17
    been different—thus appearing to raise the standard beyond
    the “reasonable probability” that Strickland imposes. We gave
    the state court the “benefit of the doubt,” holding that it had
    not acted contrary to Strickland. 
    Id.
     Instead, we observed that
    because the state court “expounded the well-known standard
    correctly on the previous page of its opinion, it is more likely
    that the court stated its conclusion imprecisely than that it ap-
    plied a different standard.” 
    Id.
     (citing Visciotti, 
    537 U.S. at 19
    ).
    More recently, in Fayemi v. Ruskin, 
    966 F.3d 591
    , 594 (7th
    Cir. 2020), we reviewed a state court decision that correctly
    stated Strickland’s prejudice definition but later twice mistak-
    enly asked whether the result of the case “would likely have
    been different.” We rejected the petitioner’s argument that the
    state court acted contrary to Strickland, instead reiterating that
    “we do not attribute to the state’s judiciary an unexplained
    replacement of the correct standard with an incorrect one.” 
    Id.
    The more respectful approach, we said, was to treat the later
    mistaken statements as shorthand for the correct definition
    that the state court noted earlier in its opinion. 15 See 
    id. 15
     To be sure, these are not the only two cases that illustrate the “benefit
    of the doubt” principle. Similar cases from our court include: Sussman v.
    Jenkins, 
    636 F.3d 329
    , 359–60 (7th Cir. 2011) (state court opinion omitting
    “reasonable probability” language in conclusion did not demonstrate that
    the state court employed incorrect standard); Woods v. Schwartz, 
    589 F.3d 368
    , 378 n.3 (7th Cir. 2009) (“We have noted numerous times that there is
    no error when a court has correctly noted the Strickland standard and then
    used an incorrect shorthand version when stating its conclusion.”). Our
    case law is also consistent with that from our sister circuits. See, e.g., Charles
    v. Stephens, 
    736 F.3d 380
    , 393 (5th Cir. 2013) (explaining that, viewed as a
    whole, the state court’s opinion “indicates that the state habeas court omit-
    ted the ‘reasonable probability’ modifier not due to its incorrect under-
    standing of the prejudice standard, but as a shorthand method to refer to
    18                                                             No. 18-3435
    Mr. Olvera attempts to distinguish his case from the others
    we have decided by noting the exact wording of the mistaken
    prejudice definition here. He correctly notes that our past
    cases involved different incorrect formulations of Strickland’s
    prejudice prong. Yet those slight differences do not make for
    a genuine distinction. We still give the state court the benefit
    of the doubt. See Visciotti, 
    537 U.S. at 24
    . And our past deci-
    sions encountered mistaken definitions of roughly equal sig-
    nificance. See, e.g., Sussman v. Jenkins, 
    636 F.3d 329
    , 359–60 (7th
    Cir. 2011) (state court opinion omitted “reasonable probabil-
    ity” when defining prejudice but was not contrary to Strick-
    land).
    Mr. Olvera admits that the state appellate court included
    the correct Strickland framework on the same page that it
    stated the mistaken definition of prejudice. As we said in
    Fayemi, the approach that most respects the state appellate
    court is to treat the second, mistaken statement as a shorthand
    for the earlier correct definition. Treating the mistaken defini-
    tion as shorthand makes even more sense in Mr. Olvera’s case
    because the state court followed the statement with a citation
    to a decision that included the correct definition of preju-
    dice. 16 Absent circumstances that would raise a grave concern
    that the state court actually applied a contrary standard, we
    see no basis to override the benefit of the doubt that § 2254
    provides to the state court’s decision.
    the correct standard”); Williams v. Roper, 
    695 F.3d 825
    , 832 (8th Cir. 2012)
    (similar); Bledsoe v. Bruce, 
    569 F.3d 1223
    , 1231–33 (10th Cir. 2009) (similar).
    16R.8-5 at ¶ 18 (citing People v. Thompson, 
    835 N.E.2d 933
     (Ill. App. Ct.
    2005)).
    No. 18-3435                                                                19
    B.
    We now examine Mr. Olvera’s contention that the state ap-
    pellate court unreasonably applied Strickland when it held
    that the affidavits that he submitted in support of his state pe-
    tition did not, individually or collectively, demonstrate inef-
    fective assistance of counsel. The state appellate court con-
    cluded that Mr. Olvera’s counsel was not ineffective. Yet the
    state court’s opinion left some aspects of Mr. Olvera’s fail-
    ure-to-investigate claim largely unaddressed. As Mr. Olvera,
    the State, and the district court note, the state appellate court’s
    analysis is rather succinct and this factor complicates our
    presentation. 17 For ease of reading, we therefore summarize
    at the outset our conclusions with respect to the merits of
    Mr. Olvera’s specific unreasonable application arguments:
    The state court did not unreasonably apply Strickland when it
    concluded that Espinoza’s affidavit failed to demonstrate
    prejudice and when it concluded that Daniel Mendoza’s,
    Damian Olvera’s, and Michael Olvera’s affidavits failed to
    demonstrate deficient performance; as for Teague’s, Del-
    gado’s, and Perez’s affidavits, even if we assume the state
    court unreasonably applied the deficient performance prong,
    we conclude that those affidavits do not establish prejudice
    (an issue we review de novo because the state court never
    reached the question).
    17 The district court’s opinion is far more thorough, but our task is to re-
    view the state appellate court’s reasoning, not the district court’s. See Wil-
    son v. Sellers, 
    138 S. Ct. 1188
    , 1192 (2018) (“[W]hen the last state court to
    decide a prisoner’s federal claim explains its decision on the merits in a
    reasoned opinion[,] … a federal habeas court simply reviews the specific
    reasons given by the state court and defers to those reasons if they are
    reasonable.”).
    20                                                   No. 18-3435
    1.
    a.
    Because our inquiry depends, in part, on an understand-
    ing of accountability and self-defense under Illinois law, we
    begin by examining the relevant provisions.
    On the fateful evening, Mr. Olvera never fired a shot. Un-
    der Illinois law, however, an individual can be accountable
    for the criminal acts of another if “either before or during the
    commission of an offense, and with the intent to promote or
    facilitate that commission, he or she solicits, aids, abets,
    agrees, or attempts to aid that other person in the planning or
    commission of the offense.” 720 ILCS 5/5-2(c). To prove the
    defendant’s accountability, “the State may present evidence
    that either (1) the defendant shared the criminal intent of the
    principal, or (2) there was a common criminal design.” People
    v. Fernandez, 
    6 N.E.3d 145
    , 149 (Ill. 2014). In Mr. Olvera’s case,
    the State argued that the facts demonstrated both a shared in-
    tent and common design.
    As for Illinois law of self-defense, an individual “is justi-
    fied in the use of force which is intended or likely to cause
    death or great bodily harm only if he reasonably believes that
    such force is necessary to prevent imminent death or great
    bodily harm to himself or another, or the commission of a for-
    cible felony.” 720 ILCS 5/7-1(a). Relying on a self-defense ar-
    gument is much more difficult for an initial aggressor. Under
    Illinois law, someone who “initially provokes the use of force
    against himself” cannot argue self-defense unless, among
    other things, he “exhausted every reasonable means to es-
    cape” the dangerous situation he started. 720 ILCS 5/7-4(c)(1).
    No. 18-3435                                                 21
    b.
    As we turn to an examination of the state appellate court’s
    evaluation of Mr. Olvera’s petition, it is important to keep
    Mr. Olvera’s contentions in proper focus. He bases his appeal
    on the seven affidavits that we recounted earlier. But,
    throughout our consideration of his contentions, we must not
    lose sight of the evidence from Mr. Olvera’s trial that remains
    untouched by his arguments. Raya testified that when
    Mr. Olvera picked her up at the party, he hurried her into the
    Buick, told her to duck down in the back seat, then sped off
    shortly before she heard loud bangs and smelled smoke.
    Rhodes, Mr. Olvera’s cellblock confidant, testified that
    Mr. Olvera confessed to ordering Delgado to shoot at the
    Latin Kings. These two witnesses were important pieces in the
    State’s case, and the core of their testimony remains intact.
    2.
    With these threshold matters in mind, we now address
    those aspects of Mr. Olvera’s petition where our analysis
    tracks the state appellate court’s, then move on to the affida-
    vits where our discussion diverges.
    a.
    In his affidavit, Espinoza claims that Montalvo lied when
    she testified that she had overheard Mr. Olvera admit to Es-
    pinoza that he ordered Delgado to shoot at the Latin Kings.
    Mr. Olvera now apparently faults his counsel for not investi-
    gating this line of impeachment.
    We agree with the state appellate court that Espinoza’s as-
    sertion does not establish prejudice. Mr. Olvera overstates the
    importance of Montalvo’s testimony. By the time she testified
    at Mr. Olvera’s trial, Montalvo had become a largely
    22                                                            No. 18-3435
    uncooperative witness. Her trial testimony consisted mostly
    of her stating that she did not remember the events in ques-
    tion, which required the prosecutor to rely on her earlier
    grand jury testimony. And as we noted, the jury still would
    have heard Raya’s testimony about the moments before the
    shooting and Rhodes’s testimony about Mr. Olvera’s prison
    confession. The state appellate court did not unreasonably ap-
    ply Strickland when it concluded that Espinoza’s information
    did not demonstrate prejudice. 18
    b.
    We next consider the state appellate court’s determination
    that Damian Olvera’s, Daniel Mendoza’s, and Michael
    Olvera’s affidavits failed to demonstrate deficient perfor-
    mance. Even though there was no evidentiary hearing on
    Mr. Olvera’s petition, it is clear from the face of these three
    affidavits that they fail to demonstrate ineffective assistance
    on the part of Mr. Olvera’s trial counsel.
    Damian Olvera states in his affidavit that he overheard Es-
    pinoza tell Montalvo to lie on the stand. He admits in his affi-
    davit, however, that he did not come forward with this infor-
    mation until 2002, well after the trial. Even a thorough
    18 We would be remiss if we did not also mention that Mr. Olvera’s trial
    counsel submitted an affidavit stating that Mr. Olvera “specifically in-
    structed” him “not to interview Mr. Espinoza.” R.8-12 at 8. When a com-
    petent defendant specifically instructs counsel not to interview a potential
    witness, that instruction significantly limits counsel’s investigative obliga-
    tions as to that witness. See Strickland v. Washington, 
    466 U.S. 668
    , 691
    (1984) (a client’s information and instructions define counsel’s responsi-
    bility to investigate). Though the state appellate court did not reach the
    deficient performance question with respect to Espinoza’s affidavit, we
    are hard-pressed to see how Mr. Olvera could make that showing.
    No. 18-3435                                                   23
    investigation would be unlikely to reach Damian Olvera, who
    played no role in the evening of the shooting. Counsel’s in-
    vestigation, therefore, was not deficient for failing to uncover
    Damian Olvera.
    As for Daniel Mendoza, the record shows that
    Mr. Olvera’s counsel did speak with Daniel Mendoza ahead
    of the trial and Mendoza refused to testify on Mr. Olvera’s
    behalf. Moreover, Daniel Mendoza ultimately did testify at
    the trial under a subpoena but overall was reluctant to say
    much of anything during his testimony. Counsel’s investiga-
    tion with respect to Daniel Mendoza was therefore reasona-
    ble; it is only well after the fact that Daniel Mendoza decided
    he has something meaningful to say.
    We reach the same conclusion regarding the state appel-
    late court’s treatment of Michael Olvera’s affidavit. There, Mi-
    chael Olvera claimed that he overheard a conversation be-
    tween Perez, Reyes, and Gutierrez, in which they stated that
    Reyes ran toward Mr. Olvera’s car and that Delgado’s shots
    appeared to have been in response to Reyes’s action. The state
    appellate court concluded that Mr. Olvera’s counsel was not
    ineffective for allegedly failing to uncover this information.
    We agree that, even without holding an evidentiary hearing,
    the state appellate court’s conclusion was reasonable. Unlike
    the other affidavits, Michael Olvera’s affidavit does not claim
    that Mr. Olvera’s counsel failed to contact him before the trial.
    As the state appellate court noted, Michael Olvera’s offer to
    testify about a conversation he allegedly overheard was rank
    hearsay not subject to any identifiable exception. Mr. Olvera
    has offered no case law suggesting counsel has an obligation
    to uncover and then use inadmissible hearsay. Moreover, Mi-
    chael Olvera, Reyes, and Gutierrez all testified at Mr. Olvera’s
    24                                                        No. 18-3435
    trial and were cross examined by Mr. Olvera’s counsel about
    the night of the shooting. 19 None offered the description of
    events that Michael Olvera now claims he overheard.
    3.
    We now turn to those situations where we believe that the
    state appellate court’s assessment of the submitted affidavits
    is problematic and where our own discussion necessarily
    must diverge from that of the state court.
    The state appellate court concluded that Delgado’s affida-
    vit failed to demonstrate deficient performance because coun-
    sel’s decision not to interview Delgado or call him as a witness
    “relate[d] to appropriate trial strategy.”20 For Teague’s affida-
    vit, the state appellate court concluded it was not deficient
    performance to fail to investigate whether Teague’s plans to
    buy Mr. Olvera’s car predated the shooting because that in-
    formation “had no relevance to the commission of the
    crime.” 21 For Perez’s affidavit, in which he says he saw Reyes
    run at Mr. Olvera’s car before the shooting, the state appellate
    court concluded counsel’s failure to interview Perez or call
    him as a witness was not deficient because Perez “was out of
    the state during the trial, and therefore counsel could not have
    called him to testify.” 22
    The issue, as we see it, is not that the state appellate court
    was necessarily wrong in concluding that Mr. Olvera failed to
    19Perez did not testify at Mr. Olvera’s trial. We address his information
    below.
    20 R.8-5   at ¶ 17.
    21 
    Id. 22
     
    Id.
    No. 18-3435                                                    25
    demonstrate deficient performance when he alleged counsel’s
    failure to investigate Teague’s, Delgado’s, and Perez’s infor-
    mation. Rather, the issue is that the state appellate court de-
    cided that counsel’s investigation of these three potential wit-
    nesses was reasonable without nearly enough information
    about the scope of counsel’s investigation. The affidavit sup-
    plied by Mr. Olvera’s trial counsel in the state court proceed-
    ings offered no insight into these three witnesses, and the
    state court held no hearing into counsel’s alleged decision not
    to investigate what information these witnesses had to offer.
    The key principle at stake here is that Strickland’s pre-
    sumption that counsel’s decisions were reasonable “applies
    only if the lawyer actually exercised judgment.” Mosley, 689
    F.3d at 848. When it comes to failure-to-investigate claims,
    Strickland does not permit courts to simply assume counsel
    acted strategically in not calling a witness when the allegation
    is that counsel never investigated what that witness would
    have to say. See id. Without knowing the scope of counsel’s
    pretrial investigation, the state court cannot answer the “crit-
    ical threshold question” of whether that investigation was ad-
    equate. Campbell, 780 F.3d at 766. That is why we have so often
    emphasized—and emphasize again today—that if a state
    court is going to conclude that trial counsel conducted a rea-
    sonable investigation, the state court must actually under-
    stand the scope of counsel’s investigation. See Gish v. Hepp,
    
    955 F.3d 597
    , 604 (7th Cir.), cert. denied, 
    141 S. Ct. 681
     (2020);
    Mosley, 689 F.3d at 848; Davis v. Lambert, 
    388 F.3d 1052
    , 1061
    (7th Cir. 2004).
    Typically, a state court can gain an understanding of trial
    counsel’s investigative decisionmaking by reviewing a de-
    tailed affidavit from counsel or holding an evidentiary
    26                                                 No. 18-3435
    hearing. Cf. Lambert, 
    388 F.3d at 1061
     (discussing the role of
    witness affidavits and state court evidentiary hearings in de-
    veloping the factual record for a failure-to-investigate claim).
    As the Supreme Court noted in Strickland, 
    466 U.S. at 691,
    “when the facts that support a certain potential line of defense
    are generally known to counsel because of what the defend-
    ant has said, the need for further investigation may be consid-
    erably diminished or eliminated altogether.” As Strickland
    also observed, the “inquiry into counsel’s conversations with
    the defendant may be critical to a proper assessment of coun-
    sel’s investigation decisions.” 
    Id.
     The facts in Delgado’s and
    Teague’s affidavits are all facts Mr. Olvera would have known
    and could have told his attorney. But, here, the state court
    lacked insight into Mr. Olvera’s conversations with his trial
    counsel. Without that insight, it could not determine whether
    counsel’s alleged failure to investigate constitutes substand-
    ard performance.
    Given these circumstances, we assume, for present pur-
    poses, that the three affidavits at issue demonstrated deficient
    performance and therefore turn our attention to prejudice. Be-
    cause the state appellate court did not address Strickland’s
    prejudice prong when examining Teague’s, Delgado’s, and
    Perez’s affidavits, our inquiry proceeds de novo. See Gish, 955
    F.3d at 604–05 (assuming the state appellate court unreasona-
    bly applied the deficient performance prong and then moving
    to review de novo the previously unaddressed prejudice
    prong).
    Mr. Olvera submits that the evidence in Teague’s, Del-
    gado’s, and Perez’s affidavits undercuts the State’s accounta-
    bility theory and presents a viable self-defense argument.
    Starting with the affidavits’ impact on the accountability
    No. 18-3435                                                   27
    theory, we conclude that the information in those affidavits
    does not undermine confidence in the jury’s verdict. Teague’s
    claim that he planned to buy Mr. Olvera’s Buick before the
    shooting is hardly significant. Although Mr. Olvera contends
    that the affidavit shows he was not trying to quickly dispose
    of evidence of his criminal activity, Teague’s affidavit easily
    cuts the other way. There is no dispute that Teague obtained
    possession of the car right after the shooting. That Teague
    showed interest in the car ahead of the shooting may just
    mean that Mr. Olvera knew he had a ready buyer and could
    rid himself of evidence that could tie him to the shooting.
    With respect to the claims in Delgado’s affidavit, these con-
    tradict the statements Delgado made during his own sentenc-
    ing that implicated Mr. Olvera in the shooting. It is notable,
    too, that Delgado’s first affidavit filed alongside this petition
    lacked any mention of the more dramatic allegations included
    in his second affidavit. 23
    Even when we consider Teague’s, Perez’s, and Delgado’s
    affidavits, there simply is no reasonable probability that the
    outcome of the case would have been different. The jury still
    would have heard Rhodes testify that Mr. Olvera admitted to
    ordering Delgado to fire at the Latin King members during
    the second shooting. Moreover, the fatal shooting happened
    on the second pass of the Latin King members, and Raya tes-
    tified that Mr. Olvera instructed her to duck down before that
    shooting occurred. That is strong evidence that Mr. Olvera
    was accountable for Delgado’s shooting Stropes.
    Mr. Olvera’s self-defense argument is even less persua-
    sive. Illinois law makes clear that an initial aggressor cannot
    23 See   supra, at 9.
    28                                                   No. 18-3435
    claim self-defense unless he “exhausted every reasonable
    means to escape” the dangerous situation he started. 720
    ILCS 5/7-4(c)(1). The evidence at trial showed that Mr. Olvera
    and Delgado fired one shot on their first pass of the Latin King
    members, then additional shots on their second pass. It was
    during the second pass that Delgado shot and killed Stropes.
    None of the affidavits undercut this basic timeline. As Ramos,
    one of the partygoers, testified at trial, Mr. Olvera pulled up,
    yelled for Raya to hurry into the car, then “took off real fast”
    shortly before Ramos heard the second set of gunshots. 24 And
    the jury could reasonably interpret Raya’s testimony about
    Mr. Olvera’s telling her to duck in the backseat to mean that
    Mr. Olvera knew more violence was about to occur.
    The shooting on the first pass made Mr. Olvera and Del-
    gado the initial aggressors. Self-defense, therefore, is unavail-
    able to them unless they exhausted every reasonable means
    of escape. It is clear that they did not. Instead, they drove back
    in the direction of the Latin Kings and fired more shots. Even
    if Reyes did run at the car, that does not change the fact that
    Mr. Olvera and Delgado were the initial aggressors. As a re-
    sult, Mr. Olvera’s new self-defense claim does not undermine
    confidence in the verdict.
    Even if we assume that Mr. Olvera’s trial counsel con-
    ducted an inadequate investigation into Delgado’s, Teague’s,
    and Perez’s information, there was no prejudice to
    Mr. Olvera. Looking at all of the information available, the
    overwhelming evidence supports the State’s accountability
    theory and precludes Mr. Olvera’s claim of self-defense.
    24 R.8-16   at 82.
    No. 18-3435                                                  29
    Conclusion
    In sum, we reject Mr. Olvera’s contention that the state
    court’s articulation of the Strickland standard was contrary to
    the Supreme Court’s clearly established law. We also hold
    that the state court did not unreasonably apply Strickland
    when it concluded that Espinoza’s affidavit failed to demon-
    strate prejudice and when it concluded that Daniel Men-
    doza’s, Damian Olvera’s, and Michael Olvera’s affidavits
    failed to demonstrate deficient performance. As for Teague’s,
    Delgado’s, and Perez’s affidavits, even if we assume the state
    court unreasonably applied the deficient performance prong,
    those affidavits do not establish prejudice.
    Consequently, Mr. Olvera has failed to demonstrate inef-
    fective assistance by his trial counsel. We therefore affirm the
    district court’s denial of his petition under 28 U.S.C. § 2254.
    AFFIRMED