Jesse Guardado v. Secretary, Florida Department of Corrections

112 F.4th 958

• Court: Court of Appeals for the Eleventh Circuit
• Decided: August 12, 2024
• Docket: 22-10957
• Status: Published

Opinion

USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 1 of 76

[PUBLISH] In the United States Court of Appeals For the Eleventh Circuit

____________________

No. 22-10957 ____________________

JESSE GUARDADO, Petitioner-Appellant, versus SECRETARY, FLORIDA DEPARTMENT OF CORRECTIONS,

Respondent-Appellee.

Appeal from the United States District Court for the Northern District of Florida D.C. Docket No. 4:15-cv-00256-RH ____________________ USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 2 of 76

2 Opinion of the Court 22-10957

Before WILLIAM PRYOR, Chief Judge, and JILL PRYOR and LUCK, Cir- cuit Judges. LUCK, Circuit Judge: On September 21, 2004, Jesse Guardado confessed to the Walton County Sheriff’s Office that, eight days earlier, he had robbed and brutally murdered seventy-five-year-old Jackie Malone in her home. After he pleaded guilty in the state trial court without the benefit of a plea agreement or the aid of counsel, counsel was appointed to represent Guardado for the penalty phase. A penalty phase jury unanimously recommended that Guardado receive the death penalty, and the state trial court sentenced him to death. Guardado appeals the denial of his petition for a writ of ha- beas corpus under 28 U.S.C. section 2254. He argues that the Flor- ida Supreme Court unreasonably applied Strickland v. Washington, 466 U.S. 668 (1984), in denying his claims that his trial counsel were constitutionally ineffective by: (1) failing to adequately investigate and present mitigating evidence for the penalty phase; and (2) fail- ing to challenge for cause or peremptorily strike Jurors Pamela Pennington, Earl Hall, and William Cornelius. After careful review of the briefs and the record, and with the benefit of oral argument, we affirm. USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 3 of 76

22-10957 Opinion of the Court 3

I. FACTUAL BACKGROUND AND PROCEDURAL HISTORY A. The Murder James Brown knelt in the supermarket aisle of a small-town Winn-Dixie. The evening was busy at the store, and Mr. Brown had been hard at work restocking groceries. Mr. Brown stooped to reach a shelf near the floor and suddenly felt the cold steel of a knife against his throat. “[G]ive [me your] wallet,” demanded the man over Mr. Brown’s shoulder. But Mr. Brown didn’t comply. He hollered for help and grabbed for the knife. The man pulled the knife back, slicing two of Mr. Brown’s fingers, and ran from the store before Mr. Brown got a good look at his face. The knifeman was Guardado. After Mr. Brown hollered for help, Guardado rushed to the Winn-Dixie parking lot and fled the scene. He’d been out of prison on conditional release for about a year and a half after spending most of his adult life behind bars. Convicted for armed robbery in 1984. Robbery with a deadly weapon in 1990. Robbery with a weapon again in 1991. And now, on September 13, 2004, he’d attempted to rob a Winn-Dixie em- ployee—all because he needed a crack cocaine fix. And he needed it bad. His crack binge was two weeks strong and counting, but he had to have more money to keep it going. As darkness fell in the late summer sky over DeFuniak Springs, Guardado’s thoughts turned to Ms. Malone. Guardado had known Ms. Malone since meeting her the year before, shortly USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 4 of 76

4 Opinion of the Court 22-10957

after his release from prison, and she’d been good to him despite his past. She helped him find a place to stay and rented him one of her properties. She provided money and helped him find his cur- rent job at the wastewater treatment plant in Niceville. She had even told him where he could find the spare key to her home and let him crash there overnight between rentals. But none of that mattered to Guardado now. What did mat- ter was that Ms. Malone lived in a secluded area. That she trusted him enough to open her home to him if he showed up in the mid- dle of the night. That she had money. And that he knew where to find it. To get the money, all he had to do was kill her. With a plan in place, Guardado swung by his truck and got a kitchen knife. He left home around ten o’clock and drove toward Ms. Malone’s house, armed with the kitchen knife and a breaker bar. Ms. Malone was asleep in her bed when Guardado arrived. Guardado appeared at the front door with the weapons tucked in the small of his back, then knocked and knocked until Ms. Malone woke up and answered. He identified himself through the door, and Ms. Malone opened the door in her nightgown and greeted him. Guardado told her he needed to use her phone, so Ms. Malone turned away to allow him inside. Then Guardado hit Ms. Malone over the head with the breaker bar. When that didn’t kill her, he struck her again and again and again. When that didn’t kill her, he grabbed his knife and stabbed her through the chest. Five times. When that didn’t kill her, he slashed her throat. Twice. USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 5 of 76

22-10957 Opinion of the Court 5

That killed her. Guardado got up, went to Ms. Malone’s bedroom, and rooted through her valuables. He took her jewelry box, briefcase, purse, checkbook, and cell phone, as well as eighty dollars in cash. With enough in hand to guarantee his next high, he left Ms. Malone’s broken body on the floor behind her couch and drove off into the night. B. Guardado’s Confession and Guilty Plea Two days later, Ms. Malone’s brother found her dead body and called the police. The Walton County Sheriff’s Office opened an investigation and quickly homed in on Guardado as a suspect. On September 21, the Niceville Police Department notified investigators in the Walton County Sheriff’s Office that Guardado wanted to speak with them. A meeting was arranged between Guardado and the Walton County investigators near Guardado’s broken-down truck in a Walmart parking lot. Once inside the in- vestigators’ car, Guardado blurted out that “that lady” didn’t de- serve what he’d done to her. The investigators suggested that he pipe down until he had a lawyer. After meeting with an attorney who advised Guardado against talking to the investigators, Guardado still wanted to speak with them. He sat down with three of them—one of whom ad- vised Guardado of his rights under Miranda v. Arizona, 384 U.S. 436 (1966)—and confessed to his crimes. USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 6 of 76

6 Opinion of the Court 22-10957

Guardado told the investigators that after he had killed Ms. Malone, he burned his clothes, Ms. Malone’s jewelry box and purse, and the knife he used in the murder. He discarded the breaker bar, the charred knife, the checkbook, the briefcase, and the jewelry—which he thought would be worthless to sell. Then he headed for his shift at the plant, where he worked for a little while before going to a local convenience store, cashing four of Ms. Malone’s checks, and buying some cigarettes. He contacted his drug dealers in DeFuniak Springs to buy more crack cocaine and continued binging crack before returning to work. Law enforcement later recovered the items Guardado dis- carded and the charred remains of the ones he burned, including the knife, the briefcase, and the breaker bar. On October 19, 2004, having waived his right to counsel, Guardado pleaded guilty to 1 murder without the benefit of a plea bargain. Guardado planned to represent himself for the rest of the proceedings but eventually took the advice of his mother, Patsy Umlauf, to accept counsel for the penalty phase. The state trial court appointed two attorneys to represent Guardado during the penalty phase: John Gontarek and Jason Cobb.

1 Guardado also pleaded guilty to the attempted robbery with a deadly weapon at the Winn-Dixie—for which he was sentenced to forty years’ im- prisonment—and robbery with a weapon at Ms. Malone’s home. USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 7 of 76

22-10957 Opinion of the Court 7

C. Penalty Phase The state trial court set the penalty phase trial to begin on September 13, 2005. 1. Penalty Phase Investigation Mr. Gontarek, a twenty-five-year lawyer from Fort Walton Beach who’d handled between forty and fifty capital litigation cases as a defense attorney, served as lead trial counsel. Mr. Gontarek first met with Guardado at the jail within a couple of weeks of his appointment. He was impressed that Guardado had taken respon- sibility for his crimes and thought that saving the county and Ms. Malone’s family from the ordeal of the trial “would go a long way in mitigation.” He planned to make Guardado’s confession and cooperation the cornerstone of his mitigation strategy. Guardado was uncooperative with Mr. Gontarek. Even so, after informing Guardado that he had a duty to try to save Guardado’s life, Mr. Gontarek had Dr. James Larson, a Pensacola- based forensic psychologist he had worked with in the past (and whose testimony had been used in several successful capital de- fenses), appointed to the case; interviewed any family members or friends who were willing to talk; and talked with any co-workers or others who could provide information about Guardado’s skills in wastewater treatment. Mr. Cobb, a newer lawyer from DeFuniak Springs, joined the defense team after approaching Mr. Gontarek about assisting because he wanted experience working on a capital case. Mr. Cobb had served as a state prosecutor for three years before beginning USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 8 of 76

8 Opinion of the Court 22-10957

his own private criminal defense practice about three years before the murder. As a prosecutor, Mr. Cobb helped prepare a capital case for prosecution, though he didn’t participate in the trial itself. After his appointment as co-counsel, Mr. Cobb met with Mr. Gon- tarek for a briefing on the case and reviewed the discovery packet. Once Mr. Cobb was up to speed, he and Mr. Gontarek met with Guardado at the jail to introduce Mr. Cobb to Guardado and dis- cuss the penalty phase process. Mr. Cobb deferred to Mr. Gontarek as lead counsel in forming a strategy for mitigation and deciding whom should be contacted as potential witnesses. Throughout counsel’s penalty phase investigation, Guardado remained uncooperative and insisted that he wanted to go back to prison, skip straight to sentencing, and even be put to death. He didn’t suggest the names of any family members who Mr. Gontarek and Mr. Cobb should contact. But Mr. Gontarek started the investigation by collecting as much information as he could on Guardado and communicating regularly with Dr. Larson. Despite Guardado’s reluctance, Mr. Gontarek compiled infor- mation about his mother, his stepfather, his uncle, and his em- ployer. Mr. Gontarek dug into Guardado’s jail records, which he preferred to use as mitigating evidence over prison records because prison records might open up prior violent felonies or other unfa- vorable evidence that the state could use. After Mr. Gontarek gathered information about Guardado’s family, he spoke with Mrs. Umlauf on numerous occasions. Most of their conversations centered on how to get Guardado to be USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 9 of 76

22-10957 Opinion of the Court 9

engaged in the process of saving his own life—a goal the two of them shared. Mr. Gontarek repeatedly asked Mrs. Umlauf to tes- tify, but she refused; he did get her to write a letter instead. Mr. Gontarek sent an investigator to talk to Guardado’s em- ployer at the Niceville wastewater treatment plant to get some background information on his water expertise. That effort proved unhelpful because the plant managers suspected Guardado of stealing equipment. Mr. Gontarek personally reached out to Major Rhodene Mathis, a now-retired warden of the Florida Department of Corrections, to discuss Guardado’s behavior during his previous stints in prison. Although Mr. Gontarek determined that Major Mathis didn’t remember anything about Guardado and thus wouldn’t be helpful as a witness, Mr. Gontarek collected records from Major Mathis to share with Dr. Larson. Mr. Gontarek also had Mr. Cobb reach out to Donna Porter, Guardado’s ex-girlfriend, who didn’t want to appear as a witness. Mr. Gontarek and Mr. Cobb chose not to subpoena Ms. Porter be- cause they believed that she wouldn’t be helpful in providing miti- gating evidence. And Mr. Gontarek reached out to Guardado’s other family members, including Guardado’s brother and stepfather, on numer- ous occasions by letter and telephone, especially during the first two months of his investigation. But the problem Mr. Gontarek kept running into was that Guardado had been in prison for so long that his family didn’t have any contact with him. Ultimately, Mr. Gontarek chose not to call any of them as witnesses. USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 10 of 76

10 Opinion of the Court 22-10957

After Dr. Larson was appointed as a mental health expert, he met with Guardado four times to identify possible mitigating cir- cumstances. Dr. Larson conducted two kinds of tests to evaluate Guardado: intelligence tests and personality tests. To evaluate Guardado’s intelligence, Dr. Larson tested Guardado’s IQ with the Wechsler Adult Intelligence Test, which Dr. Larson considered the “gold standard.” Guardado’s score on the test placed him in the sixty-third percentile, meaning he scored as well as or better than sixty-three percent of males in his age group—as Dr. Larson put it, that’s “in the upper part of the average range.” Dr. Larson also had Guardado take an “academic achievement test,” which he de- scribed as “a simple screening test for academic ability” to measure Guardado’s “reading, writing, and arithmetic” capabilities. Guardado scored “in the average range.” Next, to evaluate Guardado’s personality, Dr. Larson admin- istered the Minnesota Multi Phase Personality Inventory-2—“the most used personality test in the world.” Guardado’s results showed “no indications of mental illnesses,” but they did indicate a “slight elevation in [the] depression” scale, a small elevation of the “paranoia scale,” and elevated worry and anxiety. The MMPI-2 also analyzed “various items relating to substance abuse and attitudes toward substance abuse”—like “addiction to substances, [and] atti- tudes, beliefs, and values that support substance abuse”—which were also elevated. Finally, Dr. Larson administered the Hare Psychopathy Checklist to Guardado. Guardado’s score was “[q]uite good” and USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 11 of 76

22-10957 Opinion of the Court 11

indicated that he was not a psychopath. After evaluating Guardado, Dr. Larson prepared a report that summarized his findings and identified several factors Dr. Larson thought a jury would find mit- igating. 2. Voir Dire Guardado’s penalty phase trial began with jury selection. The state trial court broke up voir dire into two phases—general voir dire and individual voir dire. During general voir dire, the state trial court and the attorneys gave instructions and asked questions to all prospective jurors in the courtroom. Then, during individual voir dire, the state trial court examined prospective jurors who in- dicated they might have a potential bias—three at a time—in cham- bers. The general voir dire started with the prosecutor, who asked the prospective jurors if any of them knew the victim. Juror Pen- nington was among ten prospective jurors who said they did. Then, the prosecutor probed into whether the prospective jurors knew any of the witnesses that might be called to testify. Juror Hall said he recognized three law enforcement witnesses: Investigator Rome Garrett, Investigator James Lorenz, and Captain Stan Sun- day. Later, the prosecutor asked the prospective jurors whether they, or anyone they were very close with, had been the victim of a violent crime. Juror Cornelius was among twenty who answered yes because he had family members—a great aunt and a great un- cle—who had been murdered. USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 12 of 76

12 Opinion of the Court 22-10957

Then Mr. Gontarek had his turn during the general voir dire. He asked the prospective jurors to “promise . . . [they would] listen to the evidence and not make any decision about” their sentencing recommendation until they’d seen all the evidence and heard from the defense. In response to Mr. Gontarek’s request, all prospective jurors—including Jurors Pennington, Hall, and Cornelius—indi- cated they’d keep an open mind. The prospective jurors agreed they would “weigh the aggravating factors and mitigating factors according to the law,” the evidence, and the state trial court’s in- structions. And before he ended his general voir dire, Mr. Gontarek confirmed that the prospective jurors would agree to hear all the evidence and make a recommendation based on it by following the state trial court’s instructions. The state trial court then moved to individual voir dire by taking Guardado, Mr. Gontarek, Mr. Cobb, and the prosecutor into chambers to examine three prospective jurors at a time. First, Juror Cornelius had his individual voir dire. Juror Cornelius said that he was “somewhat” in favor of the death penalty, but he empha- sized—four times—that “it should be based on the incident itself ” and the specific “circumstances.” Explaining his response during general voir dire that he had family members who had been the victims of violent crimes, Juror Cornelius recounted that his great aunt and great uncle were killed in a robbery twenty-five years ear- lier, “a long time ago.” He didn’t know many details about it—they were killed when he was “small,” and the bits and pieces he knew were passed on to him by his family. The prosecutor then asked if what Juror Cornelius did know would affect his ability to consider USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 13 of 76

22-10957 Opinion of the Court 13

Guardado’s case. Juror Cornelius assured the prosecutor: “No. That doesn’t have anything to do with that.” During Mr. Gontarek’s questioning, Juror Cornelius agreed there may be circumstances where a life sentence “might be a harsher penalty than the death penalty.” Juror Cornelius had worked construction on three federal prisons and thought prison service at one he’d seen could be “a very severe punishment.” But Juror Cornelius again emphasized that he wasn’t jumping to con- clusions about the case, that his recommendation would depend on what he “hear[d] in the courtroom,” and that, “at th[at] point,” he was “totally neutral on whether or not [he] should recommend death or life” at that time. Next, Juror Pennington moved into the judge’s chambers for her individual voir dire. Juror Pennington said that she was “some- what” in favor of the death penalty and that she didn’t think it should be abolished. When asked about her answer during general voir dire that she knew Ms. Malone, Juror Pennington explained that the two crossed paths a few years before the murder when Ms. Malone brokered a real estate transaction for Juror Penning- ton’s son. In the process of finalizing the deal, the two had either met or spoken on the phone several times over several months. While Juror Pennington liked Ms. Malone and thought Ms. Malone was “a very nice lady,” she hadn’t spoken to Ms. Malone since wrapping up the deal. Juror Pennington assured everyone that she “could be fair” when questioned about whether this encounter with Ms. Malone would affect her ability to fairly evaluate the case. USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 14 of 76

14 Opinion of the Court 22-10957

And when asked a second time whether she could be fair, Juror Pennington said she could. But the prosecutor wanted a final as- surance, asking Juror Pennington if she could promise Guardado “that [she] could be fair and make a fair and legal decision based on whether or not the [s]tate proved that aggravating circumstances exist[ed] which outweigh[ed] the mitigating circumstances.” Her response: “Yes.” Mr. Gontarek picked up the questioning to see if Juror Pen- nington would agree that if she “felt that . . . one mitigating factor outweighed [a number of ] aggravating” factors, she “[w]ould . . . vote[] for life.” She did. Going back to Juror Penning- ton’s relationship with Ms. Malone, Mr. Gontarek asked if she could “set aside” her feelings and make a decision “based on the law and the evidence and not . . . any conversations” she had with Ms. Malone. She offered an unequivocal “[y]es, sir,” and insisted “[i]t was just a business knowing” Ms. Malone. Before the state trial court dismissed her from chambers, Juror Pennington agreed— again—that any feelings she had about Ms. Malone wouldn’t affect her ability to be fair and impartial. Finally, Juror Hall had his individual voir dire. He agreed that if selected to serve on the jury he had to “hold the [s]tate to its burden” before voting to impose the death penalty and that he “[m]ost definitely” would consider if “any mitigating circum- stances weigh[ed] against” imposing it. Addressing his statement during general voir dire that he knew three law enforcement offic- ers involved in the investigation, Juror Hall explained that USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 15 of 76

22-10957 Opinion of the Court 15

Investigator Garrett was a “close friend,” Investigator Lorenz was a former insurance client, and Captain Sunday was the son of a friend who went to school with Juror Hall’s wife. Despite his rela- tionships, Juror Hall maintained that he could fairly weigh any tes- timony given by the three law enforcement officers. Like with Juror Pennington, Mr. Gontarek asked if Juror Hall was aware that “if there’s one mitigating circumstance [he felt] outweigh[ed] any number of aggravating circumstances, it would be [his] duty to impose life,” and he asked if Juror Hall would “fol- low that duty.” Juror Hall agreed he would. Returning to Juror Hall’s relationship with Investigator Garrett, Mr. Gontarek pressed him on whether he’d weigh the officer’s testimony more just be- cause the two were friends. But Juror Hall made clear he would take his friend’s testimony “strictly on its value.” Jurors Pennington, Hall, and Cornelius were ultimately seated on the jury for the penalty phase. By the end of the voir dire, Mr. Gontarek exhausted all but one peremptory challenge. He had also successfully challenged four jurors for cause. 3. Penalty Phase Trial The state called nine witnesses to testify at the penalty phase trial, including Mr. Brown (the Winn-Dixie employee Guardado tried to rob), Ms. Malone’s family, law enforcement officers, and the chief medical examiner, and introduced more than a dozen ex- hibits, including the murder weapons and photographs. The de- fense called two witnesses (Dr. Larson and Guardado) and intro- duced two letters as exhibits (one from the records clerk of the USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 16 of 76

16 Opinion of the Court 22-10957

Walton County Sheriff’s Office and one from Guardado’s mother, Mrs. Umlauf ). a. Dr. Larson’s Testimony Dr. Larson testified about his psychological evaluation of Guardado, including that they met four times, and that Dr. Larson reviewed the arrest reports, depositions, family background, and criminal history associated with Guardado’s case. He began by dis- cussing Guardado’s training in wastewater treatment from his stint in prison, which Dr. Larson explained reflected on Guardado’s in- telligence and was used by Guardado both inside and outside prison. Then, Dr. Larson delved further into Guardado’s mental state and capabilities. Guardado, Dr. Larson explained to the jury, showed no signs of hallucinations, psychotic behavior, mental ill- ness, delusions, thought disorder, or major mood swings. In fact, Dr. Larson found Guardado’s thought processes “well organized, logical, [and] goal oriented”—though Guardado did show signs of depression and remorse. Focusing on evaluating Guardado through the lens of his test scores, Dr. Larson told the jury about Guardado’s sixty-third percentile score on the IQ test and his aver- age range score on the academic achievement test. Comparing Guardado’s IQ with his own observations of Guardado, Dr. Larson found Guardado’s results “consistent with [his] conversations and . . . views” about Guardado—Dr. Larson, in fact, suspected Guardado might place even higher on the IQ test. Turning to Guardado’s MMPI-2 test, Dr. Larson explained that the results USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 17 of 76

22-10957 Opinion of the Court 17

showed Guardado was not mentally ill. He did note Guardado’s elevated depression, paranoia, worry, and anxiety, but those were expected given Guardado had just committed murder, was on trial, and was facing at least a life sentence. And while Guardado’s test results showed “elevated” scores relating to substance abuse, those were also expected “given [Guardado’s] history” of “abus[ing] sub- stances since adolescence” and his prior crimes. Next, Dr. Larson described the Hare Psychopathy Checklist. Before discussing Guardado’s score, Dr. Larson took a moment to describe psychopaths to the jury. A psychopath, he explained, is someone with a “criminal personality”—“[t]hey are basically peo- ple that don’t have a conscience; they are parasitic; mooch off of people; they can frequently move; [and they] tend to have unstable jobs.” Also, “they lack empathy and caring for other people,” “they lack a conscience,” “[t]hey don’t care about stealing or robbing from others,” and they have “no respect for other people’s property or lives.” Guardado, in Dr. Larson’s view, wasn’t a psychopath. To the contrary, he explained that Guardado’s score on the psychopathy test was “[q]uite good”—about the expected score of an average prison inmate, but “not what you would expect from the average inmate on death row or a psychopath.” He explained that Guardado’s test indicated he had empathy, caring, a conscience, and remorse; that “he d[id] not fit the category of the worst of the worst”; and that Guardado was “not the psychopathic type.” Dr. Larson emphasized Guardado “absolutely would not be USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 18 of 76

18 Opinion of the Court 22-10957

considered a psychopath” and “d[id] not suffer any of the tradi- tional mental illnesses, such as bipolar disorder, schizophrenia, ma- jor depressive disorders, [or] major brain damage type of disor- ders.” Based on Guardado’s testing and history of incarceration, Dr. Larson expected Guardado “to make a good adjustment to prison,” “make a contribution,” and be “at low risk” of being a dan- ger if sentenced to life in prison. Evaluating Guardado’s mental state at the time of the mur- der, Dr. Larson opined that Guardado “was under emotional du- ress in this time frame” owing to “economic problems” and “prob- lems adjusting to society.” These problems caused Guardado to “turn[] to his old habits of using cocaine,” adding that his substance abuse habit dated “back to teenage years or early adult years.” And Guardado’s “relapse[]” turned into a “crack cocaine binge for ap- proximately two weeks” before the murder. But despite Guardado’s repeated drug use, Dr. Larson clarified that he did not “consider [Guardado] a drug addict” and that his crack cocaine binge occurred while he was under “considerable stress prior to the” murder. As to Guardado’s mental state following the murder, Dr. Larson described how Guardado expressed “genuine” remorse over Ms. Malone’s murder and called her a “very sweet lady” who “didn’t deserve to die.” Overall, in Dr. Larson’s view, Ms. Malone’s murder was a “drug related incident” motivated solely by his need for crack, mak- ing Guardado a “low risk” for violent behavior generally. Dr. Lar- son also made clear that he evaluated Guardado for two statutory USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 19 of 76

22-10957 Opinion of the Court 19

mitigating circumstances—(1) whether he acted under extreme emotional distress and (2) whether his ability to appreciate the criminality of his conduct or to conform it to the law was substan- tially impaired—but it was his opinion that neither applied. b. Guardado’s Testimony Guardado’s testimony began with questions about his fam- ily background. He was born in Detroit to the “[b]est” mother, his father died when he was young, and his mother married his “won- derful” stepfather, who “did a very fine job of raising four boys.” All three of his brothers turned out well and had respectable ca- reers. He shifted to his own career, describing the process of be- coming certified in wastewater treatment during his incarceration. Once out of prison, he was able to use his expertise and certifica- tion to get a job as the lead wastewater treatment operator at the city of DeFuniak Springs’s water treatment plant. He described how he’d be able to use his skills to assist with wastewater treat- ment and teach other inmates how to get their wastewater man- agement licenses. After talking about his good behavior in prison that earned him conditional release, Guardado pivoted to the challenge of re- adjusting to society following his release. When asked about his return to drugs, Guardado began with how he was hired at the plant. He had been working eighteen-to-twenty-hour days for sev- eral months when, late one Friday night when he thought he had some downtime and had consumed several beers, he was called into work. Because he was the only wastewater treatment USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 20 of 76

20 Opinion of the Court 22-10957

operator in town and was on call around the clock, he headed to work. On his way home, he was pulled over and arrested for DUI. This cost him his job and three months in jail. But he maintained his conditional release because of his demonstrated work ethic and references of support from others—including from the victim, Ms. Malone. He got a job with a construction company that ended badly. He started using crack cocaine around that time. Ms. Malone helped him land his job in Niceville at a wastewater treatment plant, but his crack dependency got worse and his girl- friend moved out. Guardado answered a few questions about his cooperation with the police and expressed remorse over his actions. He testified that the only reason he was in the courtroom that day was because his mother begged him to accept legal counsel, but that he wished they could have “continue[d] without this” and gone straight to sentencing. On cross-examination, Guardado went into greater detail about the period when his crack dependency started controlling his life. He testified that in the two weeks prior to the murder, it was in “every awake [sic] moment that [his] mind was geared to finding and getting crack.” Guardado described what it’s like to come off a high and that one becomes “crazy with need” for more of the drug. Driven by that need, Guardado chose Ms. Malone to rob be- cause she lived in a remote area where his crime “may go unde- tected,” and he went there determined to get the money for drugs—“whatever it took.” USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 21 of 76

22-10957 Opinion of the Court 21

c. Other Evidence Guardado also presented two letters to the jury. In the first letter, the Walton County Sheriff’s Office confirmed Guardado’s record of good behavior while awaiting trial at the Walton County Jail. In the second letter, Guardado’s mother, Mrs. Umlauf, wrote about Guardado’s descent into drugs as a young man because he “chose the wrong friends” and noted how his incarceration “changed him and played a major role in the person he is today.” She described his optimism for his new life once he got out of prison—he “came out with the best of intentions”—and his re- sponsibility in securing and performing a new job in wastewater treatment. She added that Guardado “handl[ed] his job well.” But despite his optimism, Guardado faced problems reintegrating into society and turned to drugs as a way to cope. This led to Guardado’s downward spiral as the drugs “took over his life,” and he reached out to his mom and stepfather for love and support. “My life is going down a bad road and I can’t stop,” Guardado told Mrs. Umlauf. Mrs. Umlauf thought Guardado’s crimes were “crimes of desperation by an addict” and that “[t]he need [for drugs took] over all [his] senses.” “This crime,” she asserted, “would never have happened without drug involvement.” She described Guardado as a “victim here”—of the drug trade, the stresses of life, and the Florida penal system. The jury disagreed that Guardado was the victim here. It returned a unanimous recommendation that Guardado receive the death penalty because the aggravating factors had been proven USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 22 of 76

22 Opinion of the Court 22-10957

beyond a reasonable doubt and outweighed the mitigating circum- stances. 4. Spencer Hearing After the jury’s death penalty recommendation, the state trial court held a presentence hearing under Spencer v. State, 615 So. 2d 688 (Fla. 1993), to determine whether either party had fur- ther aggravating or mitigating evidence to present outside the pres- ence of the jury. Guardado began by trying to waive the Spencer hearing and, through Mr. Gontarek, made clear he “want[ed] to be sentenced” that day. The state trial court then asked if Guardado wanted a Spencer hearing and if he had any further mitigation to present. In response, Guardado emphasized that he had “no knowledge of any further mitigation that [he could] present.” Again, the state trial court asked if Guardado knew of any other mitigation evidence. This time, Guardado answered that he wanted to speak with the court “outside of the public.” The state trial court told Guardado it couldn’t do that. With his request de- nied, Guardado had “nothing further to say.” The state trial court explained further it couldn’t have the requested conversation alone with Guardado. Guardado then wanted “to make it known that” he did not want a Spencer hearing and “wish[ed] for sentencing to be imposed” that day. “[F]rom day one,” he added, he “wanted this to be over with as expediently as possible,” but “at every turn” it had “been delayed, delayed, and delayed.” While he understood it was “of grave concern,” it was “time to put it to an end.” USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 23 of 76

22-10957 Opinion of the Court 23

At that point, the prosecutor told the state trial court that he had no further aggravating evidence to put on, but he did submit a letter from Ms. Malone’s sister. So the state trial court turned back to Guardado, asking a third time if there was any further mitiga- tion he wanted presented. Guardado then, for the first time, com- plained that he was unhappy with trial counsel’s performance. The state trial court asked what evidence that Guardado “wished [trial counsel] would present on [his] behalf.” But Guardado responded that there were “things” he couldn’t “discuss in a public environ- ment.” Pressing on, the state trial court tried again to confirm what evidence Guardado wished his counsel had presented, but he con- tinued to identify no new evidence while adding that he could not speak about “these things . . . in a public situation . . . until [the] sentence [was] imposed.” Undeterred, the state trial court again asked Guardado what evidence he had wanted his counsel to pre- sent. Rather than answering that question, Guardado said that he was “not of a legal mind” and that his counsel’s “knowledge in that area [was] greater than” his. He listed several objections he wanted his counsel to make during trial, complained about his attorneys’ “great indifference,” and reiterated that he was “going to ask, once again, that the [s]tate impose [the] sentence” at that time. Finally, the state trial court turned back to Mr. Gontarek and asked what further mitigation evidence he would like to present. Mr. Gontarek presented Dr. Larson’s written report to supplement his trial testimony. Dr. Larson’s report described Guardado’s up- bringing, noting that Guardado “considered his mother to always be a loving, thoughtful[,] and concerned mother,” and that “[h]e USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 24 of 76

24 Opinion of the Court 22-10957

later came to respect his stepfather and realized, in retrospect, that the stepfather had good intentions.” The report also explained that: (1) Guardado disclaimed prior mental health treatment but had to attend classes for sexual offenders in the mid-1980s; (2) he entered substance abuse treatment as an early teen at juvenile facil- ities in Pensacola, Marianna, and Orlando; (3) he started using ma- rijuana, alcohol, and Quaaludes in his early teen years but gradu- ated to cocaine; (4) he suffered two major traumas as a child (the crib death of a sibling and being sexually molested by a neighbor); (5) his biological father passed away before Guardado had devel- oped any lasting memories of him; and (6) his preteen years were happy, but his teen years became unhappy with increasing family discord (much of which was over his substance abuse). Dr. Larson’s report noted that Guardado’s psychosocial his- tory was significantly affected by having spent about twenty-three years of his adult life behind bars. As of the time Dr. Larson pre- pared his report, Guardado didn’t want to go into the details of Ms. Malone’s murder with him. But Guardado told Dr. Larson that he’d been sleep-deprived and using cocaine constantly in the days before the murder—as Dr. Larson put it, Guardado “basically de- scribed himself as on a two-week cocaine binge” before the mur- der. Guardado “took full responsibility” for his crime and ex- pressed to Dr. Larson that he was remorseful. Much like his testi- mony, Dr. Larson concluded that the murder “appear[ed] to be sit- uational, driven by chemical addiction.” USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 25 of 76

22-10957 Opinion of the Court 25

Turning to the possible mitigating circumstances, a jury, Dr. Larson wrote, might “find it mitigating” to know about Guardado’s efforts to work as a plumber and become certified in waste and water treatment while previously incarcerated. As he told the jury during his penalty phase testimony, Dr. Larson also opined that the jury “may also find it mitigating” that Guardado was “under emotional duress during the time frame of ” the mur- der because, for example, he was “having difficulty adjusting to a computerized society,” had lost his job, and was in the midst of “a two-week crack cocaine binge.” Finally, a jury might find it miti- gating that Guardado did “not suffer a mental illness or major emo- tional disorder” and “expressed extreme remorse” for the murder. Before adjourning the hearing, the state trial court asked Guardado if he had anything else he would like to say. Guardado answered that he would “like to be sentenced” and “have the mat- ter resolved” because it had been “continued and carried forth too long.” 5. Sentencing Based on the evidence presented at the penalty phase and the Spencer hearing, the state trial court found that five aggravating factors were proven beyond a reasonable doubt: (1) Guardado was under conditional release when he committed the murder, FLA. STAT. § 921.141(5)(a) (2005); (2) he had been convicted of another capital felony or a felony involving the use or threat of vio- lence—that is, armed robbery, robbery with a deadly weapon, rob- bery, robbery with a weapon, and attempted robbery with a deadly USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 26 of 76

26 Opinion of the Court 22-10957

weapon, id. § 921.141(5)(b); (3) he committed the murder while en- gaged in the commission of a robbery with a weapon, id. § 921.141(5)(d); (4) the murder was especially heinous, atrocious, or cruel, id. § 921.141(5)(h); and (5) he committed the murder in a cold, calculated, and premeditated manner, with no pretense of moral or legal justification, id. § 921.141(5)(i). The state trial court also found, as non-statutory mitigating circumstances, that Guardado: (1) had entered a plea of guilty without asking to bargain or for a favor (to which the state trial court gave great weight); (2) had fully accepted responsibility (great weight); (3) wasn’t a psychopath and wouldn’t be a danger in prison if given a life sentence (moderate weight); (4) could contrib- ute to the prison population as a plumber or an expert in wastewater treatment if given a life sentence (little weight); (5) had fully cooperated with law enforcement (great weight); (6) had a good jail record awaiting trial with no disciplinary reports (little weight); (7) had consistently shown remorse (great weight); (8) had suffered throughout his adult life with addiction to crack cocaine, which was the basis for his crimes (some weight); (9) had a good family and support that could help him contribute in prison (mod- erate weight); (10) would try to counsel other inmates if given life in prison (moderate weight); (11) had suffered major trauma as a child due to the crib death of a sibling (moderate weight); (12) had suffered major trauma as a child by being sexually molested by a USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 27 of 76

22-10957 Opinion of the Court 27

2 neighbor (moderate weight); (13) had a lengthy history of sub- stance abuse starting in his early teens, graduating to alcohol and cocaine, with substance abuse treatment beginning at about age fourteen or fifteen (little weight); (14) had suffered the death of his biological father before developing lasting memories of him (little weight); (15) was raised by his loving, thoughtful, and concerned mother and stepfather and recognized that discord with his family during his teen years mostly concerned his substance abuse (little weight); (16) was under emotional duress during the timeframe of the crime (little weight); (17) didn’t suffer a mental illness or major emotional disorder (little weight); (18) had offered to release his personal property and truck to his girlfriend (little weight); and (19) had previously contributed to state prison facilities as a plumber and in wastewater treatment (little weight). The trial court found no statutory mitigating circumstances. Lastly, the state trial court gave the jury’s advisory sentence great weight. Considering that recommendation and “the aggra- vating and mitigating circumstances found to exist . . . , being ever mindful that a human life [wa]s at stake,” the state trial court found, “as did the jury, that the aggravating circumstances outweigh[ed] the mitigating circumstances.” Accordingly, it sentenced Guardado to death.

2 When the state trial court reached this factor at sentencing, Guardado inter- rupted: “Objection; objection. Your Honor, I’m not—I’m not going to deal with that.” Guardado then asked to be excused from the courtroom, but the state trial court denied that request. USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 28 of 76

28 Opinion of the Court 22-10957

Guardado directly appealed his sentence to the Florida Su- preme Court, which affirmed. See Guardado v. State, 965 So. 2d 108 (Fla. 2007). The United States Supreme Court denied certiorari. See Guardado v. Florida, 552 U.S. 1197 (2008). D. Rule 3.851 Motion for Postconviction Relief Guardado moved for postconviction relief under Florida Rule of Criminal Procedure 3.851. In his motion, Guardado as- serted his trial counsel were constitutionally ineffective under Strickland by failing to: (1) investigate and present mitigating evi- dence for the penalty phase; and (2) challenge for cause or peremp- torily strike Jurors Pennington, Hall, and Cornelius. As to the first claim, Guardado identified lay witnesses he alleged trial counsel knew or should have known about but failed to present at the penalty phase: Major Mathis, the retired warden who “had first hand [sic] knowledge that Guardado was a model inmate who worked very hard and contributed positively to the prison environment for many years”; Tommy Lancaster, Mark Mestrovich, and John Harris, “civilian employees who operated vo- cational programs at Sumter [Correctional Institution]” and “could testify that Guardado was an outstanding inmate and never a disci- plinary problem”; the custodian of records at his former employer, who would’ve testified that Guardado, while working for the com- pany, contributed to public safety; and Linda Warren (Guardado’s stepsister), Bennie Guardado (his brother), Elizabeth Padgett (his friend), and Ms. Porter, Guardado’s “family members or close friends” who “could all have attested that, when not abusing drugs, USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 29 of 76

22-10957 Opinion of the Court 29

he was a good friend and relative who cared about his family and often did good things for others without any expectation of re- ward.” Guardado also alleged that trial counsel failed to present expert testimony about his mental health that would have estab- lished two statutory mitigating circumstances: (1) he “was under the influence of extreme mental or emotional disturbance,” and (2) “his ability to appreciate the criminality of his conduct or to conform his conduct to the requirements of law was substantially impaired.” See FLA. STAT. § 921.141(6)(b), (f ) (2005). For his second claim, Guardado alleged that his trial counsel were deficient in not trying to remove—either for cause or through a peremptory strike—Jurors Pennington, Hall, and Cornelius from the jury. Guardado asserted that he suffered prejudice from the three jurors being seated on the jury because, if they weren’t seated, “there [was] a distinct likelihood and reasonable probability that a majority of the jurors would have voted for life in prison.” 1. The Evidentiary Hearing The state habeas court held an evidentiary hearing on Guardado’s Rule 3.851 motion for postconviction relief. Guardado called two of the lay witnesses he identified in his motion (Major Mathis and Ms. Padgett), his mother, himself, and two mental health professionals: Joanna Johnson and Dr. Greg Prichard. The state called Mr. Cobb and Mr. Gontarek, Guardado’s trial counsel. a. Major Mathis Major Mathis testified that she worked at Sumter Correc- tional Institution, where Guardado had previously been USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 30 of 76

30 Opinion of the Court 22-10957

incarcerated, until 2006. While in prison at Sumter Correctional, Guardado worked in the forestry camp, a minimum-security op- portunity available to well-behaved prisoners. Major Mathis couldn’t remember anything specific about Guardado other than his name, that he worked in wastewater treatment, and that he was at the forestry camp. To her knowledge, he was a good worker. As to whether anyone asked her to testify during the penalty phase, she thought she remembered hearing from someone about testify- ing but didn’t recall who’d reached out to her or how she’d been contacted. Major Mathis didn’t remember reviewing any discipli- nary reports about Guardado during his time at Sumter Correc- tional, and she would’ve had access to his file to answer any ques- tions put to her at the time of the penalty phase, so long as her legal department authorized her to do so. But she acknowledged that if the state trial court already knew Guardado was incarcerated at Sumter Correctional and worked in wastewater treatment, there wasn’t much she could add. b. Mrs. Umlauf Guardado’s mother testified that Mr. Gontarek had called her about testifying during the penalty phase but said he’d felt it would upset Guardado for her to take the stand and thus told her that he thought she shouldn’t testify. She had written the letter in support of her son, which had been introduced as an exhibit, at Mr. Gontarek’s request. Had she been asked, Mrs. Umlauf would’ve been willing to testify about the trouble Guardado faced while growing up, including the death of his father and brother, his USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 31 of 76

22-10957 Opinion of the Court 31

time served at a juvenile facility (in which she didn’t recall any bad experiences), his lifelong drug problem, his difficulty reentering so- ciety from prison, and his plea for help. On cross-examination, she repeated that in addition to what she included in the letter, she would’ve testified to Guardado’s experience of losing his father and brother. c. Ms. Padgett Ms. Padgett, who struck up a friendship with Guardado af- ter his release from prison, testified that she’d known Guardado in social settings to be pleasant and fun to be around. She would often go out at night with Guardado and Guardado’s ex-girlfriend, Ms. Porter, where Guardado would act as their protector. But over time she saw a change in Guardado as he developed relationship problems with Ms. Porter (eventually leading to their breakup), drank more heavily, and became angrier. Ms. Padgett thought Guardado started using methamphetamine and cocaine, looked haggard, lost weight, and generally was on a downhill slide. She said that no one from the defense contacted her about testifying during the penalty phase. Throughout her testimony, Ms. Padgett had significant trouble remembering details about Guardado’s case, including the dates certain events occurred and when they happened in relation to other events. On cross-examination, she also acknowledged that she lacked a grasp of many facts and that she based her opinions on what she “believe[d].” USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 32 of 76

32 Opinion of the Court 22-10957

d. Guardado Guardado testified that Mr. Gontarek spent little time with him preparing for the penalty phase—about an hour total. Mr. Gontarek told him that, if he cooperated with Dr. Larson, then “everything else would be all right.” Guardado claimed he’d given Mr. Gontarek a few people to contact—his associates and people who worked at Sumter Correctional—but when asked if he’d re- quested Mr. Gontarek contact his former employers, he explained “it’s hard for [him] to answer because it’[d] been . . . seven years.” As to his meetings with Dr. Larson, Guardado said that their pur- pose wasn’t explained to him; he was told only to meet with Dr. Larson and cooperate fully. And in the lead up to the penalty phase, he tried unsuccessfully to reach out to Mr. Gontarek and Mr. Cobb multiple times and only ever saw Mr. Cobb in court. Dis- cussing what he hoped his counsel had presented at the penalty phase, Guardado remembered wanting witnesses to testify about mitigating factors. But due to the limited time he and Mr. Gon- tarek spent together, he didn’t get to discuss with his attorneys how he wanted them to approach the case. Guardado recalled being present for jury selection but couldn’t remember the extent to which he was included in the pro- cess. He didn’t remember being specifically asked his opinion about Jurors Pennington, Hall, or Cornelius, although he did seem to recall that trial counsel told him they thought it would be good to have a juror who was familiar with law enforcement because the juror would be unemotional. To wrap up his testimony, Guardado USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 33 of 76

22-10957 Opinion of the Court 33

shared details about his drug history and life after prison, as he did with his lawyers. e. Ms. Johnson and Dr. Prichard Guardado’s final two witnesses were the mental health pro- fessionals. Ms. Johnson, a social worker specializing in addiction, went first and talked about her evaluation of Guardado using tests like the Addiction Severity Index, which is “the same as a psycho- logical evaluation except that it is specific to substance abuse.” She focused on how Guardado’s combination of cocaine use, alcohol use, and sleep deprivation may have affected his ability to make de- cisions and act. Based on her evaluation, Ms. Johnson thought there was a “real possibility” that he was under such a deep influ- ence that he “might” have suffered from a “psychosis” akin to “am- nesia” that triggered a “runaway train” of actions. She character- ized Guardado’s state of mind as “the extreme emotional disturb- ance within the use of those kind[s] of substances, other than an overdose.” In this state, Ms. Johnson explained, Guardado wouldn’t have been able to conform his behavior to societal norms and would’ve been “[u]nable to control emotion, feeling, or even stop the run that he was on.” “[H]e was,” she thought, “completely under the control of the[] drugs.” Addressing Dr. Larson’s opinions from the penalty phase, Ms. Johnson disagreed with his assessment that Guardado wasn’t under an extreme emotional disturbance. Guardado’s rapid esca- lation of drug use as a young man—from marijuana to intravenous cocaine—pointed to some “extreme emotional things.” Guardado USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 34 of 76

34 Opinion of the Court 22-10957

also went into a “full blown relapse” to cope with emotional prob- lems in rejoining society after prison life, she explained, including how he faced job difficulties and relationship problems. She agreed with Dr. Larson that Guardado wasn’t insane but disagreed that he suffered from no mental illness or emotional disorder. As for what mental illness or emotional disorder Guardado might have, Ms. Johnson opined that Guardado suffered from “[c]ompulsive obsessive behavior based on substance abuse[] [and] chronic de- pendency on cocaine and alcohol” and had “many” emotional dis- orders including his substance abuse. All of this, she summed up, spoke to Guardado’s “constant need for the drug.” When asked, Ms. Johnson declined to opine on the correct- ness of Dr. Larson’s finding that Guardado had no brain damage. She “sort of disagree[d]” that Guardado had no psychosis, because chronic cocaine use mixed with alcohol can create a form of psy- chosis akin to a “blackout.” Finally, Ms. Johnson disagreed with Dr. Larson’s assessment that Guardado had the capacity to appre- ciate the wrongness and criminality of murdering Ms. Malone be- cause “what feels or is normal under the duress or use of extensive narcotics is the . . . norm at that moment.” She concluded Guardado’s decision to murder was driven solely by his desire to obtain more crack. On cross-examination, Ms. Johnson admitted that she couldn’t state with certainty that she reviewed Guardado’s mental state through the lens of the complete record; she hadn’t been given the audio of Guardado’s confession and wasn’t sure if she USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 35 of 76

22-10957 Opinion of the Court 35

had the complete penalty phase transcript. The prosecutor asked Ms. Johnson to square her testimony that Guardado had been com- pletely under the control of drugs and alcohol and was unable to appreciate the criminality of his conduct—that “he was going to obtain more drugs,” come hell or high water—with the over- whelming evidence that Guardado deliberately planned to murder and rob Ms. Malone because she was old, alone, and secluded. Ms. Johnson didn’t know and couldn’t answer why Guardado wouldn’t have just murdered and robbed the first person he met on the street instead, but whatever Guardado’s actions were, she re- mained sure they were driven solely by his goal to get more drugs. Ultimately, the difference between her and Dr. Larson’s opinions, she thought, boiled down to whether Guardado was an addict and thus suffered from a mental health disorder. Dr. Greg Prichard, a forensic psychologist, testified next. Tasked with looking at potential statutory mitigation, Dr. Prichard reviewed investigative reports on Ms. Malone’s murder, Dr. Lar- son’s report, the Florida Supreme Court opinion from Guardado’s direct appeal, other historical information on Guardado, and testi- mony at the penalty phase. Dr. Prichard also interviewed Guardado at the prison for two and a half to three hours to get anything else he had to offer. Based on his investigation, Dr. Prichard concluded that two statutory mitigating circumstances could or should have applied at the penalty phase: (1) that Guardado was under the influence of extreme mental or emotional disturbance at the time of the USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 36 of 76

36 Opinion of the Court 22-10957

murder, and (2) that Guardado’s capacity to appreciate the crimi- nality of his conduct or conform his conduct to the requirements of the law was substantially impaired. As support for the first stat- utory mitigating circumstance, Dr. Prichard cited Guardado’s em- ployment, relationship, and money troubles as “things [that] were kind of spinning out of control.” “[T]he emotional disturbance as- sociated with the[se] things that were going on in [Guardado’s] life at the time,” plus “a lot of depression [that was] present,” was the “catalyst” of Guardado’s “need for the drug.” Turning to his potential disagreements with Dr. Larson, Dr. Prichard agreed with Dr. Larson’s finding that Guardado wasn’t insane, but he thought the question of mental illness was “a little more debatable.” Guardado did not have “the severe mental illness like schizophrenia or bipolar disorder,” but there “could be a good argument for [Guardado] having a depressive illness or anx- iety illness” in Dr. Prichard’s view. Dr. Prichard squarely disagreed with Dr. Larson’s opinion that Guardado had no emotional disor- ders, saying that “addiction is an emotional issue” with “roots in childhood development.” And Guardado’s tendency to “medicate” his emotions with drugs since “an early age . . . of fourteen or fif- teen,” “an indication of severe addiction,” was “unusual” and “sug- gest[ed] very extreme emotional pain, emotional alienation, and an inability to deal with the emotional aspects” of his environment. As to the second statutory mitigating circumstance, Dr. Prichard echoed Ms. Johnson’s testimony and said that Guardado “was very much in full-blown relapse and full-blown USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 37 of 76

22-10957 Opinion of the Court 37

addiction.” Dr. Prichard described how Guardado “was using crack cocaine on a daily basis” for two weeks before the murder, “so he was actually having a binge on crack cocaine.” Guardado’s binge, in Dr. Prichard’s view, created an obsessive-compulsive motivation to attain his next high by any means necessary, causing him to lack the “moral brakes” he otherwise would’ve had to understand or appreciate the consequences of his actions until after he got off the high. Dr. Prichard identified a common “dichotomy” in addicts whereby they’re “great people” when they’re not using and “often violent and aggressive” when they are, which “seem[ed] to be pre- sent with Mr. Guardado.” Then, finding common ground with Dr. Larson, Dr. Prich- ard agreed that Guardado had no discernible, obvious brain dam- age and “suffered from no psychosis.” He equivocated on whether Guardado had the capacity to appreciate the wrongness and crimi- nality of the murder. One could argue that Guardado appreciated it, Dr. Prichard thought, but the question was how much he could appreciate it and whether he had the capacity to conform his con- duct to the law. “It was going to be a bad night for Mr. Guardado and whoever got in his way,” as Dr. Prichard put it. The addiction “was driving [Guardado],” and he murdered solely to obtain more crack. On cross-examination, when pressed about the extent of his disagreement with Dr. Larson, Dr. Prichard agreed they reached the same “bottom line”: Guardado was driven to obtain more co- caine as a result of his addiction. But Dr. Prichard believed USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 38 of 76

38 Opinion of the Court 22-10957

Dr. Larson didn’t articulate why he felt the statutory mitigating cir- cumstance related to the inability to conform one’s conduct to the law would not apply. Dr. Prichard clarified that Guardado was deal- ing with more of an emotional rather than mental disturbance. Things had been going “very well” for Guardado when he got out of prison—even though “he was drinking and using some [drugs]”—and his cocaine binge only happened after “things started going really badly” when he was arrested for DUI, lost a job he enjoyed, lost a girlfriend, and lacked a steady place to stay. “[T]he idea,” Dr. Prichard said, “is that—that it bec[a]me[] ex- treme” not because of any one thing but because of “a variety of losses, a variety of things that were not going well for him in the context of initially doing things pretty well.” Dr. Prichard acknowl- edged that Guardado had both a new job when he murdered Ms. Malone and that he’d had multiple past girlfriends, but he in- dicated that Guardado was still feeling the losses associated with his previous employment and a past relationship. f. Mr. Cobb The state also put on its witnesses. Mr. Cobb testified about how he became involved in the case, his early meetings with Mr. Gontarek, his introduction to Guardado, and his role in the case. From the outset, he got the sense that Guardado didn’t want to go forward with the penalty phase and just wanted to “plead to death.” He recalled attending an initial meeting at the jail with Mr. Gontarek and Guardado, at which Mr. Gontarek explained that if Guardado wanted to skip to the death penalty, a Spencer hearing USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 39 of 76

22-10957 Opinion of the Court 39

would still be required. But nothing substantive was discussed at this visit, and during Mr. Cobb’s meetings with Guardado “there was never really any discussion about” mitigation evidence; Guardado just wanted to get the death penalty over with. Guardado never suggested any mitigating evidence Mr. Gontarek or Mr. Cobb might present. But, in spite of Guardado’s desire to get things over with, Mr. Cobb reached out to Mrs. Umlauf and Ms. Porter to see if they had mitigating evidence to offer. And the defense briefly discussed whether to present evi- dence about Guardado’s skills in wastewater treatment, which was ultimately discussed by Guardado himself at the penalty phase. As to what he remembered about jury selection, Mr. Cobb recalled that Guardado raised an issue about a prospective juror “glaring” at him. This prospective juror was ultimately challenged either for cause or peremptorily and was not seated. Before the defense agreed to seat Jurors Pennington, Hall, and Cornelius, the “perceived problems” and “perceived positives” of having them on the jury were discussed. All considerations were weighed in Guardado’s presence, continued Mr. Cobb, so that everyone would understand the “pros and . . . cons” of having any particular indi- vidual on the jury. Ultimately, Mr. Gontarek made the final choices, and Guardado agreed with them. Recalling Juror Cornelius specif- ically, Mr. Cobb remembered discussing his opinion that a life sen- tence may be harsher than the death penalty as a positive. And for Juror Hall, his familiarity with law enforcement officers wasn’t con- sidered a problem because Mr. Gontarek’s penalty phase strategy USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 40 of 76

40 Opinion of the Court 22-10957

wasn’t to call the officers’ credibility into question. Instead, Mr. Gontarek planned to emphasize Guardado’s cooperation with the police. g. Mr. Gontarek Mr. Gontarek described his mitigation strategy and the in- vestigation he and Mr. Cobb undertook during the penalty phase. Recounting his first meeting with Guardado, Mr. Gontarek remem- bered being “impressed” that Guardado had “taken responsibility,” which Mr. Gontarek thought “would go a long way in mitigation.” His plan was “to bring in Dr. Larson, who [was] a well[-]known and respected forensic psychologist,” and “to talk to the defendant’s family who would talk to [him]—[Guardado’s] mother, . . . [Ms. Porter and other] family members or friends, [and] any employers.” He testified that Guardado’s testimony about only meeting with him for an hour, total, wasn’t accurate; they had met “[n]umerous times.” Mr. Gontarek’s billing records were introduced to confirm the time he’d spent on the case. Mr. Gontarek testified that Ms. Umlauf declined to appear as a witness, which is why he ultimately asked her to write the letter he presented at the penalty phase. As to Dr. Larson’s opinions on Guardado, Mr. Gontarek explained that since the state trial court had appointed Dr. Larson to the case, Mr. Gontarek couldn’t simply seek a more favorable opinion from a different expert once Dr. Larson determined no statutory mitigating circumstances were present. USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 41 of 76

22-10957 Opinion of the Court 41

Moving to voir dire, Mr. Gontarek testified that he would’ve moved to strike Jurors Pennington, Hall, and Cornelius if he thought they were actually biased against Guardado. He remem- bered Guardado playing an active role in jury selection but that Guardado never requested that any of the three jurors be removed. If Guardado had, Mr. Gontarek first would’ve attempted a for- cause challenge, then he would’ve used his one remaining peremp- tory strike, and finally he would’ve requested more peremptory strikes from the state trial court if needed. Next, Mr. Gontarek explained his reasons for seating the three jurors. A person who “was only somewhat in favor of the death penalty”—like Juror Pennington—was the kind of person “[he]’d want to keep.” As to Juror Hall’s ties to law enforcement, Mr. Gontarek explained his strategy was to highlight Guardado’s full cooperation, which he “thought was significant mitigation.” He thought that someone who knew the officers might have been impressed with their testimony confirming Guardado cooperated with them. And Juror Cornelius thought a life sentence could be harsher than the death penalty, which was “a reason that [Mr. Gon- tarek] may have desired that he remain on the” penalty phase jury. Later, when asked about his decision to seat Juror Cornelius, Mr. Gontarek clarified that any number of factors might’ve been considered when deciding to seat him, including his body language, tone of voice, the dynamics of the other prospective jurors, his in- teractions with counsel and Guardado, Mr. Gontarek’s conversa- tion with Guardado, and other things he said—like that he was only somewhat in favor of the death penalty. The result of the jury USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 42 of 76

42 Opinion of the Court 22-10957

selection process, Mr. Gontarek confirmed, was a jury that Guardado agreed to seat. On cross-examination, Mr. Gontarek couldn’t recall why he entered Mrs. Umlauf ’s letter into evidence without reading it to the jury, which he admitted could’ve been more effective. Going back to Dr. Larson, Mr. Gontarek explained that he provided Dr. Larson with information about the case so that Dr. Larson could deter- mine whether any statutory mitigating circumstances applied to Guardado. Mr. Gontarek described how in cases where Dr. Larson determines no statutory mitigating circumstances apply, Mr. Gon- tarek would ask him to identify any non-statutory mitigating cir- cumstances they could present to a jury instead. And before the penalty phase, Mr. Gontarek prepared Dr. Larson to testify and re- viewed his report with him. h. Other Evidence At the end of the evidentiary hearing, the state habeas court admitted into evidence four letters—received long after the penalty phase—by Ms. Warren, Bennie Guardado, Ms. Porter, and Ms. Padgett. Ms. Warren wrote that Guardado had been doing well when he first left prison, but then the pressures of adjusting to society combined with his lack of family support in DeFuniak Springs caused him to spiral out of control. Bennie Guardado added how they lost their father and another brother, how a boy in a nearby apartment complex provided Guardado with alcohol and marijuana, how another introduced Guardado to harder drugs like Quaaludes and he stole to cover the cost, how Guardado descended USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 43 of 76

22-10957 Opinion of the Court 43

into addiction and crime, and how he spent time in a juvenile facil- ity. Ms. Porter wrote about how she’d first met Guardado after his release from prison and what she’d witnessed as he coped with life on the outside. And Ms. Padgett wrote about Guardado’s relation- ship problems and the drug-related changes she’d witnessed in him. The letters maintained that Guardado could be a model pris- oner because, when he’s not high or exposed to crack and alcohol, he’s a good worker and a person who’s remorseful for his crimes. None of the letters, though, said that the author was willing and available to testify about these facts during the penalty phase. 2. The State Habeas Court Denied Guardado’s Rule 3.851 Motion The state habeas court denied Guardado’s Rule 3.851 mo- tion, concluding that he wasn’t entitled to relief on either of his two Strickland claims. As for Guardado’s claim that trial counsel were ineffective for failing to investigate and present mitigation evidence during the penalty phase, the state habeas court divided its Strick- land analysis into two parts. First, it addressed the ten lay wit- nesses, whom Guardado maintained trial counsel could have used to show his background and good behavior while not using drugs. The state habeas court found that their testimony and letters from the Rule 3.851 hearing “only contain[ed] background information” and would’ve been “cumulative” or “largely cumulative” of the ev- idence that came in about Guardado’s background and behavior during the penalty phase. Because the lay witness testimony and letters about Guardado’s background would’ve been cumulative or USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 44 of 76

44 Opinion of the Court 22-10957

largely cumulative, the state habeas court concluded that Guardado failed to establish Strickland prejudice because there was no reasonable probability of a different result if the evidence came in during the penalty phase. The state habeas court next addressed the mental health tes- timony—offered by Ms. Johnson and Dr. Prichard—which Guardado asserted trial counsel could’ve presented to show his mental state at the time of the murder. The state habeas court found that Ms. Johnson and Dr. Prichard testified to “the same in- formation” as Dr. Larson because all three witnesses described how Guardado “murdered the victim because of his addiction to cocaine.” So, the state habeas court concluded, Guardado failed to satisfy Strickland’s prejudice standard for this part of his first claim, too. As for Guardado’s second claim that trial counsel were inef- fective for failing to challenge for cause or peremptorily strike Ju- rors Pennington, Hall, and Cornelius, the state habeas court found that Guardado failed to demonstrate the jurors were unfair or bi- ased; instead, during voir dire, each juror promised to be fair. Thus, the state habeas court determined Guardado was not preju- diced by trial counsel’s failure to challenge or strike them. 3. The Florida Supreme Court Affirmed the State Habeas Court’s Denial of Guardado’s Rule 3.851 Motion The Florida Supreme Court affirmed the denial of Guardado’s Rule 3.851 motion because it “agree[d]” with the state habeas court that he failed to show Strickland prejudice for each of USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 45 of 76

22-10957 Opinion of the Court 45

his two claims. Guardado v. State, 176 So. 3d 886, 893–96, 899 (Fla. 2015). Starting with Guardado’s first claim that trial counsel failed to adequately investigate or present mitigation evidence, the Flor- ida Supreme Court—like the state habeas court—began with the lay witness testimony and letters from the Rule 3.851 hearing. See id. at 893–95. Based on its “review of the record,” the Florida Su- preme Court determined “that there was no . . . prejudice as to counsel’s failure to contact” or present these lay witnesses during the penalty phase. Id. at 894. The lay witness testimony and letters “contain[ed] mostly background information” and would have “substantively track[ed]” or been “cumulative” of evidence already presented at the penalty phase. See id. at 893–96. Turning to the mental health testimony, the Florida Su- preme Court explained that “there was no need” for Ms. Johnson’s and Dr. Prichard’s additional testimony because Dr. Larson had tes- tified to Guardado’s mental state at the time of the murder during the penalty phase. Id. at 895–96 (agreeing with the state habeas court’s conclusion that the new experts’ testimony “mirrored” Dr. Larson’s). The Florida Supreme Court acknowledged that Ms. Johnson and Dr. Prichard disagreed with Dr. Larson on how to label Guardado’s emotional stress and drug addiction. Id. But “alt- hough [Dr. Larson] was not as favorable as the defense would have liked,” “[s]imply presenting the testimony of experts . . . that [we]re inconsistent with the mental health opinion of an expert re- tained by trial counsel d[id] not rise to the level of prejudice USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 46 of 76

46 Opinion of the Court 22-10957

necessary to warrant relief.” Id. at 896 (quoting Dufour v. State, 905 So. 2d 42, 58 (Fla. 2005)) (“This was not a scenario where an expert who could provide mitigating testimony about the defendant was not called and counsel instead relied on one witness who did not provide specific details regarding mitigating information.”). Second, turning to Guardado’s claim that trial counsel were ineffective for failing to challenge for cause or peremptorily strike Jurors Pennington, Hall, and Cornelius, the Florida Supreme Court applied its own gloss on Strickland prejudice from Carratelli v. State, 961 So. 2d 312 (Fla. 2007). The Florida Supreme Court explained that, under its Carratelli test, a petitioner raising an ineffective assis- tance of counsel claim that his trial counsel failed to challenge a juror for cause or use a peremptory strike “must demonstrate that [the] juror was actually biased.” Guardado, 176 So. 3d at 899 (quot- ing Carratelli, 961 So. 2d at 324). “[A]ctual bias,” as Carratelli de- fined it, “mean[t] bias-in-fact that would prevent service as an im- partial juror.” Id. (quoting Carratelli, 961 So. 2d at 324). Applying the Carratelli test, the Florida Supreme Court determined Guardado wasn’t prejudiced by trial counsel’s failure to challenge or strike Jurors Pennington, Hall, and Cornelius because he failed to satisfy Carratelli’s actual bias test. Id. E. Section 2254 Petition After the Florida Supreme Court affirmed the denial of Guardado’s Rule 3.851 motion, Guardado petitioned the district court for a writ of habeas corpus under 28 U.S.C. section 2254. He asserted that the Florida Supreme Court unreasonably applied USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 47 of 76

22-10957 Opinion of the Court 47

Strickland in denying his claims that (1) his trial counsel were inef- fective for failing to investigate and present mitigation evidence, and (2) his trial counsel were ineffective for failing to challenge for cause or peremptorily strike Jurors Pennington, Hall, and Cor- nelius. The district court denied Guardado’s petition. As to Guardado’s first claim, the district court concluded the Florida Su- preme Court did not unreasonably determine that Guardado failed to show prejudice. The district court “compar[ed] the evidence presented during the trial and Spencer hearing, on the one hand, to the evidence presented at the [Rule 3.851] hearing, on the other hand.” Comparing the evidence, the district court explained, showed that “the mitigation case presented at trial was fundamen- tally the same as” the one Guardado presented at the Rule 3.851 hearing. Because the mitigation cases were fundamentally the same, the Florida Supreme Court’s determination that the new ev- idence was unlikely to have made any difference wasn’t unreason- able. Second, turning to Guardado’s claim that trial counsel were ineffective in failing to challenge or peremptorily strike Jurors Pen- nington, Hall, and Cornelius, the district court concluded the Flor- ida Supreme Court didn’t unreasonably determine that Guardado failed to show prejudice because there was no evidence of “juror bias.” USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 48 of 76

48 Opinion of the Court 22-10957

F. Certificate of Appealability The district court denied Guardado a certificate of appeala- bility, but we granted one on two issues: 1. Whether the Florida Supreme Court unreasonably applied Strickland on Guardado’s claim that his trial counsel was ineffective by failing to investigate and present mitigating evidence adequately, but only to the extent the particular legal theory and the specific factual foundation on which the mitigating evidence claim rests were raised and exhausted in the state courts. 2. Whether the Florida Supreme Court unreasonably applied Strickland on Guardado’s claim that his trial counsel was ineffective by failing to challenge for cause or peremptorily strike Jurors Pennington, Hall, and Cornelius. This appeal followed. II. STANDARD OF REVIEW We review de novo a district court’s denial of a federal ha- beas petition. Ward v. Hall, 592 F.3d 1144, 1155 (11th Cir. 2010). III. DISCUSSION For habeas claims resolved in state court, like Guardado’s, “we review the last state-court adjudication on the merits.” Sears v. Warden GDCP, 73 F.4th 1269, 1280 (11th Cir. 2023) (internal quota- tion marks and citation omitted). The Antiterrorism and Effective USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 49 of 76

22-10957 Opinion of the Court 49

Death Penalty Act of 1996’s (AEDPA) “highly deferential frame- work” generally “demands that [the] state-court decision[] be given the benefit of the doubt.” Id. at 1279 (citations omitted). Under this framework, AEDPA precludes federal habeas relief “unless the state court’s ‘adjudication of the claim . . . resulted in a decision that was contrary to, or involved an unreasonable application of, clearly established [f ]ederal law, as determined by the Supreme Court of the United States.’” Pye v. Warden, Ga. Diagnostic Prison, 50 F.4th 1025, 1034 (11th Cir. 2022) (en banc) (quoting 28 U.S.C. § 2254(d)(1)). If the state court unreasonably applied clearly estab- lished federal law, we do not defer to its denial of relief and, instead, review de novo the petitioner’s claim. Calhoun v. Warden, Baldwin State Prison, 92 F.4th 1338, 1346 (11th Cir. 2024). “To meet the ‘unreasonable application’ standard, ‘a pris- oner must show far more than that the state court’s decision was merely wrong or even clear error.’” Pye, 50 F.4th at 1034 (quoting Shinn v. Kayer, 592 U.S. 111, 118 (2020)). Instead, he has to show the state court’s decision was “so obviously wrong that its error lies beyond any possibility for fairminded disagreement.” Id. That means “so long as fairminded jurists could disagree on the correct- ness of the state court’s decision,” the state court did not unreason- ably apply clearly established federal law. Id. (quoting Harrington v. Richter, 562 U.S. 86, 101 (2011)). “If this standard is difficult to meet, that is because it was meant to be.” Harrington, 562 U.S. at 102–03 (“Section 2254(d) reflects the view that habeas corpus is a guard against extreme malfunctions in the state criminal justice systems, USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 50 of 76

50 Opinion of the Court 22-10957

not a substitute for ordinary error correction through appeal.” (in- ternal quotation marks and citation omitted)). But AEDPA isn’t the only “difficult to meet” standard at play here. Strickland “itself places a demanding burden on a [petitioner] to show that he was prejudiced by his counsel’s deficient perfor- mance,” Pye, 50 F.4th at 1041, and it “strongly presume[s counsel] to have rendered adequate assistance,” Cullen v. Pinholster, 563 U.S. 170, 189 (2011) (quoting Strickland, 466 U.S. at 690). Plus, the failure to establish either deficient performance or prejudice is “fatal” to an ineffective assistance claim. Tuomi v. Sec’y, Fla. Dep’t of Corr., 980 F.3d 787, 795 (11th Cir. 2020). For a petitioner to establish prejudice, he must show “there is a reasonable probability that, but for counsel’s unprofessional er- rors, the result of the proceeding would have been different.” Cul- len, 563 U.S. at 189 (quoting Strickland, 466 U.S. at 694). In the cap- ital sentencing context, that requires showing “a reasonable proba- bility that, absent [counsel’s] errors, the sentencer . . . would have concluded that the balance of aggravating and mitigating circum- stances did not warrant death.” Thornell v. Jones, 144 S. Ct. 1302, 1310 (2024) (alteration in original) (quoting Strickland, 466 U.S. at 695). A reasonable probability is one that’s “sufficient to under- mine confidence in the outcome,” which requires “a substantial, not just conceivable, likelihood of a different result.” Id. (quoting Cullen, 563 U.S. at 189); see Strickland, 466 U.S. at 687 (noting “that counsel’s errors [must have been] so serious as to deprive the de- fendant of a fair trial”). USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 51 of 76

22-10957 Opinion of the Court 51

Combining the two standards by “[a]pplying AEDPA to Strickland’s prejudice standard,” the question for the federal habeas court is “whether the state court’s conclusion that [counsel]’s per- formance at the sentencing phase of [the petitioner]’s trial didn’t prejudice him—that there was no ‘substantial likelihood’ of a dif- ferent result—was ‘so obviously wrong that its error lies beyond any possibility for fairminded disagreement.’” Pye, 50 F.4th at 1041–42 (quoting Shinn, 592 U.S. at 118); see also Johnson v. Sec’y, DOC, 643 F.3d 907, 911 (11th Cir. 2011) (“[I]t will be a rare case in which an ineffective assistance of counsel claim that was denied on the merits in state court is found to merit relief in a federal habeas proceeding.”); Harrington, 562 U.S. at 105 (“The Strickland standard is a general one, so the range of reasonable applications is substan- tial.” (citation omitted)). Unless the answer is yes, “we lack the power to grant relief.” Pye, 50 F.4th at 1042. Guardado maintains that the answer here is yes. The Florida Supreme Court, he contends, unreasonably applied Strickland’s prejudice prong in denying relief on his two ineffective assistance of counsel claims. We address his arguments as to each claim in turn. A. The Florida Supreme Court’s Prejudice Determination on Guardado’s Claim That Counsel Failed to Investi- gate and Present Mitigation Guardado first argues that the Florida Supreme Court un- reasonably applied Strickland in determining that trial counsel’s performance in investigating and presenting mitigation evidence USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 52 of 76

52 Opinion of the Court 22-10957

did not prejudice the result of the penalty phase. In Guardado’s view, the Florida Supreme Court unreasonably concluded the mit- igation evidence that trial counsel should’ve investigated and pre- sented was cumulative to the penalty phase evidence trial counsel actually presented. We disagree. “[N]o prejudice can result from the exclusion of cumulative evidence.” Raheem v. GDCP Warden, 995 F.3d 895, 925 (11th Cir. 2021) (alteration in original) (quoting Dallas v. Warden, 964 F.3d 1285, 1310 (11th Cir. 2020)). “[E]vidence presented in postconvic- tion proceedings is ‘cumulative’ or ‘largely cumulative’ to . . . that presented at trial when it tells a more detailed version of the same story . . . [,] provides more or better examples[,] or amplifies the themes presented to the jury.” Holsey v. Warden, 694 F.3d 1230, 1260–61 (11th Cir. 2012); see also Raheem, 995 F.3d at 925 (“The Su- preme Court has found evidence cumulative where it ‘substanti- ate[s],’ ‘support[s],’ or ‘explain[s]’ more general testimony provided at trial.” (alterations in original) (quoting Cullen, 563 U.S. at 200– 01)). For example, in Holsey, we concluded that the state court didn’t unreasonably find the petitioner’s postconviction evidence about his “troubled, abusive childhood” was largely cumulative be- cause “the jury at the sentencing had heard about his troubled, abu- sive upbringing too.” 694 F.3d at 1264–67. And the state court didn’t unreasonably find that expert testimony about the peti- tioner’s “limited intelligence” was largely cumulative of lay testi- mony about the petitioner’s academic struggles, which “concerned the same subject matter.” Id. at 1262–64; see also, e.g., Thornell, 144 S. Ct. at 1312 (finding no Strickland prejudice where the USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 53 of 76

22-10957 Opinion of the Court 53

postconviction evidence merely “corroborate[d]” the penalty phase evidence “about [the defendant’s] head trauma and cognitive im- pairment,” including how he was physically abused and had “three falls during childhood”); id. at 1313 (same, where the petitioner’s postconviction evidence that his grandfather introduced him to drugs and alcohol as a nine-year-old was “essentially the same” as evidence showing his substance abuse began by at least age seven- teen); Robinson v. Moore, 300 F.3d 1320, 1347–48 (11th Cir. 2002) (concluding a state court didn’t unreasonably determine the peti- tioner failed to show prejudice from trial counsel’s failure to inves- tigate and present a “different example[]” of a “good deed” he per- formed because the sentencing court heard “evidence of other good deeds”); Dallas, 964 F.3d at 1308–10 (same, where the peti- tioner offered new affidavits from family about his teenage sub- stance abuse but substance abuse was a “pillar[]” of his trial de- fense). Applying these principles, the Florida Supreme Court deter- mined Guardado failed to show Strickland prejudice because the lay witness testimony and letters from the Rule 3.851 hearing “con- taine[d] mostly background information” and “substantively track[ed]” or would’ve been “cumulative” of evidence already pre- sented at the penalty phase. See Guardado, 176 So. 3d at 893–96. As for the mental health testimony, the Florida Supreme Court deter- mined Guardado failed to show prejudice because “there was no need” to investigate and present Ms. Johnson’s and Dr. Prichard’s testimony in light of Dr. Larson’s penalty phase testimony and re- port about Guardado’s mental state during the murder. Id. at 895– USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 54 of 76

54 Opinion of the Court 22-10957

96 (agreeing with the state habeas court that the new mental health testimony “mirrored” Dr. Larson’s). Guardado needed to do more, the Florida Supreme Court added, than “[s]imply present[] the tes- timony of experts . . . that [we]re inconsistent with the mental health opinion of an expert retained by trial counsel.” Id. at 896 (citation omitted). Considering the demanding burdens that AEDPA and Strick- land place on Guardado, we cannot say the Florida Supreme Court’s prejudice determinations were “so obviously wrong . . . beyond any possibility for fairminded disagreement.” See Pye, 50 F.4th at 1041–42 (quoting Shinn, 592 U.S. at 118). To explain why, we do as the district court did by “compar[ing] the trial evidence” with the evidence that Guardado “presented during the state postconviction proceedings.” See Holsey, 694 F.3d at 1260–61. 1. Lay Witness Testimony and Letters The first bucket of mitigating evidence Guardado maintains his trial counsel should have investigated and presented was the tes- timony and letters of ten lay witnesses. The background infor- mation these witnesses would’ve provided during the penalty phase captured the circumstances that pushed Guardado to crack cocaine and how, in their view, he was a good person with a solid work ethic when not using drugs. Specifically, Mrs. Umlauf testified and Bennie Guardado wrote about Guardado’s troubled youth—how he lost his father and a brother, fell into a bad crowd who introduced him to alcohol and drugs during his teen years, and ended up in a juvenile USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 55 of 76

22-10957 Opinion of the Court 55

detention facility. They described how his troubled youth turned into an adult life spent mostly incarcerated. As far as Major Mathis knew about Guardado’s incarceration at Sumter Correctional, Guardado took advantage of a work op- portunity available for well-behaved prisoners and was a good worker. Guardado’s family and friends saw him work hard outside of prison, too. But Guardado’s family and friends saw changes in him because of his problems adjusting to life outside prison. They pointed to how Guardado went through a breakup, and to how emotional stress pushed him to abuse alcohol (like when he had a DUI) and crack cocaine. Without these stressors or the exposure to crack, Guardado could be a productive, model prisoner. The lay witness testimony and letters from the Rule 3.851 hearing echoed the penalty phase evidence. Guardado’s trial coun- sel had obtained and presented Mrs. Umlauf ’s letter, Dr. Larson’s testimony and report, and Guardado’s own testimony. That evi- dence spoke to Guardado’s troubled youth—how he lost his brother in a crib death and his father, was sexually molested, “chose the wrong friends,” began using alcohol and drugs as a teenager, and ended up in a juvenile facility. Mrs. Umlauf wrote about how this troubled youth turned into an adult life spent mostly behind bars. As to Guardado’s earlier stints in jail or prison, the Walton County Sheriff’s Office letter confirmed that Guardado had no dis- ciplinary incidents while at its jail, on top of Dr. Larson’s testimony about Guardado’s prison record and how he enjoyed working in USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 56 of 76

56 Opinion of the Court 22-10957

wastewater treatment. And Mrs. Umlauf wrote that Guardado “came out [of prison] with the best of intentions” and “handl[ed] his job well,” at least until emotional stress and crack “took over his life.” Finally, Guardado, Mrs. Umlauf, and Dr. Larson emphasized during the penalty phase that, when Guardado’s not high, he would be a model prisoner again given his previous good behavior and his remorse for murdering Ms. Malone. Based on the penalty phase evidence, the state trial court found the same mitigating circumstances that a court could’ve found had it heard the lay witness testimony and letters from the Rule 3.851 hearing. The state trial court considered mitigating cir- cumstances relating to Guardado’s upbringing (like how he suf- fered major trauma as a child due to the crib death of a sibling and losing his dad, was sexually molested, and began abusing alcohol and crack at fourteen or fifteen). It considered as mitigating cir- cumstances how Guardado’s childhood troubles continued into adulthood (like how he’s had a crack addiction throughout his adult life). And it considered mitigating circumstances accounting for Guardado’s potential to be a model prisoner (including how he had a good jail record awaiting trial with no disciplinary reports, could contribute given his wastewater treatment experience, and had good family support that could help him contribute), plus his re- morse. In other words, the lay witness testimony and letters from the Rule 3.851 hearing and the penalty phase evidence both chroni- cled Guardado’s background in essentially the same way: he lost USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 57 of 76

22-10957 Opinion of the Court 57

his father and brother as a child, used alcohol and drugs as a teen, was held in a juvenile facility, spent much of his adult life in prison, worked in wastewater treatment both in and out of prison, couldn’t adjust to life outside prison, relapsed, and is remorseful for murdering Ms. Malone. The testimony and letters from the lay witnesses would’ve merely “amplifie[d] the themes” of Guardado’s penalty phase defense, even if the lay witnesses gave “more de- tail[s]” or “better examples.” Holsey, 694 F.3d at 1260–61; see also Thornell, 144 S. Ct. at 1312–13 (finding no Strickland prejudice where the postconviction evidence merely “corroborate[d]” and was “essentially the same” as the penalty phase evidence); Robinson, 300 F.3d at 1347–48 (concluding it wasn’t unreasonable to deter- mine that the petitioner failed to show prejudice where most of his new evidence merely offered “different examples”); Dallas, 964 F.3d at 1309–10 (same, where the theme of the petitioner’s postconvic- tion evidence was a “pillar[]” of his trial defense). As in Holsey, a fairminded jurist could agree with the Florida Supreme Court that Guardado failed to show prejudice because the lay witness testi- mony and letters “contain[ed] mostly background information” and “substantively track[ed]” or would’ve been “cumulative” of the penalty phase evidence. Guardado, 176 So. 3d at 893–96; see Holsey, 694 F.3d at 1262–67. 2. Mental Health Testimony The second bucket of mitigating evidence Guardado main- tains his trial counsel should have investigated and presented was mental health testimony showing how substance abuse impacted USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 58 of 76

58 Opinion of the Court 22-10957

his mental state at the time of the murder. At the Rule 3.851 hear- ing, Guardado offered Ms. Johnson’s and Dr. Prichard’s testimony. Ms. Johnson testified that Guardado was “completely under the control of . . . drugs,” akin to an “amnesia,” which triggered a “runaway train” of actions before Ms. Malone’s murder. Ms. John- son based her opinion on how Guardado went into a “full blown relapse” of crack use for two weeks before the murder as a way to cope with his stressors in rejoining society; at the time of the mur- der, he experienced a “constant need for the drug.” Guardado’s constant need for crack, Ms. Johnson testified, was the sole reason he murdered Ms. Malone. Dr. Prichard also testified that Guardado experienced “emo- tional disturbance associated with the things that were going on . . . in his life at the time”—“things were kind of spinning out of con- trol” with Guardado’s employment, love life, and finances—and that there was “a lot of depression present.” These stressors were the “catalyst” of Guardado’s “need for the drug,” because that’s what Guardado used to “medicate” his emotions. Guardado had done so, according to Dr. Prichard, since “an early age . . . of four- teen or fifteen.” And Dr. Prichard specifically cited how Guardado “was using crack cocaine on a daily basis” just before the murder, “so he was actually having a binge on crack cocaine.” He opined that this crack binge “was driving [Guardado]” on the night of the murder and “[i]t was going to be a bad night for Mr. Guardado and whoever got in his way” of getting more crack. That’s why, in USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 59 of 76

22-10957 Opinion of the Court 59

Dr. Prichard’s view, Guardado’s need for crack was the sole reason he murdered Ms. Malone. Like the lay witness testimony and letters, the mental health testimony presented at the Rule 3.851 hearing echoed the evidence that trial counsel actually presented during the penalty phase. That’s because Dr. Larson explained during his penalty phase testi- mony that Guardado had used crack dating “back to teenage years or early adult years.” Dr. Larson described how Guardado was “un- der emotional duress” and “considerable stress” while adjusting to adult life outside prison—specifically citing his “economic prob- lems,” his DUI, and that he couldn’t hold down a job. Amidst this considerable emotional stress, Dr. Larson explained, Guardado “re- lapsed and went on a crack cocaine binge for approximately two weeks” before the murder. Dr. Larson then testified—like Ms. Johnson and Dr. Prichard—that this binge made Guardado’s murder of Ms. Malone a “drug related incident” because Guardado was motivated solely by his need for crack. Or, as Dr. Larson put it in his report, the murder was “situational” and “driven by [Guardado’s] chemical addition.” True, as Guardado points out, Ms. Johnson and Dr. Prichard disagreed with Dr. Larson on how to label Guardado’s emotional duress and need for crack. But labels aside, the circumstances Ms. Johnson, Dr. Prichard, and Dr. Larson described were largely the same—Guardado decided to murder while suffering from emo- tional stress and under the influence of crack cocaine. Indeed, re- lying on Dr. Larson’s testimony and report, the state trial court USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 60 of 76

60 Opinion of the Court 22-10957

found mitigating circumstances essentially mirroring those Ms. Johnson and Dr. Prichard described. Specifically, the state trial court weighed as mitigating circumstances how Guardado was un- der emotional duress during the murder and how his crack addic- tion caused him to murder. Because Ms. Johnson’s and Dr. Prichard’s testimony at the Rule 3.851 hearing hit similar notes to Dr. Larson’s penalty phase testimony and report, our conclusion regarding the mental health testimony is the same one we reached as to the lay witness testi- mony and letters. A fairminded jurist could agree with the Florida Supreme Court that Guardado failed to show prejudice from any failure by trial counsel to investigate and present mental health tes- timony from Ms. Johnson and Dr. Prichard because “there was no need” for their testimony on top of Dr. Larson’s. Guardado, 176 So. 3d at 895–96; see Raheem, 995 F.3d at 925–26. 3. Guardado’s Arguments Guardado disagrees with our conclusion for two reasons. First, under his view of Florida law, statutory mitigating circum- stances are inherently weightier than non-statutory ones. He ar- gues that the mental health testimony from the Rule 3.851 hearing would’ve established two statutory mitigating circumstances if in- vestigated and presented during the penalty phase: (1) he mur- dered Ms. Malone under the influence of extreme mental or emo- tional distress, and (2) his capacity to appreciate the criminality of his conduct or conform it to the law was substantially impaired. See FLA. STAT. § 921.141(6)(b), (f ) (2005). Because the state trial court USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 61 of 76

22-10957 Opinion of the Court 61

found only non-statutory mitigating circumstances, the argument goes, there was a reasonable probability that the mental health tes- timony from the Rule 3.851 hearing—which established the inher- ently weightier statutory mitigating circumstances—would have tipped the balance in favor of a life sentence. But Guardado’s view of Florida law—that statutory mitigat- ing circumstances are inherently weightier than non-statutory ones—is mistaken. The Florida Supreme Court has found that non-statutory mitigating circumstances can be weightier than stat- utory mitigating circumstances. In Abdool v. State, for example, the trial court assigned “little weight” to the same statutory mitigating circumstances that Guardado relies on here—(1) extreme mental or emotional distress and (2) a substantially impaired capacity to appreciate the criminality of conduct or conform it to the law— and gave nine non-statutory mitigating circumstances “moderate weight.” 53 So. 3d 208, 223 (Fla. 2010). Rejecting the defendant’s argument that the trial court erred in assigning weights to the mit- igating circumstances, the Florida Supreme Court “h[e]ld that the trial court did not abuse its discretion when applying weight to the mitigating circumstances,” id. at 223–24 (emphasizing that “the weight to be given to existing mitigating circumstances [is] within the discretion of the sentencing court” (citation omitted)). Abdool isn’t the only example of where the Florida Supreme Court agreed that non-statutory mitigating circumstances were weightier than statutory ones. See, e.g., Francis v. State, 808 So. 2d 110, 140–41 (Fla. 2001) (rejecting the defendant’s argument that the USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 62 of 76

62 Opinion of the Court 22-10957

statutory emotional-distress mitigator should’ve been given more than “some weight,” emphasizing that the trial court “also found, as a nonstatutory mitigator, that the defendant was mentally ill or emotionally disturbed and accorded it ‘considerable weight’”); Cov- ington v. State, 228 So. 3d 49, 60–61, 66 (Fla. 2017) (rejecting “that the trial court abused its discretion in affording [the statutory emo- tional-distress mitigating] circumstance moderate weight” where “great” weight was assigned to a non-statutory mitigating circum- stance); Perez v. State, 919 So. 2d 347, 358 & n.3, 372–74 (Fla. 2005) (concluding that the trial court didn’t abuse its discretion in assign- ing “little weight” to the statutory emotional-distress mitigator alt- hough it had given five non-statutory mitigating circumstances “moderate” or “some weight,” disagreeing with the defendant “that [the] trial court [wa]s required to assign any certain weight to the [statutory] mitigating circumstance in the abstract”). Because statutory mitigating circumstances are not inherently weightier un- der Florida law, Guardado has not shown a substantial likelihood that the result of his penalty phase would have been different if the statutory mitigating circumstances relating to his mental health were subbed in for the comparable non-statutory ones that the state trial court found and considered at sentencing. Second, Guardado compares his trial counsel’s performance in investigating and presenting mitigation evidence to Wiins v. Smith, 539 U.S. 510, 535–38 (2003), Rompilla v. Beard, 545 U.S. 374, 390–93 (2005), and Porter v. McCollum, 558 U.S. 30, 40–44 (2009), where the Supreme Court concluded that counsel’s deficient per- formance in investigating and presenting mitigation evidence USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 63 of 76

22-10957 Opinion of the Court 63

prejudiced the result of the penalty phase. But in each of those cases, the Supreme Court found Strickland prejudice because “counsel introduced little, if any, mitigating evidence at the original sentencing.” Thornell, 144 S. Ct. at 1314 (citing Wiins, 539 U.S. at 515, 534–35; Rompilla, 545 U.S. at 378, 393; Porter, 558 U.S. at 41). “[Guardado], by contrast, started with much more mitigation.” See id. For example, trial counsel presented evidence that shed light on Guardado’s childhood traumas. There was evidence showing that those traumas pushed Guardado to teenage substance abuse and a life in prison. Trial counsel also presented evidence showing that, although Guardado can work and contribute when not using crack cocaine, he struggled when readjusting to life outside prison and relapsed. And the penalty phase jury heard evidence—including expert testimony—describing Guardado’s substance abuse and how Guardado “was under considerable stress” before the murder. Based on the mitigation evidence that trial counsel pre- sented during the penalty phase, the state trial court found nine- teen mitigating circumstances and gave nine of them at least great or moderate weight, including how Guardado was traumatized by losing his brother, was sexually molested, and accepted responsibil- ity for the murder. So the mitigation case that Guardado’s trial counsel investigated and presented for his penalty phase is a far cry from the barebones mitigation cases that counsel presented in Wig- gins, Rompilla, and Porter. See id. Rompilla and Wiins are even further off the mark. In those cases, the Supreme Court “did not apply AEDPA deference to the USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 64 of 76

64 Opinion of the Court 22-10957

question of prejudice.” Gavin v. Comm’r, Ala. Dep’t of Corr., 40 F.4th 1247, 1269 (11th Cir. 2022) (quoting Cullen, 563 U.S. at 202). “Thus, as the Supreme Court has cautioned, . . . they ‘offer no guidance with respect to whether a state court has unreasonably determined that prejudice is lacking’—which is the question we must answer in this case.” Id. (quoting Cullen, 563 U.S. at 202). Here, the answer to that question, as to Guardado’s claim that trial counsel failed to adequately investigate and present miti- gation evidence during the penalty phase, is no. The Florida Su- preme Court did not unreasonably determine there was no reason- able probability of a different penalty phase outcome had trial counsel investigated and presented the lay witness testimony and letters, or the mental health testimony. And because that determi- nation was not unreasonable, the district court properly denied fed- eral habeas relief under AEDPA. See 28 U.S.C. § 2254(d)(1); Cal- houn, 92 F.4th at 1346; Sears, 73 F.4th at 1279–80. B. The Florida Supreme Court’s Prejudice Determination on Guardado’s Claim That Counsel Failed to Chal- lenge or Strike Three Biased Jurors As for his claim that trial counsel were ineffective because they failed to challenge for cause or peremptorily strike Jurors Pen- nington, Hall, and Cornelius, Guardado raises two arguments. First, he contends that the Florida Supreme Court unreasonably applied Strickland by “substitut[ing]” Carratelli’s heightened “actual bias” test for Strickland’s prejudice standard, which only requires a reasonable probability sufficient to undermine confidence in the USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 65 of 76

22-10957 Opinion of the Court 65

outcome. Second, he continues, because the Florida Supreme Court unreasonably applied Strickland’s prejudice standard, we should review de novo the Florida Supreme Court’s prejudice de- termination and find that the outcome of his penalty phase would’ve been different if Jurors Pennington, Hall, and Cornelius were challenged for cause or struck. We agree with Guardado that the Florida Supreme Court unreasonably applied Strickland’s prejudice standard by substituting Carratelli’s heightened actual bias test for the reasonable probability test. But we also conclude, applying the Strickland prejudice stand- ard de novo, Guardado has not shown that there was a substantial likelihood he’d receive a life sentence absent any error by trial coun- sel in failing to challenge for cause or peremptorily strike Jurors Pennington, Hall, and Cornelius. 1. The Florida Supreme Court’s Unreasonable Application of Strickland’s Prejudice Standard To be entitled to AEDPA deference, the Florida Supreme Court had to apply “[t]he correct standard for ineffective assistance of counsel [a]s set out in Strickland.” See Calhoun, 92 F.4th at 1347. Strickland’s prejudice standard for “juror selection claims” based on counsel’s “failure to challenge [a juror] either peremptorily or for cause” is the same as it would be for “any other Strickland claim.” Harvey v. Warden, Union Corr. Inst., 629 F.3d 1228, 1243 (11th Cir. 2011). So a court must consider, just as it must for any other Strick- land claim based on trial counsel’s deficient performance during the penalty phase, whether the petitioner has shown “a reasonable USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 66 of 76

66 Opinion of the Court 22-10957

probability that, absent [counsel’s] errors, the sentencer . . . would have concluded that the balance of aggravating and mitigating cir- cumstances did not warrant death.” Thornell, 144 S. Ct. at 1310 (al- terations in original) (quoting Strickland, 466 U.S. at 695). The Florida Supreme Court unreasonably applied this stand- ard. To explain why, we examine the standard that the Florida Su- preme Court did apply—Carratelli’s “actual bias” standard. See Guardado, 176 So. 3d at 899 (quoting Carratelli, 961 So. 2d at 324). In Carratelli, the Florida Supreme Court resolved a conflict among the state intermediate appellate courts about the application of Strickland’s prejudice prong to collateral claims of ineffective assis- 3 tance of counsel during juror selection. Carratelli, 961 So. 2d at 315. Before Carratelli, some Florida appellate courts required that the biased juror serve on the jury to establish prejudice. Id. Others applied a more lenient standard requiring the petitioner to show a reasonable doubt existed about the juror’s impartiality. Id. The Carratelli court began by observing that “the test for prejudice in conjunction with a collateral claim of ineffective assis- tance” is “much more strict” than the “test for prejudicial error in conjunction with a direct appeal.” Id. at 317–18 (quotation omit- ted). On the one hand, “the standard for obtaining a reversal upon

3 The specific claim in Carratelli was ineffective assistance in “failure to preserve a challenge to a potential juror.” 961 So. 2d at 315 (emphasis added). But, in its holding, the Florida Supreme Court set out a unified prejudice standard for claims of ineffective assistance in “failing to preserve or raise a cause challenge before a jury is sworn.” Id. at 327 (emphasis added). USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 67 of 76

22-10957 Opinion of the Court 67

the erroneous denial of a cause challenge is relatively lenient: a defendant need only show that an objectionable juror—whether or not actually biased—sat on the jury.” Id. at 320; see id. at 318–20 (“Where the record demonstrates a reasonable doubt about a ju- ror’s ability to be impartial, the trial court abuse[s] its discretion in denying [a] cause challenge.”). On the other hand, because “once a conviction has been affirmed on direct appeal a presumption of finality and legality attaches to the conviction and sentence,” a court’s “consideration of postconviction claims . . . is more restric- tive.” Id. at 320 (cleaned up). After citing this finality concern, the Carratelli court noted how Strickland’s prejudice standard requires “show[ing] that there is a reasonable probability that, but for coun- sel’s unprofessional errors, the result of the proceeding would have been different.” Id. (quoting Strickland, 466 U.S. at 694). But instead of stopping at the “reasonable probability” standard, as Strickland required, see 466 U.S. at 694, the Carratelli court went further. The court “determined that the prejudice standard applicable to [a] postconviction claim” regarding counsel’s failure to challenge a juror “is whether the juror [wa]s actually bi- ased”: In the context of the denial of challenges for cause, [Strickland] prejudice can be shown only where one who was actually biased against the defendant sat as a juror. We therefore hold that where a postconviction motion alleges that trial counsel was ineffective for failing to raise or preserve a cause challenge, the USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 68 of 76

68 Opinion of the Court 22-10957

defendant must demonstrate that a juror was actually biased.

Carratelli, 961 So. 2d at 324–25. “[A]ctual bias,” as Carratelli defined it, “mean[t] bias-in-fact that would prevent service as an impartial juror.” Id. at 324 (citation omitted). In Guardado’s case, the Florida Supreme Court applied the Carratelli test to affirm the denial of his claim, concluding he was not prejudiced by counsel’s failure to challenge for cause or strike Jurors Pennington, Hall, and Cornelius because he failed to show actual bias. Guardado, 176 So. 3d at 899. The problem, as Guardado argues, is that Carratelli’s actual bias-in-fact standard requires more than “a reasonable probability that . . . the result of the proceeding would have been different.” Strickland, 466 U.S. at 694. The reasonable probability standard only requires showing that counsel’s deficient performance “under- mine[s] confidence in the outcome.” Thornell, 144 S. Ct. at 1310 (quoting Cullen, 563 U.S. at 189). And that more lenient standard is the one a petitioner must satisfy for a claim that counsel failed to challenge for cause or peremptorily strike a juror. See Harvey, 629 F.3d at 1243 (“We evaluate juror selection claims as we would any other Strickland claim.”). Showing the juror was actually biased could be enough to establish prejudice, but Strickland doesn’t re- quire a showing of actual bias. Cf., e.g., Smith v. Gearinger, 888 F.2d 1334, 1337–39 (11th Cir. 1989) (reasoning “it may be sufficient” for a petitioner to satisfy the reasonable probability standard by show- ing counsel failed to challenge for cause “jurors . . . expected to sympathize with the victim”). USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 69 of 76

22-10957 Opinion of the Court 69

As we recently explained in Calhoun, “this type of error”— where “[t]he correct prejudice standard puts a lesser burden on the petitioner” but the state court holds the petitioner to a “stricter prejudice standard”—“ordinarily strips a state court decision of AEDPA deference.” Calhoun, 92 F.4th at 1347–49 (concluding a state supreme court’s application of Strickland wasn’t due AEDPA deference because the court “used [a] preponderance of the evi- dence/‘would have’ standard instead of the reasonable probabil- ity/confidence-in-the-outcome standard that Strickland man- dates”). Because the Florida Supreme Court’s application of Car- ratelli held Guardado to a stricter prejudice standard than Strick- land’s reasonable probability standard, the ordinary rule applies here just as it did in Calhoun. The state disagrees. It primarily relies on ten cases that, in its view, require a stricter prejudice standard—actual bias—for in- effective assistance claims that counsel failed to challenge for cause or peremptorily strike a juror. Having carefully reviewed the ten cases, we are not persuaded. First, the state argues that in two of our cases, Teasley v. War- den, Macon State Prison, 978 F.3d 1349 (11th Cir. 2020), and Owen v. Florida Department of Corrections, 686 F.3d 1181 (11th Cir. 2012), we read Strickland’s prejudice standard to require actual bias when the claim is that counsel failed to challenge for cause or peremptorily strike a juror. But we did not read Strickland’s prejudice standard to require actual bias in either case. Starting with Teasley, the peti- tioner raised two arguments on appeal: (a) the state habeas court USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 70 of 76

70 Opinion of the Court 22-10957

made an unreasonable factual determination that a juror wasn’t ac- tually biased; and (b) the state habeas court unreasonably applied Strickland in concluding the petitioner wasn’t prejudiced by coun- sel’s failure to raise the juror’s alleged bias on direct appeal. 978 F.3d at 1355–56, 1358. In resolving the actual bias argument, we only addressed whether the state habeas court’s finding of fact was “unreasonable . . . in light of the evidence.” See id. at 1355–58 (quoting 28 U.S.C. § 2254(d)(2)). We did not hold that the peti- tioner had to prove actual bias to prevail on an ineffective assistance claim that counsel failed to challenge or strike a juror. See id. In fact, in resolving the Teasley petitioner’s second argument that the state habeas court unreasonably applied Strickland, we did not use any “actual bias” gloss on the prejudice standard. See id. at 1358–59. Instead, we assessed the state habeas court’s no-prejudice determination against Strickland’s well-established reasonable probability test—concluding that, in light of the state habeas court’s factual finding, “there [wa]s no reason to believe that the re- sult of the appeal would have been different if appellate counsel had raised the issue.” Id. (emphasis added). That’s the Strickland rea- sonable probability standard. See 466 U.S. at 694. If anything, Tea- sley reaffirmed that the reasonable probability standard is the cor- rect one for establishing prejudice when the claim is that counsel failed to challenge or strike a juror. Owen is even less helpful to the state than Teasley. That’s because in Owen—as the state concedes—we expressly did “not de- cide whether the Carratelli actual-bias test for prejudice impose[d] USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 71 of 76

22-10957 Opinion of the Court 71

a higher burden or contradict[ed] the governing Strickland preju- dice standard.” 686 F.3d at 1201. Instead, we assumed that “the Carratelli prejudice test . . . w[as] contrary to Strickland” because it was clear that, under de novo review, the petitioner couldn’t prevail on his Strickland claim. Id. Second, the state argues that, under Weaver v. Massachusetts, 582 U.S. 286 (2017), courts must apply a higher prejudice standard on postconviction review than on direct appeal. As the state sees it, without Carratelli’s heightened prejudice standard, habeas peti- tioners asserting a Strickland claim that counsel failed to challenge or strike a juror would have a lighter burden than if they raised the juror issue on direct appeal. But the state’s premise—that the Strickland standard would be more favorable than the direct appeal standard without Car- ratelli—is wrong. On direct appeal, Florida courts presume preju- dice where a defendant claims that a biased prospective juror should’ve been removed. See, e.g., Busby v. State, 894 So. 2d 88, 92, 96–97 (Fla. 2004) (holding that the defendant, who exhausted his peremptory challenges, identified a biased juror, was denied a chal- lenge for cause, and was denied additional peremptory challenges, was prejudiced without analyzing the probability of a different out- come). By contrast, when asserting a postconviction Strickland claim, prejudice “cannot be presumed.” Teasley, 978 F.3d at 1358; see also Weaver, 582 U.S. at 300–01 (holding that prejudice isn’t pre- sumed when a defendant raises a violation of right to a public trial on an ineffective-assistance claim). Instead, the petitioner must USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 72 of 76

72 Opinion of the Court 22-10957

satisfy a “highly demanding” burden of showing that, but for coun- sel’s deficient performance, there was a reasonable probability of a different outcome. Shinn, 592 U.S. at 118 (citation omitted). Third, the state directs us to four cases that were decided on direct appeal from civil or criminal trials: McDonough Power Equip- ment, Inc. v. Greenwood, 464 U.S. 548 (1984); United States v. Tsarnaev, 595 U.S. 302 (2022); Skilling v. United States, 561 U.S. 358 (2010); and United States v. Martinez-Salazar, 528 U.S. 304 (2000). But because these four cases are not habeas cases, or even criminal cases, they are unhelpful in answering the question we have to answer— whether Strickland’s prejudice standard applies to a habeas claim 4 that counsel failed to challenge for cause or strike a juror. Cf. Wil- liams v. Singletary, 114 F.3d 177, 180–81 (11th Cir. 1997) (rejecting a habeas petitioner’s reliance on Zafiro v. United States, 506 U.S. 534 (1993), where it “was not a habeas case” and “[i]nstead, . . . involved a direct appeal of a federal criminal conviction”); Pierre v. Vannoy, 891 F.3d 224, 228 n.3 (5th Cir. 2018) (explaining that a habeas peti- tioner’s reliance on similar case law was “unhelpful”). Finally, the state cites three cases from other circuits “for the proposition that a showing of actual bias . . . is the typical standard

4 McDonough Power is particularly unhelpful. That case followed a civil prod- ucts liability trial, and the Supreme Court decided it nearly four months before adopting the two-part test for ineffective assistance claims in Strickland. See McDonough Power, 464 U.S. at 549; Strickland, 466 U.S. at 686–87. The earlier McDonough Power decision could not tell us much, if anything, about how to apply the later Strickland habeas standard. USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 73 of 76

22-10957 Opinion of the Court 73

for Strickland prejudice when the claim . . . involves a juror.” But, like the other cases the state relies on, these three do not address the question we must decide: whether a state court unreasonably applies Strickland’s prejudice standard when it adopts a test requiring actual bias rather than a reasonable probability of a different out- come. See Dickey v. Davis, 69 F.4th 624, 634 & n.3, 645–46 (9th Cir. 2023) (addressing the state habeas court’s “unexplained denial” of the petitioner’s claim that he was prejudiced by counsel’s strategy of selecting jurors predisposed to vote for death); Haight v. Jordan, 59 F.4th 817, 832–33 (6th Cir. 2023) (applying Strickland’s prejudice standard de novo because the state court “did not decide [that] is- sue on the merits”); Virgil v. Dretke, 446 F.3d 598, 611–14 (5th Cir. 2006) (applying Strickland’s “well-rehearsed” reasonable probability standard and concluding the petitioner satisfied it by showing “two persons, each expressly stating that they were unable to serve as fair and impartial jurors,” were seated on his jury). As we’ve explained, the answer to that question is yes—the Florida Supreme Court un- reasonably applied Strickland’s prejudice standard by substituting Carratelli’s heightened actual bias test for the reasonable probability test. 2. De Novo Review Because the Florida Supreme Court unreasonably applied Strickland’s prejudice prong, we must “undertake a de novo review of the record” and determine for ourselves whether Guardado was prejudiced by counsel’s failure to challenge for cause or perempto- rily strike Jurors Pennington, Hall, and Cornelius. See Debruce v. USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 74 of 76

74 Opinion of the Court 22-10957

Comm’r, Ala. Dep’t of Corr., 758 F.3d 1263, 1266 (11th Cir. 2014) (cleaned up). Based on our de novo review, we conclude that he was not. The three jurors confirmed during individual voir dire that, despite their connections to the case, they would be fair if seated on the jury. Starting with Juror Pennington, which Guardado de- scribes as his best case for prejudice, she was asked whether her business relationship with Ms. Malone would affect her ability to be fair. Juror Pennington testified that she “could be fair.” She gave the same answer a second time. And when asked again if she could “assure Mr. Guardado that [she] could be fair and make a fair and legal decision,” Juror Pennington said, “[y]es,” she could. Before the end of Juror Pennington’s individual voir dire, she was asked two more times whether her previous relationship with Ms. Malone would affect her ability to make a fair and impartial decision based on the law; she confirmed both times that she would make her decision based on the law. Her knowing Ms. Malone “was just . . . business.” Like Juror Pennington, Juror Hall assured the state trial court and the parties that he could be fair despite his connection to the case. During general voir dire, Juror Hall said that he knew three law enforcement officers involved in the investigation. Then, when asked during individual voir dire if he could “fairly weigh [their] testimony with that of other witnesses,” Juror Hall testified that he could. Mr. Gontarek later pressed Juror Hall on whether he would weigh Investigator Garrett’s testimony more than anyone USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 75 of 76

22-10957 Opinion of the Court 75

else’s, and Juror Hall again answered that he would not. Instead, as Juror Hall said, he would take the testimony “strictly on its value.” Finally, Juror Cornelius was asked if having family members who were victims of violent crime would affect his ability to serve as a juror in this case. He answered: “No. That doesn’t have any- thing to do with that.” Juror Cornelius emphasized that he was “small,” and he didn’t know many details about the murders of his great aunt and great uncle; the murders had taken place twenty- five years earlier. All three jurors, in other words, “insisted—repeatedly and unequivocally—that they could follow the law.” Owen, 686 F.3d at 1201 (reviewing the petitioner’s Strickland claim de novo and con- cluding he failed to show prejudice where three jurors similarly em- phasized their impartiality). Guardado has not met his burden to show that Jurors Pennington, Hall, and Cornelius could not be fair or that trial counsel’s acceptance of them as jurors otherwise prej- udiced the result of the penalty phase trial. For that reason, Guardado has not satisfied Strickland’s “highly demanding” stand- ard of showing a substantial likelihood that, absent any error by trial counsel in failing to challenge or strike the three jurors, he would’ve received a life sentence instead of death. See Thornell, 144 S. Ct. at 1310; Shinn, 592 U.S. at 118 (citation omitted). Guardado isn’t entitled to habeas relief on his second claim. USCA11 Case: 22-10957 Document: 56-1 Date Filed: 08/12/2024 Page: 76 of 76

76 Opinion of the Court 22-10957

IV. CONCLUSION Guardado committed a heinous crime. After a full and fair penalty phase trial, he was sentenced to death. We conclude that the Florida Supreme Court didn’t unreasonably apply Strickland to his claim that trial counsel were ineffective by failing to investigate and present mitigating evidence. And despite the Florida Supreme Court’s unreasonable application of Strickland to his claim that trial counsel were ineffective by failing to challenge for cause or per- emptorily strike three jurors, on de novo review, Guardado failed to show counsel’s performance prejudiced the result of his penalty phase trial. The district court’s denial of Guardado’s petition under section 2254 is affirmed. AFFIRMED.