United States v. Matthew William Wheeler ( 2021 )


Menu:
  • USCA11 Case: 17-15003     Date Filed: 10/21/2021   Page: 1 of 53
    [PUBLISH]
    In the
    United States Court of Appeals
    For the Eleventh Circuit
    ____________________
    No. 17-15003
    ____________________
    UNITED STATES OF AMERICA,
    Plaintiff-Appellant-Cross-Appellee,
    versus
    MATTHEW WILLIAM WHEELER,
    a.k.a. Matthew Williams,
    JAMES WAYNE LONG,
    a.k.a. J.W. Long,
    Defendants-Appellees-Cross-Appellants.
    ____________________
    Appeal from the United States District Court
    for the Southern District of Florida
    D.C. Docket No. 1:16-cr-20715-MGC-9
    ____________________
    USCA11 Case: 17-15003    Date Filed: 10/21/2021   Page: 2 of 53
    2                   Opinion of the Court                17-15003
    ____________________
    No. 17-15030
    ____________________
    UNITED STATES OF AMERICA,
    Plaintiff-Appellee,
    versus
    ANITA SGARRO,
    a.k.a. Anita Simone,
    CHARLES DAVID SMIGROD,
    a.k.a. Charles David,
    CHARLES K. TOPPING,
    a.k.a. CharlieKenn,
    Defendants-Appellants.
    ____________________
    Appeal from the United States District Court
    for the Southern District of Florida
    D.C. Docket No. 1:16-cr-20715-MGC-5
    ____________________
    USCA11 Case: 17-15003      Date Filed: 10/21/2021   Page: 3 of 53
    17-15003              Opinion of the Court                      3
    ____________________
    No. 17-15379
    ____________________
    UNITED STATES OF AMERICA,
    Plaintiff-Appellant-Cross-Appellee,
    versus
    JAMES WAYNE LONG,
    MATTHEW WILLIAM WHEELER,
    a.k.a. Matthew Williams,
    Defendants-Appellees-Cross-Appellants.
    ____________________
    Appeal from the United States District Court
    for the Southern District of Florida
    D.C. Docket No. 1:16-cr-20715-MGC-12
    ____________________
    Before WILSON, LAGOA, and BRASHER, Circuit Judges.
    PER CURIAM:
    USCA11 Case: 17-15003        Date Filed: 10/21/2021     Page: 4 of 53
    4                      Opinion of the Court                 17-15003
    This is a criminal appeal involving five co-defendants who
    were charged with wire fraud, mail fraud, and conspiracy for their
    alleged involvement in a telemarketing scheme to defraud stock
    investors. After an eight-week trial, in which the defendants made
    several motions for mistrial, the jury found each defendant guilty
    on all counts. At a post-trial hearing, the district court found that
    the prosecution had acted improperly in closing argument but de-
    nied the defendants’ motions for mistrial. The court then granted
    judgments of acquittal based on insufficient evidence as to defend-
    ants Matthew Wheeler and James Wayne Long. As to the other
    three defendants—Charles Smigrod, Anita Sgarro, and Charles
    Topping—the court denied their motions for judgment of acquit-
    tal.
    The government appeals the judgments of acquittal granted
    to Wheeler and Long. Smigrod, Sgarro, and Topping appeal the
    court’s denial of their motions for judgment of acquittal and argue
    that, in the alternative, prosecutorial misconduct warranted a mis-
    trial. Additionally, Sgarro and Topping argue that we should re-
    verse their convictions based on a variety of erroneous evidentiary
    rulings and improper jury instructions. They also challenge their
    sentences. Wheeler and Long join their co-defendants in arguing
    on cross-appeal that even if we find that the jury’s verdict was based
    on sufficient evidence, prosecutorial misconduct tainted the trial
    and requires us to reverse the convictions. We consolidated the
    appeals and cross-appeal after briefing on the merits.
    USCA11 Case: 17-15003       Date Filed: 10/21/2021     Page: 5 of 53
    17-15003               Opinion of the Court                        5
    After careful review and with the benefit of oral argument,
    we reverse the judgments of acquittal granted to Wheeler and
    Long because there is a reasonable construction of the evidence
    that supports the jury’s verdicts as to those two defendants. We
    also find that sufficient evidence supported the convictions of
    Sgarro, Smigrod, and Topping. On the question of prosecutorial
    misconduct, we find that the prosecution’s behavior at trial did not
    rise to the level of misconduct. Nor do any of the district court’s
    evidentiary rulings or the jury instructions warrant reversal. Fi-
    nally, we find no error in the district court’s sentencing of Sgarro,
    Smigrod, and Topping.
    I.    Factual and Procedural Background
    A.     The Stock-Selling Operation
    The government charged the defendants in this case for
    their alleged roles in a telemarketing scheme that tricked investors
    into making stock purchases and misappropriated their money.
    The defendants operated from two phone rooms. One was located
    in Florida and managed by Miguel Mesa; the other was located in
    California and managed by Sgarro. Both phone rooms initially sold
    the stock of a company called Sanomedics International Holdings;
    the Florida phone room later switched to selling the stock of a gam-
    ing company called Fun Cool Free, Inc. (FCF). Sanomedics, which
    was run by Craig Sizer (owner and founder), Keith Houlihan
    (CEO), and Mesa (in charge of Sanomedics’ stock-selling
    USCA11 Case: 17-15003          Date Filed: 10/21/2021        Page: 6 of 53
    6                        Opinion of the Court                    17-15003
    operation),1 manufactured and marketed a non-contact infrared
    thermometer. Its stock was listed on the over-the-counter mar-
    ket. 2 FCF sold a video game that was available for purchase on
    Apple’s App Store.
    Each of the five defendants had a particular role in the stock-
    selling operation. Wheeler, Long, and Smigrod worked as sales-
    people in the Florida phone room. Wheeler, who sometimes used
    the alias “Matt Williams” with investors, sold Sanomedics stock
    from 2009 to 2012 and FCF stock during 2015. Long, using his real
    name with investors, joined the operation near the tail end in 2015.
    He sold FCF stock for about six months and had no involvement
    in Sanomedics stock. Smigrod, going by the name “Charles David”
    (his first and middle names), started out selling Sanomedics stock
    and later transitioned to selling FCF stock.
    The government alleged that Sgarro and Topping per-
    formed a broader range of functions. Sgarro, as the manager of the
    California office, instructed salespeople on how to pitch Sanomed-
    ics stock. She also personally pitched and sold the stock, introduc-
    ing herself to investors as “Anita Simone” from Sanomedics’ “in-
    vestor relations.” Topping, using the moniker “Charlie Kenn,”
    1 Sizer, Houlihan, and Mesa were all charged in the same indictment as the
    defendants but pleaded guilty before trial.
    2 An “over-the-counter” market is a securities market that functions without
    a central exchange. Alan R. Bromberg & Lewis D. Lowenfels, Bromberg &
    Lowenfels on Securities Fraud § 1:2 (2d ed.).
    USCA11 Case: 17-15003       Date Filed: 10/21/2021     Page: 7 of 53
    17-15003               Opinion of the Court                        7
    worked in the Florida office where he was a “loader.” As a loader,
    Topping played an important role in a two-part process that was
    often used by both the California and Florida locations. First, a
    salesperson would cold call a potential investor using an autodialer
    and would pitch an investment opportunity. Then, if the salesper-
    son secured an initial investment, he would offer to connect the
    investor with a high-level executive who would be able to sell spe-
    cial “institutional” shares—or so investors were told. The govern-
    ment alleged that Topping would play the role of that high-level
    executive. If Topping successfully sold the investor additional “in-
    stitutional” shares, both he and the referring salesperson would
    make a commission on that transaction. Topping first sold Sano-
    medics stock and later sold FCF stock.
    Both the Florida and California phone rooms used prepared
    written materials to aid their pitches to investors. There were press
    releases, for example, that were created by Houlihan, posted
    online, and distributed to the sales team by Mesa. Sizer and Mesa
    also developed scripts for salespeople to use in their pitches. The
    press releases and scripts contained some accurate information but
    also included exaggerations and fabrications. Working partly off
    these press releases and scripts, salespeople made false representa-
    tions to potential investors, including that they were paid only in
    company stock and did not work for commissions, and that im-
    portant businesspeople or celebrities were involved in the com-
    pany they were pitching.
    USCA11 Case: 17-15003          Date Filed: 10/21/2021       Page: 8 of 53
    8                        Opinion of the Court                    17-15003
    In addition, although Sanomedics and FCF were real com-
    panies with real products—and investors were really issued stock
    in the companies—investors were misled as to how their money
    would be used. Salespeople often assured investors that their
    money would fund business development and expansion plans, but
    very little of the money actually was used for such purposes. Most
    of the $23 million or so that investors sunk into Sanomedics and
    FCF stock went to Sizer and Mesa. A little less than twenty percent
    was split among ten salespeople, including the five trial defendants.
    Between Sanomedics and FCF, the defendants took in the follow-
    ing sums: $1,207,000 for Topping, $1,070,000 for Sgarro, $162,000
    for Wheeler, $148,000 for Smigrod, and $17,000 for Long.
    B.      The Indictment
    A September 2016 indictment charged thirteen co-defend-
    ants with substantive mail and wire fraud and conspiracy to com-
    mit mail and wire fraud. The indictment alleged conspiracy to
    commit mail and wire fraud under 
    18 U.S.C. § 1349
    : Count 1 in
    connection with Sanomedics and Count 11 in connection with
    FCF. Counts 2 through 10 alleged mail fraud, 
    18 U.S.C. § 1341
    , and
    wire fraud, 
    18 U.S.C. § 1343
    , in connection with the sale of Sano-
    medics stock. 3 Counts 12 and 13 alleged mail fraud in connection
    with the sale of FCF stock.
    3 The government voluntarily dismissed Counts 4 and 5 before trial.
    USCA11 Case: 17-15003       Date Filed: 10/21/2021   Page: 9 of 53
    17-15003              Opinion of the Court                       9
    Eight defendants—including ringleaders Sizer and Mesa—
    pleaded guilty. The other five—Wheeler, Long, Smigrod, Sgarro,
    and Topping—went to trial.
    C.     The Trial
    Trial began in April 2017 and lasted eight weeks. Among the
    witnesses who testified in the government’s case-in-chief were two
    of the alleged co-conspirators who had pleaded guilty, sixteen in-
    vestors who had bought stock from the defendants, and a confiden-
    tial informant, Stuart Rubens, who took a job undercover in the
    Florida phone room. Rubens testified that he had previously been
    convicted of fraud for participating in a scheme similar to Sano-
    medics. He highlighted similarities between the Florida phone
    room and other criminal operations he had experience with, which
    he referred to as “boiler rooms.” The government also presented
    evidence of the defendants’ sales scripts and commission sheets,
    along with other documents seized in an FBI raid of the Florida
    phone room.
    The record on appeal shows that the trial evidence included
    the following. All the defendants but Long used an alias or nick-
    name with investors. The defendants often claimed that they were
    employees of the company they were pitching, although in truth
    they worked for a company called Hemisphere Management that
    was owned by Mesa. They claimed that they were paid only in
    company stock, though they really made commissions of roughly
    fifteen to twenty percent. They told investors that they were of-
    fering stock in a company that was making—or was on the verge
    USCA11 Case: 17-15003       Date Filed: 10/21/2021    Page: 10 of 53
    10                     Opinion of the Court                17-15003
    of making—millions in profits, although the company was never
    really close to generating profits of that scale. And they sometimes
    told investors that the company they were pitching was associated
    with a celebrity, a well-known business executive, or another suc-
    cessful company—representations that were based on half-truths
    or totally false.
    After the government rested, the defendants called several
    of their own witnesses. Houlihan, as the author of the Sanomedics
    press releases, testified that much of the information in the press
    releases was true although embellished. And Gary Miller, a sales-
    person who worked for Sgarro in the California phone room, testi-
    fied that he pitched Sanomedics stock to investors using infor-
    mation he gathered from conference calls with Mesa. According
    to Miller’s testimony, he believed that the information from Mesa
    was true, and that Sanomedics was a good investment.
    Miller’s testimony took a twist on cross-examination when
    the prosecution asked whether he had “a different business” with
    Sgarro. Miller denied having any business association with Sgarro
    aside from selling Sanomedics stock. After some discussion with
    counsel at sidebar, the court determined—over objections from the
    defendants—that the prosecution could impeach Miller with a
    prior inconsistent statement he had made to an FBI agent, to the
    effect that he sold drugs with Sgarro. The district court ruled that,
    under Federal Rule of Evidence 613(b), the prosecution could ask
    Miller: “[D]id you ever tell Agent X that you sold marijuana to de-
    fendant Sgarro[?]” And “[i]f the witness denies making the
    USCA11 Case: 17-15003       Date Filed: 10/21/2021    Page: 11 of 53
    17-15003               Opinion of the Court                       11
    statement,” the court determined, “the agent can be called to say
    he made the statement.” Before proceeding, the court read the fol-
    lowing limiting instruction to the jury:
    [Y]ou determine the credibility of the wit-
    nesses in this case. You decide who to believe and
    what to believe. You are about to hear testimony that
    this witness testified differently than his trial testi-
    mony. . . . The testimony is not offered to prove the
    truth of the statement. It is offered only to assist you
    in determining the credibility of this witness.
    The prosecution proceeded to cross-examine Miller. In
    what was apparently a volatile exchange, the prosecutor asked Mil-
    ler several times if he told the FBI that he sold marijuana with
    Sgarro. Miller repeatedly denied having done so. In rebuttal, the
    prosecution called an FBI agent who testified that Miller had in fact
    made statements about selling drugs with Sgarro.
    After the defendants rested, the parties’ attention turned to
    the jury instructions. To provide some context, this case went to
    trial in 2017 on the heels of our decision in United States v. Takha-
    lov, 
    827 F.3d 1307
     (11th Cir.), opinion modified on denial of
    reh’g, 
    838 F.3d 1168
     (11th Cir. 2016) (per curiam). Based on the
    recent Takhalov decision, the defendants sought a jury instruction
    after the close of evidence to convey the distinction between an
    intent to deceive and an intent to defraud. The defendants’ theory
    was this: evidence that the defendants used aliases or misrepre-
    sented their employer, for example, might prove the defendants’
    USCA11 Case: 17-15003          Date Filed: 10/21/2021        Page: 12 of 53
    12                        Opinion of the Court                    17-15003
    intent to deceive, but it would not prove their intent to defraud so
    long as the investors got the stock they bargained for at the agreed-
    upon price.
    The defendants requested that the court insert clarifying lan-
    guage from Takhalov in the jury instruction on the elements of the
    offenses. Because the pattern instructions had not been updated
    since Takhalov, the defendants argued that there was “confusion
    between . . . the . . . instructions as they exist in the pattern versus
    what Takhalov has cleared up.” 4 The court decided that it would
    give the standard pattern instructions unaltered, but would also
    give a theory-of-defense instruction that incorporated Takhalov.
    The government would be allowed to attack the theory of defense,
    the court explained, by arguing to the jury that the theory “doesn’t
    match up with the evidence.”
    In a charge conference on the twenty-seventh day of trial,
    the government argued that telling the jury about Takhalov’s hold-
    ing—even in a theory-of-defense instruction—would only confuse
    them. The court maintained that while it would not alter the in-
    structions on the elements of wire fraud, it would allow an instruc-
    tion that presented the defendants’ theory that they lacked an in-
    tent to harm under our holding in Takhalov.
    4 We have since adopted new pattern jury instructions for mail and wire fraud
    that incorporate Takhalov’s holding. See 11th Cir. Std. (Crim.) Jury Instr.
    O50.3, O51 (2019) (“Proving intent to deceive alone, without the intent to
    cause loss or injury, is not sufficient to prove intent to defraud.”).
    USCA11 Case: 17-15003       Date Filed: 10/21/2021      Page: 13 of 53
    17-15003               Opinion of the Court                        13
    The prosecutor expressed concern that the jury would lend
    too much credence to the instruction if it came from the court.
    [PROSECUTOR]: [The defendants] . . . can argue all
    that until they get blue, and if they create a reasonable
    doubt, they win, God bless them, but when it comes
    from Your Honor’s mouth instructing the jury on
    that theory, that’s where I take issue. I believe it’s un-
    necessary.
    The court stressed, however, that the law did not require
    the jury to accept the defendants’ Takhalov defense. The applica-
    tion of the defense theory to the facts of this case was for the jury
    to determine.
    THE COURT: I don’t want the jury to think this is
    part of the actual substantive instruction. I want
    them to understand this is defendants’ theory. . . .
    [PROSECUTOR]: Judge, we’re making some pro-
    gress. We agree that as long as it’s clear that this is
    the defendants’ theory, not coming from Your
    Honor—
    ...
    THE COURT: That’s why I’m trying to take out an-
    ything that says the law requires or this requires.
    ...
    THE COURT: “[T]his is a theory, not a legal instruc-
    tion.”
    USCA11 Case: 17-15003       Date Filed: 10/21/2021    Page: 14 of 53
    14                     Opinion of the Court                17-15003
    When the charge conference resumed the next day of trial,
    the defense continued to argue that Takhalov was the law—it had
    clarified that proving deception alone is not enough to sustain a
    conviction. That legal framework should be presented to the jury,
    the defendants argued, so that the jury could decide whether the
    government proved more than mere deception in this case. The
    prosecution, for its part, continued to spar with the court about
    whether the theory-of-defense instruction, which highlighted that
    the government had to prove more than an intent to deceive, was
    necessary.
    THE COURT: If you read the opening paragraph of
    Takhalov, it says the wire fraud statute does not enact
    as federal law the 9th commandment given to Moses
    on Sinai. . . .
    [D]eceiving does not always involve harming another
    person. Defrauding does. That a defendant merely
    induced the victim to enter into a transaction that he
    otherwise would have avoided is, therefore, insuffi-
    cient to show wire fraud. . . .
    [PROSECUTOR]: But that’s not the holding, Judge.
    That’s not the holding. That’s their introductory re-
    marks. . . .
    ...
    THE COURT: I don’t agree . . . .
    Still, the prosecutor continued to argue that proving the de-
    fendants’ intent to deceive was sufficient and that a contrary ruling
    would be irreconcilable with the text of the mail fraud statute.
    USCA11 Case: 17-15003       Date Filed: 10/21/2021    Page: 15 of 53
    17-15003               Opinion of the Court                       15
    [PROSECUTOR]: Do you believe Takhalov over-
    ruled the mail fraud statute, Your Honor, or changed
    it?
    ...
    Judge, the wire fraud and mail fraud statutes have the
    words.
    ...
    You read it yourself, deceit is a form in which to prove
    the fraud.
    THE COURT: We’re going to have to agree to disa-
    gree here.
    The district court thus maintained that it would not alter the
    instructions on the wire-fraud elements, but that it would include
    a separate theory-of-defense instruction. The finalized instruction
    read as follows:
    It is Defendants’ Theory of Defense that they
    did not knowingly and intentionally participate in a
    fraud, and that the fraud actually occurred when, un-
    beknownst to the Defendants, the owners and offic-
    ers of the corporations, Craig Sizer and Keith Houli-
    han, together with Miguel Mesa, stole millions of dol-
    lars in investor funds. Although they may have been
    aggressive salesmen, they did not intend to defraud
    the investors, and that they were unwitting pawns in
    the scheme orchestrated by the owners and Mesa to
    steal investors’ funds [sic].
    USCA11 Case: 17-15003      Date Filed: 10/21/2021     Page: 16 of 53
    16                     Opinion of the Court                17-15003
    Defendants contend that there is a difference
    between deceiving and defrauding. To defraud the
    defendants must have intended to use deception to
    cause financial harm to investors.
    In other words, if a defendant selling stock rea-
    sonably believes that it had value, and the price
    charged reflects that value, his or her use of deceit
    would not constitute a scheme to defraud. Deceiving
    does not always involve harming another person; de-
    frauding does.
    Fraud requires proof of a material misrepre-
    sentation, or the omission or concealment of a mate-
    rial fact. Mere “puffing” or “seller’s talk” is insuffi-
    cient.
    Although the defendants had objected to framing the in-
    struction as a theory of defense rather than an instruction on the
    elements of the offense, they relied heavily on the instruction in
    closing argument. Even if the defendants unwittingly deceived in-
    vestors, they argued, the investors still got the number of shares
    they bargained for at the price they bargained for. There was no
    intent to harm investors, the argument went, because the defend-
    ants thought they were selling valuable stock; they were unaware
    that higher-ups in the organization were victimizing the investors
    by pocketing the bulk of their investments.
    Next came the prosecution’s rebuttal. The prosecution ar-
    gued to the jury that the theory-of-defense instruction was “not the
    Court’s instruction” [but] merely [the defense’s] theory.”
    USCA11 Case: 17-15003        Date Filed: 10/21/2021       Page: 17 of 53
    17-15003                Opinion of the Court                         17
    “[H]undreds of years ago,” he said, “it was a theory that the sun
    revolved around the earth, and that the earth itself was flat. But as
    we know, evidence has disproven those theories, similar to the case
    at bar.” Shortly thereafter, the prosecution returned to the theory
    of defense, directing the jury’s attention to page 23 of the court’s
    instructions. “This page is very special,” he told the jury. “It’s spe-
    cial because it’s not the law. It is not the law. Of the 32 pages in
    your hands, 23 is no more—” Defense counsel then objected, but
    the court overruled the objection and the prosecutor continued.
    [PROSECUTOR]: It’s a theory of defense. Okay? 23
    is, as you can see at the top—we’ll go through it real
    quick—it is the defendant’s theory of defense. That’s
    their theory. That’s the one where it says they were
    unwitting. The next paragraph says the defendants
    contend. They already told you what they contend.
    They already had a shot. They all went. Some talked
    for an hour. Some talked for an hour and 40 minutes.
    . . . The rest of this page, again, is what they believe. .
    . . So again, page 23, it’s in the packet. The Judge ob-
    viously thought it was important for you to have it.
    But ladies and gentlemen, this is not the law. This is
    a theory of defense.
    [DEFENSE COUNSEL]: Objection, Judge.
    THE COURT: Sustained.
    After the prosecutor concluded his closing argument and the
    jury exited the courtroom, the court immediately admonished the
    prosecutor for his “this-is-not-the-law” argument.
    USCA11 Case: 17-15003        Date Filed: 10/21/2021       Page: 18 of 53
    18                     Opinion of the Court                    17-15003
    THE COURT: You know, [ ] we worked very hard to
    craft instructions in this case that maybe not everyone
    agreed with but [ ] that I thought would pass legal
    muster and for you to take the defendants’ theory of
    the case instruction and basically tell the jury it’s a
    joke, that was wrong. Wrong. Now, whether it rises
    to the level of mistrial, . . . I’ll take care of at a later
    point . . . . [B]ut you basically trashed the instruction
    they asked for . . . .
    The defense didn’t like it, you didn’t like it. It’s not
    up to whether people like it, it’s as to whether I think
    it’s a correct statement of the law, and I thought it
    was.
    The court then brought the jury back into the courtroom
    and restated its general charge, including the following instruction:
    “You must follow the law as I explained it to you, even if you do
    not agree with it. And you must follow all of my instructions as a
    whole. You must not single out or disregard any of the instructions
    that I gave you in this trial.”
    After sending the jury to deliberate, the court reiterated to
    counsel that the theory-of-defense instruction “was appropriate
    given the law and the facts of this case.” The court questioned why
    the prosecution “would take our hard work . . . [in] crafting [the
    instruction], and basically almost tell the jury it doesn’t matter.”
    The government “can’t just disregard an instruction because . . .
    they don’t like it,” the court added.
    USCA11 Case: 17-15003       Date Filed: 10/21/2021     Page: 19 of 53
    17-15003               Opinion of the Court                        19
    Subsequently, the court heard oral argument on the defend-
    ants’ motions for mistrial. Counsel for Sgarro pointed out that a
    theory-of-defense instruction incorporates the defense’s view of
    the facts, “but the law is what [the court] give[s] the jury.” “We
    may not agree with the instruction that you gave,” he said, “but
    that was your ruling.” He argued that the government’s decision
    to “try to get this jury to do something contrary to what this Court
    did,” was emblematic of a “pattern of misconduct” that had perme-
    ated the trial. Counsel for the other defendants voiced similar con-
    cerns. The prosecutor responded that he was “merely showing the
    jury that [the defendants’] theory [was] not one supported by the
    facts at hand.” And in his view, he reiterated, the theory of defense
    was not the law. The court took the motions for mistrial under
    advisement. After two days of deliberation, the jury convicted
    each of the five defendants on all counts.
    D.     Post-Trial Hearing and Sentencing
    Following the trial, the defendants moved for judgments of
    acquittal, and the district court convened a hearing to take up the
    post-trial motions. In reference to the motions for mistrial, the
    court expressed at the hearing that, in hindsight, it “should have
    sustained [the defense’s] original objection instead of overruling it”
    and “should have maybe given a more forceful instruction to the
    jury including rereading the theory of defense instruction.” The
    court questioned the propriety of the government’s argument and
    noted the magnitude of the harm it likely caused.
    USCA11 Case: 17-15003       Date Filed: 10/21/2021      Page: 20 of 53
    20                     Opinion of the Court                  17-15003
    THE COURT: [H]ere’s my problem . . . . Constantly
    through[out] the trial . . . , the Court tells the jurors,
    listen, at the end of the trial, I am going to give you
    instructions, and this is the law that you are to follow.
    Essentially what [the prosecutor] did was to negate
    every time I said that. He basically said, just don’t
    worry your pretty little heads about that instruction.
    The prosecution insisted that its closing argument essen-
    tially aligned with the court’s directions at the charge conference
    that the theory of defense was not a legal instruction. The court
    countered that what is said between the parties and the court when
    “hammering out an instruction” is quite different than what can be
    properly argued to the jury based on the court’s final ruling.
    Nevertheless, the court denied the defendants’ motion for a
    mistrial. It found that the defendants were entitled to a theory-of-
    defense instruction, and that “the Government . . . chose an im-
    proper path” in its closing argument. But the court also stated that
    it could not find support for the proposition that a theory-of-de-
    fense instruction is the law. For that reason, the court denied the
    defendants’ motion.
    The court, however, granted judgments of acquittal to
    Wheeler and Long, holding that the government did not prove that
    those defendants knew about the overarching scheme orchestrated
    by Sizer and Mesa to misappropriate investors’ money. As to the
    remaining three defendants, the district court denied their motions
    for judgment of acquittal.
    USCA11 Case: 17-15003       Date Filed: 10/21/2021     Page: 21 of 53
    17-15003               Opinion of the Court                        21
    At sentencing, the Presentence Investigation Reports (PSI)
    held Smigrod and Sgarro responsible for a loss of $22,278,000 and
    Topping responsible for a loss of $22,456,186. Sgarro’s PSI applied
    a two-level enhancement for the use of sophisticated means during
    the offense. Topping’s PSI included a three-level enhancement for
    acting as a manager or supervisor. All three defendants had a crim-
    inal history category of I. The district court sentenced each of them
    to a prison term: 48 months for Smigrod, 116 months for Sgarro,
    and 113 months for Topping. The court also ordered those three
    defendants to serve a three-year term of supervised release and to
    pay restitution.
    II.   Discussion
    On appeal, Smigrod, Sgarro, and Topping argue that there
    was insufficient evidence to support a number of their convictions.
    The government cross-appeals the judgments of acquittal as to
    Wheeler and Long, arguing that the guilty verdicts should be rein-
    stated because they were supported by sufficient evidence. The
    defendants argue that even if sufficient evidence supported their
    convictions, the district court should have granted a mistrial be-
    cause the prosecution engaged in misconduct. Sgarro and Topping
    also appeal a handful of other issues related to evidentiary and sen-
    tencing rulings, as well as the jury instructions.
    In the following sections, we will address: (1) the sufficiency
    of the evidence, (2) the defendants’ allegations of prosecutorial mis-
    conduct, (3) the district court’s evidentiary rulings and the jury in-
    structions; and (4) the district court’s sentencing determinations.
    USCA11 Case: 17-15003        Date Filed: 10/21/2021      Page: 22 of 53
    22                      Opinion of the Court                  17-15003
    When considering the sufficiency of the evidence, we review de
    novo both the grant and denial of a judgment of acquittal, viewing
    the evidence in the light most favorable to the government and
    drawing all reasonable inferences in its favor. Butcher v. United
    States, 
    368 F.3d 1290
    , 1297 (11th Cir. 2004); United States v. Capers,
    
    708 F.3d 1286
    , 1296 (11th Cir. 2013). We will not overturn the jury
    verdict “if any reasonable construction of the evidence would have
    allowed the jury to find the defendant guilty beyond a reasonable
    doubt.” Capers, 708 F.3d at 1296. When reviewing allegations of
    prosecutorial misconduct in closing argument, our review is de
    novo. United States v. Eckhardt, 
    466 F.3d 938
    , 947 (11th Cir. 2006).
    We review for abuse of discretion a district court’s evidentiary rul-
    ings. United States v. Williford, 
    764 F.2d 1493
    , 1497 (11th Cir.
    1985). And in reviewing a district court’s application of the Sen-
    tencing Guidelines, “we review purely legal questions de novo, and
    the district court’s factual findings for clear error.” United States v.
    Monzo, 
    852 F.3d 1343
    , 1348 (11th Cir. 2017).
    A.     Sufficiency of the Evidence
    This appeal requires us to determine whether sufficient evi-
    dence supported both the conspiracy and substantive counts of
    mail and wire fraud as to the five defendants. To prove a conspir-
    acy to commit mail and wire fraud under 
    18 U.S.C. § 1349
    , the gov-
    ernment must prove that: “(1) a conspiracy existed; (2) the defend-
    ant knew of it; and (3) the defendant knowingly and voluntarily
    joined it.” United States v. Moran, 
    778 F.3d 942
    , 960 (11th Cir.
    2015).
    USCA11 Case: 17-15003       Date Filed: 10/21/2021    Page: 23 of 53
    17-15003               Opinion of the Court                       23
    As to substantive mail and wire fraud, the government must
    prove that a person: “(1) intentionally participate[d] in a scheme or
    artifice to defraud another of money or property, and (2) use[d] or
    ‘cause[d]’ the use of the mails or wires for the purpose of executing
    the scheme or artifice.” United States v. Langford, 
    647 F.3d 1309
    ,
    1320 (11th Cir. 2011). To prove that a defendant had the intent to
    defraud, the Government has to prove that the defendants either
    knew they were making false representations or acted with “reck-
    less indifference to the truth.” See United States v. Simon, 
    839 F.2d 1461
    , 1466, 1470 (11th Cir. 1988).
    We have explained that the government must prove not
    only that the defendants had the intent to deceive, but also that
    they intended to harm the victim, meaning that they intended to
    deceive the victim about something that affected the value of the
    bargain. Takhalov, 827 F.3d at 1313. A defendant can do this in
    two ways: (1) lying about the price; or (2) lying about the charac-
    teristics of the good. Id. at 1313–14. However, if a defendant lies
    and says, for example, “that he is the long-lost cousin of a prospec-
    tive buyer—then he has not . . . ‘schemed to defraud’” because that
    misrepresentation does not go to the value of the bargain. Id. at
    1314.
    Takhalov used two hypotheticals to draw the key distinction
    between deceptions that do not affect the value of the bargain and
    those that do:
    Consider the following two scenarios. In the
    first, a man wants to exchange a dollar into four
    USCA11 Case: 17-15003       Date Filed: 10/21/2021     Page: 24 of 53
    24                     Opinion of the Court                 17-15003
    quarters without going to the bank. He calls his
    neighbor on his cell phone and says that his child is
    very ill. His neighbor runs over, and when she arrives
    he asks her to make change for him. She agrees; the
    quarters pass to the man; the dollar passes to the
    woman; and they part ways. She later learns that the
    child was just fine all along. The second scenario is
    identical to the first, except that instead of giving the
    woman a true dollar, he gives her a counterfeit one.
    The first scenario is not wire fraud; the second
    one is. Although the transaction would not have oc-
    curred but-for the lie in the first scenario—the
    woman would have remained home except for the
    phony sickness—the man nevertheless did not intend
    to deprive the woman of something of value by trick,
    deceit, and so on. But in the second scenario he did
    intend to do so.
    Id. at 1313 (footnote, citation, and quotation marks omitted and
    alteration adopted). So to prove substantive mail and wire fraud,
    the government must prove more than deceit; it must prove that
    the defendant intended to deprive the victim of something of value.
    Id.
    1.     Wheeler’s and Long’s Convictions
    We discuss first the government’s appeal as to the suffi-
    ciency of the evidence supporting the convictions of Wheeler and
    Long. The indictment charged Long under Counts 11 (FCF con-
    spiracy) and 12 (substantive mail fraud involving the sale of FCF
    stock to an investor identified as R.L.), and it charged Wheeler
    USCA11 Case: 17-15003        Date Filed: 10/21/2021   Page: 25 of 53
    17-15003                 Opinion of the Court                     25
    under Counts 1 (Sanomedics conspiracy), 11 (FCF conspiracy) and
    13 (substantive mail fraud involving the sale of FCF stock to an in-
    vestor identified as B.M.).
    a.      Substantive Counts
    On the substantive counts—Count 12 against Long and
    Count 13 against Wheeler—the district court found that the gov-
    ernment did not prove the defendants’ intent to defraud investors.
    On appeal, the government points to evidence that Wheeler and
    Long lied about their employer, lied about their compensation
    structure, and misled investors about the value of the stock they
    were selling. After a thorough review of the record, we agree with
    the government.
    It is true that some of the misrepresentations Wheeler and
    Long made to investors were not the type of fraudulent misrepre-
    sentations that Takhalov requires the government to prove. Take,
    for example, the claims that Wheeler and Long were employees of
    the company they were pitching, or that Wheeler’s name was
    “Matt Williams.” Like pretending to be the victim’s long-lost
    cousin or feigning the illness of one’s child, these deceptions would
    not affect the price or characteristics of the good being sold. See
    Takhalov, 827 F.3d at 1313–14. Therefore, neither the salesper-
    son’s employment status nor his use of an alias would amount to a
    scheme to defraud.
    There is evidence in the record, however, of Wheeler and
    Long making misrepresentations or failing to disclose information
    USCA11 Case: 17-15003        Date Filed: 10/21/2021      Page: 26 of 53
    26                      Opinion of the Court                  17-15003
    that a reasonable jury could find affected the nature of the bargain.
    Both Wheeler and Long misled investors to believe that FCF had
    made millions of dollars in profit and was closely associated with
    high-profile companies and executives. Long told investors that
    Apple was a partner with FCF and that John Sculley, a renowned
    CEO, was involved in the company. Wheeler similarly told inves-
    tors, including B.M., that Sculley was closely involved in FCF.
    Wheeler and Long also misrepresented their form of compensa-
    tion, telling investors that they were paid only in company stock
    and did not make commissions on stock sales. The investors’
    money, they said, would go back into the company. Finally, Long
    misled investor R.L. to believe that FCF would be listed on the
    NASDAQ stock exchange and “would be going public.”
    A reasonable jury could infer that facts like FCF’s profits, its
    association with a famous executive and a globally recognized tech-
    nology company, and a potential listing on a major stock exchange
    are essential characteristics of the stock that would alter the nature
    of the bargain. See id. at 1313–14. The same is true of the defend-
    ants’ misrepresentations about commissions. Although investors
    presumably knew the salespeople were being paid somehow (and
    there was testimony to that effect), a reasonable jury could have
    found that it would decrease the value investors got from the bar-
    gain if their money was going to a salesperson’s pocket in the form
    of commissions, rather than injecting capital for FCF to expand or
    to conduct research and development. See Simon, 
    839 F.2d at
    1468
    USCA11 Case: 17-15003        Date Filed: 10/21/2021   Page: 27 of 53
    17-15003                 Opinion of the Court                    27
    (salesperson falsely telling client she did not work for commissions
    provided support for jury’s finding of intent to defraud).
    And a reasonable jury could have inferred that Wheeler and
    Long made these misrepresentations knowingly, or, at a mini-
    mum, with “reckless indifference to the truth.” 
    Id. at 1470
    . Nota-
    bly, at least some of the defendants’ misrepresentations went be-
    yond the information contained in scripts and press releases. For
    example, a press release from Houlihan touted a connection be-
    tween Sculley and Sanomedics based on a small investment that a
    Sculley-owned company made in Sanomedics. But there were no
    press releases connecting Sculley—or Apple for that matter—to
    FCF. Further, there was evidence that the salespeople filled out
    their own commission sheets to claim their share of each stock sale
    and thus would have known they were paid in commissions.
    The evidence against Long on Count 12 and against
    Wheeler on Count 13 was not overwhelming. But considering the
    evidence in the light most favorable to the government and draw-
    ing all reasonable inferences and credibility choices in the govern-
    ment’s favor, the evidence provided a basis for a reasonable jury to
    conclude that Wheeler and Long schemed to defraud investors.
    b.      Conspiracy Counts
    The district court also granted judgment of acquittal on the
    conspiracy charges against Wheeler and Long, finding that even if
    they made some misrepresentations to investors, they did not
    know about the “overarching fraudulent scheme” perpetrated by
    USCA11 Case: 17-15003        Date Filed: 10/21/2021     Page: 28 of 53
    28                      Opinion of the Court                 17-15003
    Mesa and Sizer to make off with investors’ money. The govern-
    ment argues on appeal that no such knowledge is necessary to
    prove a conspiracy to defraud. Wheeler and Long, defending the
    district court’s ruling, point us to several cases. In United States v.
    Toler, for example, we said that the “government must show an
    ‘interdependence’ among the alleged co-conspirators in order to
    prove that the indicted conspiracy was a single unified conspiracy
    as opposed to a series of smaller, uncoordinated conspiracies.” 
    144 F.3d 1423
    , 1426 (11th Cir. 1998).
    We articulated a similar rule in United States v. Chandler,
    
    376 F.3d 1303
     (11th Cir. 2004). There, the defendants were charged
    in a two-part mail fraud conspiracy. 
    Id. at 1306
    . The “scheme was
    a classic ‘hub-and-spoke’ conspiracy, in which a central core of con-
    spirators recruits separate groups of co-conspirators to carry out
    the various functions of the illegal enterprise.” 
    Id. at 1316
    . Part
    one of the conspiracy involved a single actor embezzling promo-
    tional game stamps from McDonald’s Corporation. 
    Id. at 1306
    . In
    part two of the conspiracy, the embezzler would give the stamps
    to his friends and family—the defendants—to redeem them. 
    Id.
     A
    jury found the defendants guilty of conspiracy to commit mail
    fraud, and the district court denied their motions for judgment of
    acquittal. 
    Id. at 1305
    .
    We reversed the denial of those motions on appeal and va-
    cated the convictions. Although the defendants made misrepresen-
    tations to McDonald’s in redeeming the game stamps, we stressed
    that the indictment alleged a single conspiracy to accomplish two
    USCA11 Case: 17-15003       Date Filed: 10/21/2021     Page: 29 of 53
    17-15003               Opinion of the Court                        29
    objectives: (1) embezzle game stamps and (2) redeem those stamps.
    
    Id. at 1318
    . The defendants agreed to help with the second objec-
    tive, but—fatal to the government’s case—there was no evidence
    that the defendants knew the game stamps had been embezzled.
    
    Id.
     “[W]ithout proof of some connection” between the two objec-
    tives, we said, the government had not proved the single conspir-
    acy it alleged in the indictment. 
    Id.
    In reaching this conclusion, the Chandler court found per-
    suasive the Third Circuit’s decision in United States v. Pearlstein,
    
    576 F.2d 531
     (3d Cir. 1978). There, the defendants were salespeople
    who allegedly used aliases and made misleading statements in their
    sales pitches to customers. The salespeople generally used a script
    developed by higher-ups in the organization but sometimes “added
    a few exaggerations and misstatements” of their own. 
    Id. at 539
    .
    The Third Circuit reversed the defendants’ convictions for conspir-
    acy and substantive mail fraud, reasoning that “although the de-
    fendants might have made fraudulent misrepresentations during
    the course of their individual sales presentations, the jury could not
    reasonably infer that the salesmen knew of the fraudulent purpose
    of the overall [fraudulent] scheme.” 
    Id. at 545
    .
    Wheeler and Long argue that, just like in those cases, the
    government failed to prove that they knew about the “top to bot-
    tom” conspiracy alleged in the indictment, which centered on Sizer
    and Mesa pocketing most of the investors’ money. Long points to
    testimony from Houlihan that Sizer operated “in a secretive man-
    ner,” and argues that even if salespeople made some
    USCA11 Case: 17-15003       Date Filed: 10/21/2021     Page: 30 of 53
    30                     Opinion of the Court                 17-15003
    misrepresentations to induce sales on the front end, they thought
    the companies were legitimate and there was no evidence whatso-
    ever that they knew Sizer and Mesa were looting most of the
    money on the back end.
    The government, on the other hand, points us to Simon,
    where we recognized Pearlstein as “persuasive” but distinguished
    it on the basis that the Simon defendants’ individual misrepresen-
    tations were related to the overall scheme charged in the indict-
    ment. Simon, 
    839 F.2d at 1468
    . Unlike in Pearlstein, the jury’s ver-
    dict in Simon was supported by sufficient evidence that the defend-
    ants knew about the alleged scheme and conspired to execute it.
    
    Id.
    Harmonizing the cases cited by the defendants and the gov-
    ernment, we can derive the following rule: to prove that a defend-
    ant was part of a conspiracy, there must be some evidence that the
    defendant knew the objective of the conspiracy charged in the in-
    dictment and decided to join it. In this case, there was sufficient
    evidence to support a finding that Wheeler and Long knew about
    a conspiracy to sell stock by lying to investors; but there was little
    to no evidence that Wheeler and Long knew about a plot by Sizer
    and Mesa to misappropriate a large percentage of investor funds
    for themselves. So the core issue here is: which of those conspira-
    cies did the indictment charge?
    To make that determination, we start with the language of
    the indictment. Count 11 described the purpose of the FCF con-
    spiracy as follows: for the defendants to “unlawfully enrich
    USCA11 Case: 17-15003      Date Filed: 10/21/2021     Page: 31 of 53
    17-15003               Opinion of the Court                      31
    themselves by misappropriating investor money for their personal
    use and benefit” through misrepresentations and failing to disclose
    material facts, including “the safety and profitability of Fun Cool
    Free stock, and the defendants[’] and their co-conspirators’ exces-
    sive commissions and fees.” Count 1 alleged that the Sanomedics
    conspiracy had a parallel purpose.
    This language is not limited to a higher-level conspiracy by
    Sizer and Mesa to pocket investments; it matches the conspiracies
    that the government attempted to prove at trial in which salespeo-
    ple made misrepresentations and concealed their excessive com-
    missions to sell stock. As a result, the government could prove that
    the defendants knew about the conspiracy charged in the indict-
    ment without proving that they knew what Sizer and Mesa were
    up to.
    There was sufficient evidence on which the jury could have
    concluded that the government did so. One way for the govern-
    ment to show that a defendant knew about and joined the charged
    conspiracy is “through proof of surrounding circumstances such as
    acts committed by the defendant which furthered the purpose of
    the conspiracy.” United States v. Gonzalez, 
    834 F.3d 1206
    , 1215
    (11th Cir. 2016). As detailed above in reference to the substantive
    mail fraud counts, there is evidence that Wheeler and Long made
    claims they knew were false in order to further the purpose of the
    FCF conspiracy. Even if they did not know the extent of the fraud
    perpetrated by Sizer and Mesa, the jury could infer from the de-
    fendants’ actions that they joined the FCF conspiracy charged in
    USCA11 Case: 17-15003           Date Filed: 10/21/2021        Page: 32 of 53
    32                        Opinion of the Court                      17-15003
    the indictment.    Therefore, sufficient evidence supported
    Wheeler’s and Long’s convictions on Count 11.
    The same can be said of Wheeler’s involvement in the Sano-
    medics conspiracy charged in Count 1. While selling Sanomedics
    stock, Wheeler made many of the same misrepresentations he
    made while selling FCF stock. There was also testimony that he
    concealed from investors that they would be indefinitely restricted
    from selling their Sanomedics stock. 5 Therefore, there was evi-
    dence to support a finding that Wheeler knew about and joined the
    conspiracy to defraud investors in Sanomedics.
    Because there was evidence to support the jury’s verdict on
    the substantive and conspiracy counts against Wheeler and Long,
    we reverse the district court’s judgment of acquittal as to those de-
    fendants. We turn next to the convictions of their three co-defend-
    ants.
    2.      Smigrod’s, Sgarro’s, and Topping’s Convic-
    tions
    Counts 1 and 11 of the indictment charged Smigrod with the
    Sanomedics and FCF conspiracies, and Count 9 charged him with
    5 Unbeknownst to the investors, they were not buying shares directly from
    Sanomedics; they were buying Sizer’s personal shares which were restricted.
    Juan Perez Ortega, a salesperson who worked in the Florida phone room, tes-
    tified that Wheeler knew about the stock restriction but did not disclose it to
    investors, which provides support for the inference that Wheeler knew about
    the conspiracy to defraud investors.
    USCA11 Case: 17-15003          Date Filed: 10/21/2021      Page: 33 of 53
    17-15003                   Opinion of the Court                         33
    substantive mail fraud based on a July 2014 sale of Sanomedics
    stock to an investor identified as D.M. Sgarro was charged in the
    Sanomedics conspiracy in Count 1 and for substantive mail fraud
    involving investors identified as G.S. and A.O. in Counts 2 and 10. 6
    Notably, she left Sanomedics in 2011; but on her way out, she re-
    ferred clients—including G.S. and A.O—to Sizer, who sold them
    additional shares. Finally, the indictment charged Topping in both
    the Sanomedics and FCF conspiracies and in a slew of substantive
    counts. Altogether, he was charged in Counts 1, 2, 3, 6, 7, 8, 9, 10,
    11, and 13. We first consider the defendants’ substantive convic-
    tions and then turn to the conspiracy convictions.
    a.      Substantive Counts
    Beginning with Smigrod, he says that he lacked any intent
    to defraud investors. First, he argues that his misrepresentations
    were minor and did not go to the nature of the bargain. Second,
    he argues that the government did not prove that he knew the
    statements he made were inaccurate. We find, however, that alt-
    hough the evidence against Smigrod was not overwhelming, it was
    sufficient to support the jury’s guilty verdict. Like Wheeler and
    Long, Smigrod hid from investors that he made commissions as
    high as twenty percent. There was also testimony from one of Smi-
    grod’s co-workers that he knew the Sanomedics stock he sold was
    restricted, yet he failed to disclose the restriction to investors.
    6 Sgarro was also charged and convicted on Count 1, but she does not appear
    to challenge the sufficiency of the evidence supporting that conviction.
    USCA11 Case: 17-15003       Date Filed: 10/21/2021     Page: 34 of 53
    34                     Opinion of the Court                 17-15003
    Based on these facts, it is at least a “reasonable construction of the
    evidence” that Smigrod intended to defraud victims by deceiving
    them about the value of Sanomedics stock. Capers, 708 F.3d at
    1297.
    As to Sgarro, she argues that she did not participate in the
    stock sales to G.S. and A.O. that are charged in Counts 2 and 10.
    Regarding Count 2, Sgarro first sold stock to G.S. in early 2010.
    Later, in 2011, G.S. bought additional shares from Sizer which form
    the basis for Count 2. Sgarro argues that she cannot be liable for
    the 2011 transaction because she was no longer in the organization
    and did not participate in that sale. She also stresses that she tried
    to dissuade G.S. from buying more Sanomedics stock, pointing to
    testimony that she told him he already had enough.
    We disagree. As the government explains, there was suffi-
    cient evidence that Sgarro aided and abetted the transaction in two
    ways, even if she did not personally participate. First, she provided
    Sizer with G.S.’s contact information, and there was sufficient evi-
    dence to allow the jury to infer that she did so knowing Sizer would
    defraud G.S. Second, she told G.S. that if he reported her to the
    authorities, she would deny everything. We have held that dis-
    couraging a victim from going to the authorities amounts to aiding
    and abetting the crime. United States v. Hansen, 
    262 F.3d 1217
    ,
    1236 (11th Cir. 2001) (per curiam).
    On Count 10, Sgarro was convicted of defrauding investor
    A.O. That investor testified that Sgarro called her and presented
    an investment opportunity in Sanomedics stock. Sgarro told A.O.
    USCA11 Case: 17-15003       Date Filed: 10/21/2021     Page: 35 of 53
    17-15003               Opinion of the Court                        35
    that Sanomedics was going to secure an endorsement deal with Ce-
    sar Millan, a popular television personality known as the “dog
    whisperer,” and that Sanomedics had recently received two large
    orders that would generate over $10,000,000 in revenue. She also
    said that Sanomedics was in clinical trials at Yale Medical School
    and Miami Children’s Hospital. None of these representations
    were true.
    A.O. invested $100,000 in Sanomedics stock with Sgarro.
    When Sgarro left the operation in early 2012, she referred A.O. to
    Sizer, and A.O. continued to invest in Sanomedics with Sizer in the
    summer of 2012. Although Sgarro did not personally orchestrate
    the 2012 transaction on which Count 10 is based, A.O. testified that
    she executed that transaction based on “the viability of the corpo-
    ration as told to me by Anita [Sgarro] and Craig Sizer.” Specifically,
    A.O. testified that she relied on Sgarro’s representations that Sano-
    medics was in clinical trials, that it had high-profile endorsements,
    and that it was making millions in profit. While Sgarro argues that
    people typically make stock purchase decisions based on currently
    available information, we cannot substitute our judgment for that
    of the jury, who apparently credited A.O.’s testimony about her
    reliance on Sgarro’s earlier representations. Therefore, we reject
    Sgarro’s sufficiency-of-the-evidence argument.
    Finally, Topping’s substantive convictions. Topping’s pri-
    mary argument is that there was no basis on which the jury could
    have convicted him of Counts 2, 6, and 10 because there was no
    evidence that he was involved in those transactions or that he had
    USCA11 Case: 17-15003         Date Filed: 10/21/2021       Page: 36 of 53
    36                       Opinion of the Court                    17-15003
    any contact with the victims named in those counts. We rejected
    a similar argument in United States v. Munoz, 
    430 F.3d 1357
    , 1368–
    69 (11th Cir. 2005). There, we held that a defendant can be guilty
    of mail fraud even if there is no evidence of his personal interaction
    with a particular victim. It is enough to show that the defendant
    “participated in a scheme to defraud all buyers.” Id.; see also
    United States v. Toney, 
    598 F.2d 1349
    , 1355 (5th Cir. 1979) (“It is
    well settled in this circuit that so long as one participant in a fraud-
    ulent scheme causes a use of the mails in execution of the fraud, all
    other knowing participants in the scheme are legally liable for that
    use of the mails.”). 7 Because there was evidence that Topping ac-
    tively participated in the scheme to defraud at the time of the trans-
    actions described in Counts 2, 6, and 10, the government need not
    show that he was personally involved with every victim. See
    Munoz, 
    430 F.3d at
    1368–69.
    More broadly, Topping argues that there was insufficient ev-
    idence to support all of his convictions because the government
    failed to prove that he made misrepresentations that went to the
    nature of the bargain with investors. His argument is that despite
    any misstatements he made, investors got the shares of stock they
    bargained for. He also argues that the government failed to prove
    that he knew any misstatements were false. We are satisfied,
    7 We adopted as binding precedent all decisions of the former Fifth Circuit
    handed down before October 1, 1981. Bonner v. City of Prichard, 
    661 F.2d 1206
    , 1209 (11th Cir. 1981) (en banc).
    USCA11 Case: 17-15003      Date Filed: 10/21/2021    Page: 37 of 53
    17-15003               Opinion of the Court                      37
    however, that the record provides sufficient evidence for a reason-
    able jury to conclude both that Topping’s representations were es-
    sential to the bargain and that Topping knew they were false.
    True, Topping’s use of an alias or his claiming to work for Sano-
    medics would not establish an intent to defraud. But to highlight
    some facts that could establish such intent, there was evidence that
    Topping:
    • said that Sanomedics was profiting millions of dollars (its an-
    nual revenue never actually reached $100,000);
    • represented that FCF was “bring[ing] on John Sculley, for-
    mer Apple CEO, to facilitate success and growth”; he also
    played up connections with other famous people like billion-
    aire investor Phillip Frost and the founder of Papa John’s
    Pizza, whereas, in reality, none of these people were actively
    involved in Sanomedics;
    • created the illusion that he was offering investors a special
    opportunity to buy “institutional shares,” although there
    was nothing special about the shares he was selling; and
    • concealed that he made commissions and that the stock he
    was selling would be restricted.
    And the jury could have drawn the inference that Topping
    made these claims knowing they were false. Many of these state-
    ments went beyond the information in the written press releases,
    and Topping even admitted in an interview with law enforcement
    that he misled investors about how he was compensated. This
    USCA11 Case: 17-15003        Date Filed: 10/21/2021    Page: 38 of 53
    38                       Opinion of the Court               17-15003
    evidence was sufficient to support Topping’s convictions on the
    substantive counts.
    b.      Conspiracy Counts
    Smigrod and Topping also challenge the sufficiency of the
    evidence supporting their convictions on Count 1 (the Sanomedics
    conspiracy) and Count 11 (the FCF conspiracy). Both claim that
    they did not know of the conspiracies to defraud investors—that
    they were unwitting pawns in a conspiracy hatched by Sizer and
    Mesa. But, as with Wheeler and Long, there was sufficient evi-
    dence for the jury to conclude that Smigrod and Topping know-
    ingly joined the conspiracies charged in the indictment. Again,
    Smigrod’s and Topping’s misrepresentations often went beyond
    the information provided to them in scripts or press releases, allow-
    ing the inference that they invented falsehoods to further the pur-
    pose of the conspiracy. See Gonzalez, 834 F.3d at 1215. And there
    was additional evidence that the defendants were not innocent
    dupes. For example, Mesa instructed salespeople not to pitch
    members of law enforcement, financial professionals, or lawyers,
    and there were even signs posted to that effect in the Florida phone
    room. Finding that sufficient evidence supported the jury’s ver-
    dicts, we affirm the district court’s denial of the motions for judg-
    ment of acquittal.
    B.     Prosecutorial Misconduct
    Even if the defendants are not entitled to a judgment of ac-
    quittal, we still must consider their arguments that they are entitled
    USCA11 Case: 17-15003        Date Filed: 10/21/2021      Page: 39 of 53
    17-15003                Opinion of the Court                         39
    to a new trial. See United States v. Thomas, 
    86 F.3d 647
    , 651 (7th
    Cir. 1996) (finding that evidence was sufficient to convict the de-
    fendants but reversing the convictions and remanding for a new
    trial because improper jury instructions did not amount to harm-
    less error). We begin with the defendants’ argument that the dis-
    trict court should have ordered a new trial because of prosecutorial
    misconduct.
    Prosecutors have a special role in our justice system. “The
    United States Attorney is the representative not of an ordinary
    party to a controversy, but of a sovereignty . . . whose interest,
    therefore, in a criminal prosecution is not that it shall win a case,
    but that justice shall be done.” Berger v. United States, 
    295 U.S. 78
    ,
    88 (1935). So we must carefully review accusations that a criminal
    conviction is marred by prosecutorial misconduct. Our two-step
    framework for assessing allegations of prosecutorial misconduct is
    well-established: “(1) the remarks must be improper, and (2) the
    remarks must prejudicially affect the substantial rights of the de-
    fendant.” See Eckhardt, 466 F.3d at 947.
    We begin with the first prong: whether the prosecutor made
    improper remarks. While a prosecutor can attack a defense theory
    during closing argument, see United States v. Sosa, 
    777 F.3d 1279
    ,
    1297 (11th Cir. 2015), he is not to misstate the law or tell the jury it
    can ignore the law, see Spivey v. Head, 
    207 F.3d 1263
    , 1275 (11th
    Cir. 2000).
    Whether the prosecutor acted improperly here is a close
    question. The theory-of-defense instruction explained, crucially,
    USCA11 Case: 17-15003       Date Filed: 10/21/2021     Page: 40 of 53
    40                     Opinion of the Court                 17-15003
    that there is a “difference between deceiving and defrauding.” It is
    cause for concern that the prosecutor told the jury that this instruc-
    tion was “not the law.” When considered in context, however, we
    cannot say that the prosecutor’s remarks were improper. The dis-
    trict judge repeatedly emphasized to the lawyers that the theory-
    of-defense instruction was not an instruction about the law and did
    not affect the legal elements for mail and wire fraud. When one
    prosecutor raised concerns that the proposed wording of the the-
    ory of defense instruction was “describing it as the law,” the judge
    responded “[n]o, I say up front it is the defendants’ theory of de-
    fense.” When one defense lawyer proposed a sentence beginning
    “the law makes clear,” the judge responded, “saying the law makes
    clear is not probably the best way to caption a paragraph for a the-
    ory of defense.” When another complained about removing sev-
    eral paragraphs from his proposed instruction, the judge empha-
    sized: “I think you have a right to say what your theory is, not a
    right to, through the theory, back door the instruction” on intent
    to harm. She went on to admonish them: “I don’t want the jury to
    think this is part of the actual substantive instruction.” As she ed-
    ited the instruction, she underscored that she was “trying to take
    out anything that says the law requires or this requires.”
    Moreover, that the prosecutor confusingly characterized the
    theory-of-defense instruction does not necessarily mean that the
    prosecutor acted improperly. One of our precedents, in particular,
    underscores this point. In United States v. Chirinos, 
    112 F.3d 1089
    (11th Cir. 1997), we held that a prosecutor did not commit
    USCA11 Case: 17-15003       Date Filed: 10/21/2021    Page: 41 of 53
    17-15003               Opinion of the Court                       41
    misconduct when he discussed during his opening statement
    “highly prejudicial evidence” of a defendant’s unrelated criminal
    activity. 
    Id. at 1097
    . Even though the district court later excluded
    that evidence, we did not fault the prosecutor for making the re-
    marks. These remarks were “not improper,” we reasoned, because
    the prosecutor thought “the district court would admit the 404(b)
    evidence at the time he made such statements.” 
    Id. at 1098
    . The
    same is true here: based on the district judge’s statements during
    the charge conference, the prosecutor could reasonably have be-
    lieved his remarks were proper. And the district judge even agreed
    the remarks were proper in overruling the defendants’ first objec-
    tion. Accordingly, we find that the prosecutor did not make im-
    proper remarks and thus we need not consider whether the com-
    ments were prejudicial. See Eckhardt, 466 F.3d at 947.
    C.     Evidentiary Issues and Right to Unanimous Verdict
    We turn next to the defendants’ contentions that the district
    court abused its discretion by allowing inadmissible evidence at
    trial in several instances, and Topping’s contention that the jury
    instructions deprived him of his right to a unanimous jury verdict.
    Beginning with the evidentiary issues, Topping argues, first,
    that the district court erred by allowing Rubens, a government in-
    formant who secretly recorded the Florida phone room, to offer
    improper lay opinions. Rubens testified that the Florida phone
    room was a “boiler room”—a term used to describe a salesroom
    that uses high-pressure selling tactics. Rubens also identified and
    defined the roles that were commonly assigned to salespeople in
    USCA11 Case: 17-15003        Date Filed: 10/21/2021     Page: 42 of 53
    42                      Opinion of the Court                 17-15003
    boiler-room operations. For example, front-end salespeople who
    cold-called potential investors were called “fronters.” After a
    fronter made a sale, a second salesperson, known as a “loader,”
    would follow up and pitch the investor on making additional pur-
    chases.
    The district court did not abuse its discretion in allowing Ru-
    bens’s testimony. Under the rules of evidence, a lay witness can
    give opinion testimony that is: “(a) rationally based on the witness’s
    perception; (b) helpful to clearly understanding the witness’s testi-
    mony or to determining a fact in issue; and (c) not based on scien-
    tific, technical, or other specialized knowledge within the scope
    of Rule 702.” Fed. R. Evid. 701. Rubens’s testimony met these cri-
    teria. He was able to assist the jury by explaining, for example,
    some of the jargon used in conversations between Rubens and
    Mesa. And Rubens was able to draw on his own experience in the
    field to explain to the jury how “boiler rooms” typically functioned,
    as well as the various roles that salespeople played in such opera-
    tions. Because Rubens’s testimony was based on his own percep-
    tions, was helpful to the jury, and was not based on scientific or
    technical knowledge, Rubens did not give an improper lay opinion.
    See Fed. R. Evid. 701.
    Next, Topping argues that the district court allowed inflam-
    matory victim impact testimony. To prove mail fraud, the Gov-
    ernment does not need to prove that the victims suffered an actual
    loss. As a result, Topping argues, the court erred by allowed vic-
    tims to testify about the amount of their loss, and—even worse—
    USCA11 Case: 17-15003        Date Filed: 10/21/2021     Page: 43 of 53
    17-15003                Opinion of the Court                        43
    the impact of that loss, such as victims struggling to pay for their
    child’s surgery or for their grandson’s college. We find, however,
    that the testimony at issue came into evidence to illustrate that the
    defendants would learn about the victims’ personal lives, gain their
    trust, and then use that trust to exploit them. And, in any event,
    the district court gave a limiting instruction, stressing that the jury
    could not “be influenced in any way by either sympathy for or prej-
    udice against the defendant.” As a result, we do not find reversible
    error based on the victims’ testimony.
    Finally, both Topping and Sgarro argue that the district
    court abused its discretion when it allowed the government to im-
    properly impeach defense witness Gary Miller on a collateral mat-
    ter in order to label Sgarro as a drug dealer. Although this is a close
    question, we cannot say that the district court abused its discretion
    in this regard. “[E]xtrinsic evidence is not admissible to prove spe-
    cific instances of a witness’s conduct in order to attack or support
    the witness’s character for truthfulness.” Fed. R. Evid. 608(b).
    However, extrinsic evidence can be “offered for other grounds of
    impeachment (such as contradiction, prior inconsistent statement,
    bias and mental capacity).” Fed. R. Evid. 608(b), advisory commit-
    tee’s notes to 2003 amendments. Typically, extrinsic evidence of a
    prior inconsistent statement is admissible because it is relevant to
    the witness’s character for untruthfulness, but the witness must be
    given “an opportunity to explain or deny the statement,” and an
    adverse party must be given “an opportunity to examine the wit-
    ness about it.” Fed. R. Evid. 613(b).
    USCA11 Case: 17-15003        Date Filed: 10/21/2021     Page: 44 of 53
    44                      Opinion of the Court                 17-15003
    We have held, however, that the Government cannot begin
    a line of questioning that is irrelevant, only to impeach the witness
    with a prior inconsistent statement. In United States v. Reed, for
    example, a defendant who stood trial for embezzlement and ob-
    struction of the mail testified in his own defense. 
    700 F.2d 638
    , 643
    (11th Cir. 1983). During cross-examination, the prosecution asked
    whether the defendant had admitted to investigators that he
    smokes marijuana. 
    Id.
     After the defendant denied making the
    statement, the prosecution called the investigator in rebuttal to im-
    peach the defendant with a prior inconsistent statement. 
    Id. at 644
    .
    We reversed the defendant’s conviction in Reed because ir-
    relevant references to the defendant’s drug use amounted to prej-
    udicial error. 
    Id.
     at 646–47. We explained:
    Though the defendant’s answers may have indeed
    been false, the issue before us is the propriety of the
    prosecutor’s questions and the introduction of the de-
    fendant’s possession and use of marijuana as an issue
    in the trial; the prosecutor can hardly contend that his
    question was intended to highlight a prior incon-
    sistent statement when no inconsistent statement had
    been uttered until after the prosecutor broached the
    marijuana issue.
    
    Id. at 644
    . Sgarro urges that the same result should follow
    here.
    Although this case bears some similarity to Reed, the gov-
    ernment points out a key distinction. In this case—unlike in
    USCA11 Case: 17-15003        Date Filed: 10/21/2021      Page: 45 of 53
    17-15003                Opinion of the Court                         45
    Reed—the prosecution’s initial line of questioning was relevant to
    establish that the witness might be biased. If Miller had an illicit
    drug-dealing business with Sgarro, he might have been likely to
    slant his testimony in her favor.
    In that regard, this case is analogous to United States v. Abel,
    
    469 U.S. 45
     (1984). There, the prosecution pursued a line of ques-
    tioning related to the defense witness and the defendant having
    membership in a secret prison gang. When the defense witness
    denied that association, the prosecution used extrinsic evidence to
    impeach him. The prosecution’s line of questioning was relevant,
    the Court held, to establishing that the defense witness might be
    biased in the defendant’s favor based on shared membership in the
    gang.
    Here, the district court, in its discretion, found that the pros-
    ecution’s inquiry was relevant to illustrate Miller’s bias, rather than
    simply to portray Sgarro as a drug dealer. In light of the Supreme
    Court’s decision in Abel, the district court did not abuse its discre-
    tion in admitting this evidence.
    Having addressed the defendants’ contentions that the dis-
    trict court’s evidentiary rulings were an abuse of discretion, we
    turn to Topping’s argument that the district court erred in its jury
    instructions. Topping argues that the jury instructions were erro-
    neous because they charged “overlapping,” rather than separate
    conspiracies. Because the conspiracies were presented to the jury
    as overlapping, the argument goes, the jury could convict as long
    as each juror believed Topping was guilty as to one conspiracy.
    USCA11 Case: 17-15003       Date Filed: 10/21/2021     Page: 46 of 53
    46                     Opinion of the Court                 17-15003
    Such a result would run afoul of the principle that a jury cannot
    convict without a unanimous finding of guilt as to a single conspir-
    acy. See Richardson v. United States, 
    526 U.S. 813
     (1999).
    After reviewing the jury instructions at issue here, we reject
    Topping’s argument. As the government points out, the instruc-
    tions conveyed that conspiracies were “overlapping” in the sense
    that the participants overlapped. The instructions read: “As you
    can see, the Indictment alleges two overlapping conspiracies.
    Please note that Defendant Long is not charged in Count 1, which
    is the first conspiracy count, and Defendant Sgarro is not charged
    in Count 11, the other conspiracy count.”
    Moreover, the jury instructions directed the jury not to
    lump together separate charges. In relevant part, the instructions
    provided that:
    Each count of the indictment charges a separate
    crime against one or more of the defendants. You
    must consider each crime and the evidence relating
    to it separately. And you must consider the case of
    each defendant separately and individually. If you
    find a defendant guilty of one crime, that must not
    affect your verdict for any other crime or any other
    defendant.
    USCA11 Case: 17-15003      Date Filed: 10/21/2021     Page: 47 of 53
    17-15003               Opinion of the Court                      47
    Because the jury instructions explained that each crime was
    to be considered separately, they did not deprive Topping (or any
    defendant) of the right to a unanimous jury verdict.
    D.     Sentencing
    Having determined that the defendants’ convictions should
    be affirmed, we turn to a number of sentencing-related issues
    raised by the defendants. Sgarro and Topping each argue that the
    court improperly applied sentencing enhancements. As to Sgarro,
    she argues that the court should not have applied the sophisticated-
    means enhancement because there was nothing intricate or com-
    plex about the conspiracy. The Sentencing Guidelines include a
    two-level enhancement for an offense that involved sophisticated
    means. U.S.S.G. § 2B1.1(b)(10)(C). An offense is sophisticated if it
    involves “especially complex or especially intricate offense conduct
    pertaining to the execution or concealment of an offense.” Id. §
    2B1.1, application notes 9(B). We have held that a sophisticated-
    means enhancement can be applied when some—but not all—as-
    pects of a scheme are sophisticated. See United States v. Ghertler,
    
    605 F.3d 1256
    , 1267–68 (11th Cir. 2010). Here, parts of the Sano-
    medics scheme were sophisticated. For example, the evidence
    showed that Sgarro paid her salespeople through a company not
    associated with Sanomedics: Ladybug Ventures. Sgarro also had
    her salespeople research target investors’ financial assets, some-
    times rewarding salespeople for finding investors with IRA ac-
    counts. Based on this evidence, the district court’s application of
    USCA11 Case: 17-15003       Date Filed: 10/21/2021    Page: 48 of 53
    48                     Opinion of the Court                17-15003
    the sophisticated-means enhancement was not clearly erroneous.
    See Monzo, 852 F.3d at 1348.
    As to Topping, he argues that the court had no basis to im-
    pose a managerial-role sentencing enhancement. The Sentencing
    Guidelines include a three-level enhancement for acting as “a man-
    ager or supervisor.” Id. § 3B1.1(b). Even if a defendant could not
    force others to engage in criminal conduct, he can still be sentenced
    as a manager or supervisor if he hires and trains others to partici-
    pate in the criminal operation. United States v. Matthews, 
    168 F.3d 1234
    , 1249–50 (11th Cir. 1999). Here, there was evidence that Top-
    ping led sales meetings and trained new salespeople, and he was
    the only one in the Florida operation with an office, other than
    Mesa (the director of the Florida operation). Accordingly, the dis-
    trict court did not clearly err in applying a sentencing enhancement
    for Topping’s managerial role. See Monzo, 852 F.3d at 1348.
    Sgarro and Topping also contest the amount of loss for
    which they were held responsible. Sgarro argues that, because she
    worked only in the California operation and because she left the
    conspiracy in early 2011, the district court erred in attributing to
    her a $22.3 million loss, which represented victims’ losses from
    both the Florida and California operations. Similarly, Topping ar-
    gues that the evidence does not support the $22.5 million loss at-
    tributed to him. He argues that the district court did not even at-
    tempt to make an individualized finding as to the extent of his cul-
    pability.
    USCA11 Case: 17-15003       Date Filed: 10/21/2021     Page: 49 of 53
    17-15003               Opinion of the Court                        49
    We do not find these arguments persuasive. The district
    court need only make a reasonable estimate of the loss. United
    States v. Campbell, 
    765 F.3d 1291
    , 1301 (11th Cir. 2014). And as the
    government notes, a defendant is liable for the total loss amount of
    the conspiracy when she is actively involved in furthering the con-
    spiracy’s overall objective. United States v. McCrimmon, 
    362 F.3d 725
    , 731–33 (11th Cir. 2004). Here, there was evidence that Sgarro
    knew the overall scheme and actively participated in furthering it.
    The following two facts, among others, support that conclusion.
    First, Sgarro was in close contact with Mesa and the Miami phone
    room, often exchanging information about sales and investor fi-
    nancial information. Second, she introduced investors to Sizer,
    who oversaw both the Florida and California operations. There-
    fore, the district court did not err in holding Sgarro accountable for
    losses caused by the Florida and California phone rooms. Likewise,
    because Topping actively participated in the scheme to defraud, he
    is liable for the amount of loss to all the victims who were de-
    frauded through the scheme. See 
    id.
    In conclusion, we affirm the convictions and sentences of
    Smigrod, Sgarro, and Topping. As to Wheeler and Long, we re-
    verse their judgments of acquittal and remand to the district court
    for proceedings consistent with this opinion.
    AFFIRMED IN PART; VACATED AND REMANDED IN
    PART.
    USCA11 Case: 17-15003        Date Filed: 10/21/2021     Page: 50 of 53
    50                      Opinion of the Court                 17-15003
    WILSON, Circuit Judge, joined by LAGOA, Circuit Judge, con-
    curring:
    I concur in the court’s judgment. I write separately to ex-
    press my concern with the conduct of the prosecution during clos-
    ing argument and throughout trial in this case.
    To address what transpired in closing argument, the theory-
    of-defense instruction drafted by the district court was quite unu-
    sual. It framed statements of law as the defendants’ “con-
    ten[tions].” And, as the court’s opinion explains, the district judge
    told the parties during a charge conference that the instruction in
    question was “a theory, not a legal instruction.” Perhaps the pros-
    ecutor still should have known better than to argue that an instruc-
    tion from the court was “not the law.” It is well-settled that district
    courts can instruct the jury on a theory of defense only if it is based
    on a correct statement of law. See Chirinos, 112 F.3d at 1101 (“The
    district court should instruct the jury on the defendant’s defense
    theory if the theory has a foundation in evidence and legal sup-
    port.”). As the district court correctly told the parties, the prosecu-
    tion should have focused its argument on whether the theory of
    defense “match[ed] up to the evidence”—not whether the instruc-
    tion accurately stated the law. Still, the district court’s commen-
    tary on the instruction, as well as the manner in which the instruc-
    tion was phrased, gave mixed messages. With that in mind, I can-
    not say that the prosecution’s closing argument amounted to prej-
    udicial misconduct under our precedents.
    USCA11 Case: 17-15003        Date Filed: 10/21/2021     Page: 51 of 53
    17-15003                Opinion of the Court                        51
    More broadly, however, I feel compelled to address the
    prosecution’s conduct and the tactics it employed throughout the
    trial. The prosecution fell short of the high level of professionalism
    that we expect prosecutors to embody, even if their actions did not
    rise to the level of misconduct. An unfortunate but notable feature
    of this trial was that the district court exerted considerable time and
    energy corralling the prosecution’s often wayward tactics. Starting
    in voir dire and continuing through the testimony of multiple wit-
    nesses, the prosecution frequently appeared to ignore the court’s
    rulings when it disagreed with them, eliciting remarks from the
    court including:
    • “Counsel, you know that’s improper.”
    • “[W]hy would you go there?”
    • “We went over this. . . . I may be wrong, but I ruled. Let’s
    go.”
    • “We’ve had this conversation through other witnesses.
    Counsel, move on.”
    • “We did this yesterday. I’m not revisiting. Anybody [who]
    wants to go back to the transcript, can.”
    • “I don’t know . . . how many other languages to speak to
    you. . . . I said how to proceed. Proceed that way.”
    • “What you have to do is to remember we have had some
    rules in this trial and somehow they seem to have been for-
    gotten.”
    USCA11 Case: 17-15003       Date Filed: 10/21/2021     Page: 52 of 53
    52                     Opinion of the Court                 17-15003
    The court’s admonitions, it seems, had little effect. After a
    particularly volatile exchange between a prosecutor and defense
    witness Miller, in which the prosecutor admitted that he lost his
    composure, the district court warned that he was “close” to causing
    a mistrial. Afterward, outside the presence of the jury, she admon-
    ished the prosecutor, telling him: “You’re better than having to go
    to the lowest part of your anger in order to examine this witness. .
    . . [Y]ou’re an experienced cross-examiner. You didn’t have to do
    that. I would have expected that of someone of less experience
    than you.” The district court lamented that things “got very messy
    and uncontrolled.”
    Certainly, some missteps are to be expected over the course
    of a long, contentious trial. And to be sure, the court felt compelled
    on occasion to correct the conduct of the defense lawyers, as well
    as the prosecution. A review of the transcript, however, leaves no
    doubt that the prosecution’s professionalism and demeanor was
    particularly lacking. The district court, after having had time to
    reflect, singled out one prosecutor’s conduct at trial as “disturbing”
    and recounted multiple instances in which he came “so close” to
    causing a mistrial.
    Prosecutors must hold themselves to a higher standard. As
    they carry out their duty to seek justice, they must strive to conduct
    themselves with the utmost professionalism, “even when the ethi-
    cal rules and constitutional requirements do not bar the behavior”
    in question. See Charles R. Wilson, “That Justice Shall Be Done”—
    Constitutional Requirements, Ethical Rules, and the Professional
    USCA11 Case: 17-15003      Date Filed: 10/21/2021     Page: 53 of 53
    17-15003               Opinion of the Court                      53
    Ideal of Federal Prosecution, 
    36 N. Ill. U. L. Rev. 111
    , 135 (2015).
    Although the prosecution’s actions in this case did not amount to
    misconduct under our precedents and on the unique facts pre-
    sented here, I expect far better from prosecutors in future cases.