Andrew R Allred v. Secretary, Florida Department of Corrections

USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 1 of 38 [DO NOT PUBLISH] In the United States Court of Appeals For the Eleventh Circuit ____________________ No. 22-12331 ____________________ ANDREW R. ALLRED, Petitioner-Appellant, versus SECRETARY, FLORIDA DEPARTMENT OF CORRECTIONS, ATTORNEY GENERAL, STATE OF FLORIDA, Respondents-Appellees. ____________________ Appeal from the United States District Court for the Middle District of Florida D.C. Docket No. 6:16-cv-00560-PGB-LHP ____________________ USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 2 of 38 2 Opinion of the Court 22-12331 Before JORDAN, JILL PRYOR, and BRASHER, Circuit Judges. PER CURIAM: Florida death-row prisoner Andrew Allred appeals the dis- trict court’s denial of his petition for a writ of habeas corpus under 28 U.S.C. § 2254. This appeal concerns his claim that his trial coun- sel was constitutionally ineffective for failing to ensure that he un- derwent a reasonably competent mental health evaluation for use during the penalty phase of his criminal trial. After a thorough re- view of the record and with the benefit of oral argument, we affirm the district court’s denial of the ineffective assistance of counsel claim. I. BACKGROUND Allred pled guilty before a Florida court to two counts of first-degree murder as well as armed burglary, aggravated battery with a firearm, and criminal mischief. Because under Florida law his convictions made him eligible for the death penalty, he was en- titled to a penalty-phase trial to determine his sentence. See Fla. Stat. § 921.141(1). He waived his right to a jury during the penalty phase. Thus, a judge was tasked with weighing aggravating and mitigating factors to decide whether to sentence Allred to death or life imprisonment. See id. § 921.141(3)(b). The trial court sentenced Allred to death for both murders. After his death sentence was up- held on direct appeal, Allred pursued postconviction relief in the Florida state courts and then in federal court. Below we describe the evidence presented at Allred’s penalty-phase trial, his USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 3 of 38 22-12331 Opinion of the Court 3 sentencing, his state postconviction proceedings, and his federal ha- beas proceedings. A. The Penalty Phase After Allred pled guilty, his case proceeded directly to a pen- alty-phase bench trial. During the penalty phase, the State intro- duced evidence of the following facts. Allred had a romantic relationship with one of the two vic- tims in this case, Tiffany Barwick. Before her murder, Allred and Barwick were living together at Allred’s parents’ home. They ended their relationship with a fight on August 25, 2007. The fight and breakup happened at the Allred home during Allred’s 21st birthday party, which was attended by 50 guests. In attendance was Michael Ruschak, the other murder victim, who was Allred’s clos- est male friend at the time. Unable to cope with the breakup, Allred began harassing Barwick. Days after the breakup, he purchased a handgun. Imme- diately after obtaining the weapon, he used pictures of Barwick for target practice. He emailed her a photo of the bullet-riddled pic- tures, which hung on the wall of his room. After the breakup, Allred learned that Barwick and Ruschak had begun a sexual relationship. The day before the murders, he messaged his friend Michael Siler, “I pretty much just need to start killing people.” Allred v. State (Allred I), 55 So. 3d 1267, 1272 (Fla. 2010). Siler replied, “[Y]ou’re just depressed.” USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 4 of 38 4 Opinion of the Court 22-12331 The next day, Allred sent threatening messages to and about Barwick and Ruschak. In Allred’s direct appeal, the Florida Su- preme Court summarized the messages as follows: In an instant message chat with Siler in the morning, Allred stated, “I’m pretty much gonna kill him . . . Ru- schak . . . and her.” In an electronic conversation with victim Ruschak on that same day, Allred told him, “If [I] see you again, [I] will kill you, and yes that is a threat.” Finally, Allred and Barwick engaged in a heated and lengthy computer exchange on the day of the murder. Allred informed Barwick that he had hacked into her computer, changed the passwords, deleted files, and sent emails to people on her contacts list. He also transferred all of the funds in her bank account to pay her credit card debt. Calling her a “whore” because of her relationship with Ruschak, Allred said he could not forgive her for that and threatened, “[I]f [] I ever see [Ruschak] again I will kill him.” Allred I, 55 So. 3d at 1272–73 (alterations in original). That evening, Allred told Ruschak that he was coming over to Ruschak’s house. Allred took his handgun and drove to the home of Eric Roberts, where Ruschak lived and Barwick lived tem- porarily. Allred expected to find both Barwick and Ruschak at the house. Several guests who were acquainted with Allred were also at the house. Ruschak warned them of Allred’s impending arrival. According to one guest, Kathryn Cochran, when Barwick learned that Allred was coming over, she went into “panic mode.” USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 5 of 38 22-12331 Opinion of the Court 5 A few minutes after Ruschak’s announcement, Allred pulled up to the house. He repeatedly rammed his truck into Barwick’s car, which was parked outside. Guests inside the house heard the collisions. Cochran described hearing a noise like a mortar blast. Allred exited his truck and attempted to enter the home, but he was unable to enter because Ruschak had locked the front door in anticipation of his arrival. Allred banged loudly on the front door, yelling “[l]et me in,” but no one opened it. Allred I, 55 So. 3d at 1273. Allred then walked to back of the house and banged on a sliding glass door that opened to the living room, where some of the guests had gathered. Barwick ran away to hide. When no one let him in, Allred shot through the door. He walked through the broken glass into the house, gun in hand. The occupants scattered. Allred noticed Ruschak peering from the kitchen and pur- sued him, shooting him four times. Ruschak was killed instantly. Allred continued to the bathroom, where he found Barwick hiding in the bathtub. He shot her six times. She, too, died instantly. As Allred moved through the house, Roberts grabbed him, trying to stop him. Allred shot Roberts in the leg and escaped his grasp. Allred left the crime scene and drove home. He called 911 and reported that he had killed two people. He threatened to com- mit suicide. When law enforcement officers arrived at Allred’s home, he was standing at the end of the driveway with a handgun on the ground next to him. He told his arresting officer, “I’m the guy you’re looking for.” Id. at 1274. After he was secured, Allred USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 6 of 38 6 Opinion of the Court 22-12331 asked “if the people were dead,” and in the patrol car he said, “I knew I killed someone, I shot fourteen times.” Id. After his arrest, detectives interviewed Allred about the murders. In the interview, he demanded to know “what happened to . . . the people that got shot” before confessing to shooting Ru- schak and Barwick. He confessed to the events of the shooting sub- stantially as described above. 1 He also confessed to using Barwick’s pictures for target practice. He told detectives that he killed Ru- schak because Ruschak was an “asshole” who sided with Barwick in the couple’s breakup. Id. at 1275. But Allred denied planning the murders, buying the gun to facilitate the murders, and going to Roberts’s house that night with the intent to shoot Barwick and Ruschak. He said that he drove to the scene only to ram Barwick’s car. Contrary to Allred’s denials, the State’s evidence indicated that the murders were premeditated. The State pointed to digital evidence of the threats he sent to and about Ruschak and Barwick leading up to the murders. Cochran, who had a conversation with Barwick about the threats at Roberts’s gathering, testified that Bar- wick was so “freaked out” about Allred’s communications that she called the police when she heard Allred was coming to the house. 1 Although Allred described the events of the shooting in the police interview, in postconviction proceedings, his mental health expert, Dr. Glenn Caddy, opined that Allred’s reporting of the murders was imperfect or “fragmented.” We discuss Caddy’s opinion in more detail in Section B below. USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 7 of 38 22-12331 Opinion of the Court 7 The eyewitnesses who had been at Roberts’s home testified about Allred’s deliberate pursuit of Ruschak at the scene. They de- scribed Allred as singularly focused when, after shooting his way through the door, he silently moved past the guests in the living room to shoot Ruschak in the kitchen. Roberts testified to Allred’s cool affect during the murders, characterizing Allred’s tone of voice during their struggle as “somewhat calm” or “louder than calm.” Corroborating the eyewitness accounts, Allred confessed to detec- tives that he was “specifically looking for Michael” when he walked into Roberts’s kitchen. To prove that Barwick’s murder was especially heinous, atrocious, or cruel, an aggravating factor under state law, the State played the 911 call that Barwick made while hiding from Allred in the bathroom to illustrate the fear she experienced immediately be- fore her death. On the call, she screamed as she heard the gunshots that killed Ruschak and continued screaming as she was fatally shot moments later. For the defense in the penalty phase, Allred’s trial counsel, Timothy Caudill, called seven witnesses in mitigation, mostly Allred’s family members and former teachers. Caudill had difficulty organizing a mitigation case because Allred was sometimes unco- operative and generally reluctant to participate in mitigation ef- forts. Allred waived his right to be present and did not attend his penalty-phase trial. Caudill presented evidence that Allred was socially and emo- tionally developmentally delayed and that his family life was USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 8 of 38 8 Opinion of the Court 22-12331 difficult. He argued that Allred’s developmental delays meant that he should be considered younger than 21 for sentencing purposes. The evidence about Allred’s developmental delays included testimony from his mother, Tora Allred, about his childhood. She noticed changes in Allred’s behavior beginning around the age of five or six, when he suddenly became hyper and emotional. Con- cerned about his behavior, Tora took him to a pediatrician who attributed his behavior to possible sexual abuse and referred him to a psychiatrist. The psychiatrist instead diagnosed Allred with a “well-defined tic disorder” (Allred licked his hand and rubbed his eye repetitively) and attention deficit hyperactivity disorder (“ADHD”). Id. The psychiatrist prescribed Allred medication to treat the ADHD. The pediatrician’s suggestion of sexual abuse was never sub- stantiated. Although there was alleged sexual abuse in Allred’s fam- ily—Allred’s older cousin filed a report that their grandfather and great uncle had sexually abused him—Allred never claimed to have been sexually abused. Allred’s grandfather testified that he and his grandson had a good relationship and that Allred had a good child- hood. But other testimony suggested that, in many ways, Allred had a troubled childhood, at school and home. Caudill argued that Allred’s developmental difficulties impacted his education and so- cial growth. To illustrate, Allred’s teachers testified that he did not interact with his peers or participate in class. His third-grade teacher testified that he was “withdrawn and ‘standoffish.’” Id. at USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 9 of 38 22-12331 Opinion of the Court 9 1276. Tora testified that around third grade, school progress re- ports suggested that her son might have a learning disability. Yet subsequent school testing revealed that Allred had a high IQ. As a result, he enrolled in gifted education classes in middle school. He did not adjust well to the program, however. Both Allred’s middle school teacher and Tora testified that he became even less social— more of a “loner”—throughout his middle school years. He even- tually stopped taking gifted classes. About Allred’s home life, his parents testified that he was ex- posed to alcohol abuse and domestic violence throughout child- hood. Allred’s father, David, and Tora agreed that David had an alcohol problem. They testified that they were physically violent toward each other. They described one violent incident involving Allred. One evening, David picked up a shotgun—Tora said he threatened to shoot himself; David said he was shooting at a tree to blow off steam—and assaulted Tora as she tried to take the weapon away. Allred, who was about 12 at the time, witnessed the assault and called the police. David was arrested. There was no tes- timony that Allred himself was abused. Despite Allred’s difficult childhood and developmental defi- cits, Tora admitted on cross-examination that by the time of the murders, her son was self-sufficient. Although he dropped out of high school in eleventh grade, he later earned his diploma and a two-year associate’s degree in accounting. Until the day of the mur- ders, when his employment was terminated, he worked full time, teaching the use of software. He lived independently in a large USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 10 of 38 10 Opinion of the Court 22-12331 room, an addition to the family home that only he could access. For three or four months, Barwick lived there too, and, according to Tora, the couple was happy. Because of Allred’s emotional difficulty handling his breakup with Barwick, Caudill argued that Allred was suffering from extreme emotional disturbance at the time of the murders. In support, he presented Tora’s testimony that Allred became even more withdrawn after Barwick moved out. David testified that he was worried his son was suicidal after the breakup. Caudill’s arguments in the penalty phase about mitigating circumstances focused on Allred’s social and emotional develop- mental delays, difficult childhood, and emotional distress, yet he called no mental health expert to testify. Before trial, Caudill had identified Dr. Deborah Day, a psychologist, as a testifying expert. But Caudill decided not to call Day as a witness. 2 After the penalty-phase trial concluded, the trial court sen- tenced Allred to death. The court found three statutory aggravat- ing factors for each murder. For Ruschak’s murder, the court con- cluded that (1) the murder was cold, calculated, and premeditated; (2) the murder was committed while Allred was engaged in a bur- glary; and (3) Allred had a prior capital conviction for Barwick’s contemporaneous murder. For Barwick’s murder, the court 2 During postconviction proceedings, Caudill testified about his decision not to call Day. We discuss his testimony in Section B below. USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 11 of 38 22-12331 Opinion of the Court 11 concluded that (1) the murder was especially heinous, atrocious, or cruel; (2) the murder was cold, calculated, and premeditated; and (3) Allred had a prior capital conviction for Ruschak’s contempora- neous murder. The court assigned great weight to all these aggra- vators except the contemporaneous burglary. The court found no statutory mitigating factors and four nonstatutory mitigating factors, all of which it assigned little or moderate weight. As nonstatutory mitigation, the court weighed the facts that Allred: (1) accepted responsibility by entering guilty pleas; (2) cooperated with law enforcement; (3) suffered from an emotional disturbance after his breakup with Barwick; and (4) had developmental problems at a young age that impacted his educa- tional and social development, but not his later education.3 3 The court considered and rejected several other mitigating factors. It deter- mined that Allred’s emotional and developmental age at the time of the crime corresponded to his chronological age of 21 and therefore declined to weigh a younger emotional age as a mitigating circumstance. See Fla. Stat. § 921.141(7)(g) (identifying as a mitigating factor “the age of the defendant at the time of the crime”). And although the court recognized that Allred suf- fered from an emotional disturbance after the breakup, it did not apply the statutory mitigator for extreme emotional disturbance. See id. § 921.141(7)(b) (identifying as a mitigating factor that “[t]he capital felony was committed while the defendant was under the influence of extreme . . . emotional disturb- ance”). The court found that the “careful thought and planning of the mur- ders” negated such application. The court also stated that it was “not con- vinced that [Allred] was incapable of conforming his conduct to the require- ments of law.” See id. § 921.141(7)(f) (identifying as a mitigating factor that “[t]he capacity of the defendant to appreciate the criminality of his or her USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 12 of 38 12 Opinion of the Court 22-12331 Weighing the aggravating and mitigating circumstances, the court found that the aggravating circumstances far outweighed the mitigating circumstances. The court sentenced Allred to death for the murders of both Ruschak and Barwick. B. State Postconviction Proceedings The Florida Supreme Court affirmed Allred’s death sentence on direct appeal. See Allred I, 55 So. 3d at 1284. Allred then initiated state postconviction proceedings, in which, as relevant here, he claimed that Caudill, his penalty-phase trial counsel, rendered con- stitutionally ineffective assistance under Strickland v. Washington, 466 U.S. 668 (1984), by failing to ensure that he received a reasona- bly competent mental health evaluation. 4 Allred alleged that Cau- dill performed deficiently in three respects. First, he alleged that Caudill unreasonably relied on the opinion of Day, the mental health expert Caudill retained. Allred alleged that Caudill mistak- enly believed Day would testify that Allred had antisocial person- ality disorder, sociopathy, or psychopathy and relied on this under- standing in deciding to withdraw her testimony in the penalty phase. Second, Allred alleged that Caudill failed to conduct an ade- quate background investigation and therefore failed to adequately prepare Day. Third, Allred alleged that Caudill failed to present the conduct or to conform his or her conduct to the requirements of law was sub- stantially impaired”). 4 In postconviction proceedings, Allred raised other challenges to his convic- tions and sentence. We limit our discussion to Allred’s ineffective assistance claim regarding the mental health evaluation, the only claim in this appeal. USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 13 of 38 22-12331 Opinion of the Court 13 testimony of a mental health expert as mitigation in the penalty phase. Allred alleged that Caudill was obligated to pursue other ex- perts, specifically those that were “tailored to the needs of the case” and had expertise with Allred’s developmental issues. This deficient performance, Allred alleged, prejudiced his case such that had counsel performed effectively, there was a rea- sonable probability that he would not have received a death sen- tence. See Strickland, 466 U.S. at 687 (explaining that, to establish ineffective assistance of counsel, the defendant “must show that counsel’s performance was deficient” and “that the deficient per- formance prejudiced the defense”). 1. The Postconviction Evidentiary Hearing The postconviction trial court conducted an evidentiary hearing on several of Allred’s claims, including the claim relevant here: that Caudill rendered ineffective assistance by failing to en- sure that Allred underwent a reasonably competent mental health evaluation for use in the penalty phase. At the hearing, Allred’s postconviction counsel introduced two types of mental health evi- dence that Allred contended supported his theories of deficient per- formance. First, counsel introduced evidence of possibly mitigating mental health conditions—dissociation and autism spectrum disor- der—that Allred argued Caudill should have developed. Second, counsel introduced evidence to undermine Day’s purported diag- nosis of antisocial personality disorder, a potentially aggravating mental health condition. USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 14 of 38 14 Opinion of the Court 22-12331 Allred’s counsel called two licensed psychologists. Both eval- uated Allred in preparation for the evidentiary hearing. The first psychologist, Dr. Glenn Caddy, testified that Allred was in a disso- ciative state during the murders. This meant that Allred was “dis- connect[ed] from a clear understanding” of his actions. Caddy opined that the dissociative state was triggered by Allred’s pro- found “ego disintegration”—his sense of degradation and emo- tional distress—following his public breakup with Barwick and the discovery that she was involved with Ruschak. Caddy opined that, because of the dissociation and Allred’s deteriorating emotional state, Allred satisfied a statutory mitigator: he had diminished ca- pacity at the time of the crime. See Fla. Stat. § 921.141(7)(f) (listing as a mitigating factor that “[t]he capacity of the defendant to appre- ciate the criminality of his or her conduct or to conform his or her conduct to the requirements of law was substantially impaired”). Caddy testified that Allred’s postconviction reporting of the murders showed a fragmented memory of what happened inside Roberts’s house. In Caddy’s opinion, Allred’s postarrest interview indicated fragmentation. To Caddy, the disconnect between Allred’s behavior and his inability to explain it indicated dissocia- tion. But Caddy conceded on cross-examination that a fragmented memory could also have been Allred’s response to the trauma of the shootings. He agreed it was possible that Allred’s memory frag- mented under the stress of the violence he inflicted rather than be- cause of dissociation before the shootings. Thus, Caddy could not say whether the symptoms of dissociation he observed actually “set[] the stage” for the shooting. USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 15 of 38 22-12331 Opinion of the Court 15 Allred’s postconviction counsel also called Dr. Gary Ge- ffken, an autism specialist, who opined that Allred had a high-func- tioning form of autism spectrum disorder or “at the very least[,] pervasive developmental disorder.” Geffken based his opinion on social and emotional deficiencies he observed upon meeting Allred and assessing his background. Geffken opined that Allred was emo- tionally underdeveloped: he lacked sympathy, empathy, and the ability to discuss emotions like jealousy or embarrassment. Geffken also opined that Allred was socially underdeveloped. When Ge- ffken met with Allred, he found him to be intentionally solitary, cutting himself off from human interaction. Records established that Allred had also been this way in childhood and had difficulties interacting with peers. In Geffken’s opinion, Allred also exhibited other characteristics of autism spectrum disorder: Allred adapted poorly to change and had exhibited unusual repetitive behaviors and tics in childhood. Geffken concluded that Allred’s autism was a plausible ex- planation for why he murdered Ruschak and Barwick. Geffken pos- ited that when Allred and Barwick broke up, Allred was unable to cope with the loss because he did not have the social and emotional skills of a normal adult his age. Allred’s actions stemmed from this inability to process his emotions. Geffken opined that the for- mation of intense, consuming attachments—and an inability to cope with their loss—was characteristic of behavior across the au- tism spectrum. But Geffken acknowledged on cross-examination that the type of violence Allred carried out was “clearly” atypical of USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 16 of 38 16 Opinion of the Court 22-12331 those on the autism spectrum. Geffken unequivocally stated that Allred was not remorseful and that his actions had been deliberate. But Caddy thought that Allred had recently demonstrated an “intellectual” remorse for what he had done. Anticipating the testimony of the State’s witnesses, including withdrawn penalty- phase mental health expert Day, Caddy unequivocally opined that Allred did not have antisocial personality disorder. Allred did not meet the diagnostic criteria for antisocial personality disorder be- cause he had no history of a childhood conduct disorder. And Caddy observed that Allred showed no history of failing to corre- spond to social norms and noted that before the murders he had no history of getting into significant trouble. Caddy concluded that Allred displayed empathy that was contraindicative of antisocial personality disorder. Postconviction counsel then called Allred’s trial counsel, Caudill, attempting to show that Caudill unreasonably relied on this contraindicated diagnosis of antisocial personality disorder to justify his decision to not present a mental health expert in the pen- alty phase. Caudill’s reliance on Day’s purported antisocial person- ality disorder diagnosis, postconviction counsel argued, prevented him from adequately investigating more favorable mental health diagnoses like dissociation and autism spectrum disorder. Caudill testified to his understanding of Day’s evaluation of Allred, their conversations about her potential testimony as to an- tisocial personality disorder and other aggravators, and his decision not to call Day or another mental health expert. He testified that USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 17 of 38 22-12331 Opinion of the Court 17 he retained Day as a mental health expert before the penalty phase. After Day evaluated Allred, in the time leading up to the trial, Cau- dill had multiple conversations with her about her evaluation. Dur- ing one of these conversations, Day relayed that she had nothing to offer in mitigation in Allred’s case. Caudill understood from the conversation that if Day testified and had to offer a diagnosis, the only diagnosis she could offer would be antisocial personality dis- order—or worse, that Allred was a psychopath or sociopath. Put differently, her findings were consistent with a diagnosis of antiso- cial personality disorder. 5 And though Caudill understood that an antisocial personality disorder diagnosis must be premised on a childhood conduct disorder, he believed there was evidence of such a disorder in Allred’s past, such that an antisocial personality disorder diagnosis was not contraindicated. According to Caudill, in his experience, antisocial personal- ity disorder, sociopathy, and psychopathy were not mitigating. He further believed that calling Day, or any mental health expert, would elicit damaging testimony on Allred’s reporting of the mur- ders. Caudill viewed Allred’s explanation for the murders—that the victims were deserving—as aggravating. He also thought that men- tal health expert testimony as to Allred’s animosity toward the 5 A contemporaneous memo written by Caudill’s cocounsel stated that Day had “concluded that Mr. Allred is a sociopath or psychopath.” This statement was not accurate, however. Day never diagnosed Allred. Both the postconvic- tion trial court and the Florida Supreme Court determined that despite the memo’s overstatement of Day’s conclusion, Caudill did not misunderstand her evaluation. The record does not contradict this finding of fact. USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 18 of 38 18 Opinion of the Court 22-12331 victims, expressed in Allred’s own words, would be more impactful than the State’s comparable evidence. For all these reasons, he de- cided to withdraw Day as a testifying witness. Caudill testified that he never had cause to question Day’s conclusions, nor did he consider seeking a second opinion or ex- pert. He did not believe that another expert would have come to a different conclusion or been able to provide mental health mitiga- tion. He was also concerned about delaying the penalty phase to pursue other experts, given Allred’s hostility toward developing a mitigation case. The State called Day at the postconviction hearing. Through Day, the State aimed to establish that Caudill had correctly con- cluded that Day’s expert testimony would have been more aggra- vating than mitigating. Therefore, the State argued, Caudill made a reasonable strategic decision to forgo the presentation of mental health evidence in the penalty phase. Day testified that, at Caudill’s behest she, along with two psychologists in her employ, had conducted a comprehensive eval- uation of Allred before his penalty-phase trial. Based on her evalu- ation, Day opined at the postconviction hearing that Allred had “features” of antisocial personality disorder: he met some, but not all, of the criteria for the disorder described in the Diagnostic and Statistical Manual of Mental Disorders. She determined that Allred displayed an inability to have empathy, had interpersonal difficul- ties, had difficulties obeying societal norms, and evidenced a reck- less disregard for others. But she also concluded that he did not USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 19 of 38 22-12331 Opinion of the Court 19 have a childhood or adolescent conduct disorder; she saw evidence of antisocial traits emerging only past the age of 15. Because such a conduct disorder is a required criterion for a diagnosis of antisocial personality disorder, Day could not formally diagnose him. According to Day, Allred demonstrated features of psychop- athy, which she described as a “descriptor[] of behavior” rather than a separate diagnosis. She opined that many of Allred’s antiso- cial traits—deceptiveness, irresponsibility, poor behavioral control, lack of empathy, and disregard for others—were also characteristic of psychopathy. Allred’s clinically elevated scores on the psycho- pathic deviate scale of a personality test she administered, the Min- nesota Multiphasic Personality Inventory 2 (“MMPI-2”), were con- sistent with her impressions. 6 Day testified consistently with Caudill’s impressions of her conclusions. If she had testified at the penalty phase, she said, her testimony would have been aggravating rather than mitigating. She could offer little in the way of statutory mitigation, and her opinions would have supported aggravators, including that the murders were cold, calculated, and premeditated. She opined that Allred’s behavior leading up to the murders evidenced premedita- tion and observed that he displayed a lack of remorse. 6 Day also testified that Allred had clinically elevated scores for depression and trauma. She did not explain how those scores related to her impressions of Allred’s traits, background, or behavior, though she did report that Allred ex- perienced situational depression in jail. USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 20 of 38 20 Opinion of the Court 22-12331 The State called a second mental health expert, psychiatrist Dr. Jeffrey Danziger, who evaluated Allred and reviewed the other experts’ conclusions. 7 Danziger refuted Caddy’s diagnosis of disso- ciation and Geffken’s diagnosis of autism spectrum disorder. Like Caddy, Danziger opined that Allred did not have anti- social personality disorder. Aside from the murders and the sur- rounding events, Danziger found that Allred lacked traits that would support an antisocial personality disorder diagnosis. But he “agree[d] wholeheartedly” with Day’s evaluation and postconvic- tion testimony and opined that, accounting for the murders, Allred displayed features of the disorder. Danziger also agreed with Caddy and Day that Allred would not meet the criteria for a “full-scale diagnosis” of antisocial personality disorder, in part because he had no conduct disorder before the age of 15. Danziger saw no perva- sive pattern of criminal conduct in Allred’s behavior. He opined that Allred was not a sociopath or a psychopath. Danziger disagreed entirely with Caddy’s finding that Allred was in a dissociative emotional state during the murders. He opined that the evidence of Allred’s actions surrounding the mur- ders suggested premeditation rather than confusion, disorienta- tion, or a lack of control. He characterized Allred’s imperfect 7 Danziger had met previously with Allred, at Caudill’s request, shortly after Allred’s arrest in 2007. There was conflicting testimony as to whether this pre- trial evaluation was intended to evaluate Allred’s competency only or also to assess the possibility of an insanity defense. Danziger met with Allred again in 2013 to evaluate him at the State’s behest. USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 21 of 38 22-12331 Opinion of the Court 21 memory as “a very thin reading to base a diagnosis of dissociative disorder on.” Danziger also disagreed with Geffken’s diagnosis of autism spectrum disorder. Although Danziger believed that Allred had so- cial deficits, he explained that to be diagnosed with autism spec- trum disorder, a person must have a restrictive repetitive pattern of behavior, interest, or activities. He did not find that Allred pre- sented such a pattern, though he failed to consider Allred’s child- hood tic disorder when making this conclusion. Danziger further opined that an autism spectrum disorder diagnosis would not be mitigating even if correct. If Allred were autistic, he still would have been able to appreciate the wrongfulness of his conduct and conform his conduct to law. And Danziger saw no evidence that Allred was experiencing extreme emotional distress. Danziger summarized that had he testified at trial, he could not have offered any testimony in mitigation other than the evi- dence that had been offered in the penalty phase. He noted the ev- idence of Allred’s exposure to domestic violence and his historic diagnoses of ADHD and tic disorder, but the court had acknowl- edged those facts at sentencing. 2. The State Postconviction Trial Court’s Order The postconviction trial court denied Allred’s claim that trial counsel rendered constitutionally ineffective assistance in failing to ensure that Allred received a reasonably competent mental health evaluation. It concluded that Allred failed to show that Caudill USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 22 of 38 22 Opinion of the Court 22-12331 performed deficiently or that any deficiency prejudiced him. See Strickland, 466 U.S. at 687. The postconviction trial court assessed Caudill’s perfor- mance as to each of the three deficiencies Allred alleged: that Cau- dill unreasonably relied on Day’s antisocial personality disorder di- agnosis, that Caudill failed to conduct an adequate background in- vestigation, and that Caudill failed to seek out additional mental health experts to testify in the penalty phase. First, the court concluded that Caudill reasonably relied on his understanding of Day’s opinion. Caudill correctly understood that “Day had not diagnosed the Defendant with anti-social per- sonality disorder, but rather, she indicated that if she was forced to give a diagnosis the only diagnosis she could offer was anti-social personality disorder.” State v. Allred, No. 07-4890, 2013 WL 12450438, at *3 (Fla. Cir. Ct. 2013). Second, the court concluded Allred offered no evidence that Caudill conducted an inadequate background investigation or in- adequately prepared Day. Allred’s argument to the contrary had no support. Consequently, Caudill was entitled to rely on Day’s opinion. Given Day’s opinion that Allred had antisocial traits and her indi- cation that she could not provide helpful testimony, the court con- cluded that Caudill’s choice not to present Day’s testimony was a “reasonable strategic decision” rather than deficient performance. Id. at *4. It was also a “successful” decision in that it precluded the USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 23 of 38 22-12331 Opinion of the Court 23 introduction of additional damaging evidence as to Allred’s pre- meditation and lack of remorse. Id. Because the court concluded that Caudill’s reliance on Day’s opinion was reasonable, it followed that he “was not required to continue searching for an expert who would give a more favorable assessment of [Allred’s] mental status.” Id. (internal quotation marks omitted). The court therefore rejected Allred’s third argu- ment, that Caudill was obligated to seek another mental health ex- pert to testify at the penalty phase, concluding that Caudill’s failure to do so did not amount to deficient performance. Despite finding no deficient performance, the postconvic- tion trial court proceeded to analyze whether Caudill’s failure to present a mental health expert prejudiced Allred. The court looked to whether the new mental health evidence adduced in postconvic- tion—the diagnoses provided by Caddy and Geffken—would have altered the sentencing outcome. The court found the testimony of both experts not credible. It found that Caddy’s diagnosis of dissociation at the time of the murders was unsupported, in part, because the diagnosis relied on Allred’s “alleged fragmented memory” of the murders, yet Caddy “acknowledged that the Defendant’s fragmented memory could be the result of a physiological reaction to the trauma of the shooting.” Id. at *5. And the court credited Danziger’s contrary opinion that there was no evidence of Allred’s having been in a dis- sociative state. USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 24 of 38 24 Opinion of the Court 22-12331 The court also found not credible Geffken’s testimony that Allred had autism spectrum disorder. It credited Danziger’s testi- mony that, to be diagnosed with autism spectrum disorder, one must display a restrictive repetitive pattern of behavior. It found a “lack of evidence” that Allred had restrictive repetitive behavior patterns. Id. Because it found Caddy’s and Geffken’s testimony not cred- ible and credited Danziger’s, the court concluded that Allred failed to establish that he was in a dissociative state at the time of the murders or that he had autism spectrum disorder. But even if Allred had credibly demonstrated those diagnoses, the court found, their introduction would not have impacted the penalty phase’s outcome because his experts’ testimony failed to show that Allred’s capacity was, in fact, diminished, and that he lacked premeditated intent. 3. The Florida Supreme Court’s Decision The Florida Supreme Court affirmed the postconviction trial court’s denial of Allred’s ineffective assistance claim. Allred v. State (Allred II), 186 So. 3d 530, 539 (Fla. 2016). It ruled that Allred “demonstrated neither deficiency nor prejudice” as to Caudill’s fail- ure to obtain a competent mental health evaluation during the pen- alty phase. Id. at 539. Like the postconviction trial court, the Florida Supreme Court evaluated Caudill’s performance as to each of the three defi- ciencies that Allred alleged. First, the court concluded that Caudill reasonably relied on Day’s opinion. Id. at 536–37. It held that Day USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 25 of 38 22-12331 Opinion of the Court 25 had “clearly indicated,” and Caudill correctly understood, “that her testimony would be more aggravating than mitigating.” Id. at 537. Although Day “did not reach the [antisocial personality disorder] diagnosis, her testimony that Allred met all but one of the factors essential to the diagnosis would not have been mitigating in nature. Neither would her testimony that he had some of the traits of a sociopath and a psychopath.” Id. Further, by withdrawing Day’s testimony, “Caudill kept out testimony about Allred’s lack of em- pathy or remorse.” Id. His reliance on her “professional assess- ment” was not unreasonable, and his decision to withdraw her tes- timony was a reasonable strategic choice. Id. Second, the court concluded that Caudill’s background in- vestigation was not deficient. Id. at 537–38. It determined that Cau- dill conducted an adequate background investigation to prepare his mitigation case in the penalty phase; the court cited his efforts to track down records, witnesses, and evidence that Allred was possi- bly sexually abused by his grandfather. Id. at 538. Third, the Florida Supreme Court determined that Caudill’s decision not to call Day “did not require the continued search for a more favorable mental health opinion.” Id. Caudill’s failure to find and present a favorable mental health expert in the penalty phase did not constitute deficient performance. Id. In addition, Caddy and Geffken were not credible, and the evidence presented at the post- conviction hearing “established neither that [Allred] was in a disso- ciative state nor that he suffered from an autism spectrum disor- der.” Id. at 538–39. Thus, Allred failed to present evidence of a USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 26 of 38 26 Opinion of the Court 22-12331 favorable opinion. But even if he had, the court stated that “secur- ing a more favorable expert opinion” would not “undermine the sufficiency” of Day’s opinion, on which Caudill was entitled to rely. Id. at 539. C. Federal Habeas Proceedings Allred then filed a petition for a writ of habeas corpus in fed- eral district court, raising several claims, including a claim that trial counsel rendered ineffective assistance by failing to obtain a rea- sonably competent mental health evaluation. The district court de- nied Allred relief and then denied him a certificate of appealability. We granted Allred a certificate of appealability on his claim that trial counsel was constitutionally ineffective in two ways, for failing to (1) conduct a sufficient background investigation and (2) ensure a reasonably competent mental health evaluation for Allred’s pen- alty phase. II. STANDARDS OF REVIEW “When reviewing a district court’s grant or denial of habeas relief, we review questions of law and mixed questions of law and fact de novo, and findings of fact for clear error.” Reaves v. Sec’y, Fla. Dep’t of Corr., 717 F.3d 886, 899 (11th Cir. 2013) (internal quotation marks omitted). An ineffective assistance of counsel claim “pre- sents a mixed question of law and fact that we review de novo.” Pope v. Sec’y, Fla. Dep’t of Corr., 752 F.3d 1254, 1261 (11th Cir. 2014). The Antiterrorism and Effective Death Penalty Act of 1996 (“AEDPA”) governs our review of federal habeas petitions. “AEDPA prescribes a highly deferential framework for evaluating USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 27 of 38 22-12331 Opinion of the Court 27 issues previously decided in state court.” Sears v. Warden GDCP, 73 F.4th 1269, 1279 (11th Cir. 2023). AEDPA bars federal courts from granting habeas relief to a petitioner on a claim that was “ad- judicated on the merits in [s]tate court” unless the decision (1) “was contrary to, or involved an unreasonable application of, clearly es- tablished Federal law, as determined by the Supreme Court of the United States,” or (2) “was based on an unreasonable determina- tion of the facts in light of the evidence presented in the [s]tate court proceeding.” 28 U.S.C. § 2254(d). A state-court decision is “contrary to” clearly established law if the court “applie[d] a rule that contradicts the governing law” set forth by the Supreme Court or the state court confronted facts that were “materially indistinguishable” from Supreme Court prece- dent but arrived at a different result. Williams v. Taylor, 529 U.S. 362, 405–06 (2000). To meet the “unreasonable application” stand- ard, “a prisoner must show far more than that the state court’s de- cision was merely wrong or even clear error.” Shinn v. Kayer, 592 U.S. 111, 118 (2020) (internal quotation marks omitted). Rather, the decision must be “so obviously wrong that its error lies beyond any possibility for fair minded disagreement.” Id. (internal quotation marks omitted). This standard is “difficult to meet and . . . demands that state-court decisions be given the benefit of the doubt.” Raulerson v. Warden, 928 F.3d 987, 996 (11th Cir. 2019) (internal quotation marks omitted). A federal habeas court must defer to a state court’s determi- nation of the facts unless the state court decision “was based on an USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 28 of 38 28 Opinion of the Court 22-12331 unreasonable determination of the facts in light of the evidence presented in the [s]tate court proceeding.” 28 U.S.C. § 2254(d)(2). We are required to give state courts “substantial deference” under § 2254(d)(2). Brumfield v. Cain, 576 U.S. 305, 314 (2015). “We may not characterize . . . state-court factual determinations as unreason- able merely because we would have reached a different conclusion in the first instance.” Id. at 313–14 (alteration adopted) (internal quotation marks omitted). We presume a state court’s factual de- terminations are correct absent clear and convincing evidence to the contrary. See Pye v. Warden, Ga. Diagnostic Prison, 50 F.4th 1025, 1035 (11th Cir. 2022) (en banc). On each claimed basis for relief, we review “the last state- court adjudication on the merits.” See Greene v. Fisher, 565 U.S. 34, 40 (2011). “When a federal claim has been presented to a state court and the state court has denied relief,” we presume “the state court adjudicated the claim on the merits in the absence of any indication or state-law procedural principles to the contrary.” Harrington v. Richter, 562 U.S. 86, 99 (2011). Allred’s relevant ineffective assis- tance claim was adjudicated on the merits, and thus we review it under AEDPA’s standards. III. DISCUSSION In this appeal, Allred argues that his trial counsel was consti- tutionally ineffective for failing to ensure a reasonably competent mental health evaluation during the penalty phase, as he did in USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 29 of 38 22-12331 Opinion of the Court 29 state postconviction proceedings.8 He argues that Caudill’s failure to present mental health evidence was deficient in two ways. First, Allred argues that Caudill unreasonably relied on a misunderstanding of Day’s evaluation of Allred’s mental state with- out adequate investigation into her opinion. This led Caudill to provide ineffective assistance of counsel by deciding not to call Day, or another mental health expert, in the penalty phase. Allred contends that Caudill erroneously understood Day’s opinion to be that if “she had to diagnose Mr. Allred, she would state that he suf- fered from Antisocial Personality Disorder (ASPD), sociopathy, or psychopathy.” Appellant’s Br. 9. According to Allred, Caudill was required to second-guess this opinion and investigate the supposed diagnosis of antisocial personality disorder further. Had he done so, Allred contends, Day would have continued her work and offered the “firm conclusion” that Allred did not have antisocial personality disorder due to his lack of adolescent conduct disorder. Reply Br. 8 As noted above, we granted a certificate of appealability on two aspects of Allred’s ineffective assistance of counsel claim: that trial counsel was ineffec- tive for failing to (1) conduct a sufficient background investigation and (2) en- sure a reasonably competent mental health evaluation for the penalty phase. Allred’s argument on appeal, however, addresses only (2), counsel’s failure to ensure a reasonably competent mental health evaluation. He does not argue that Caudill performed a deficient background investigation; instead, he now argues that Caudill failed to investigate the basis of Day’s evaluation and opin- ion, which is simply another way of saying that Caudill unreasonably relied on Day’s opinion in deciding not to call her as a witness. We therefore discuss only the claim that Caudill was ineffective in failing to ensure Allred under- went a reasonably competent mental health evaluation for the penalty phase. USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 30 of 38 30 Opinion of the Court 22-12331 4–6. Because Caudill did not question Day’s evaluation, Allred ar- gues, his decision to withdraw her testimony was based on insuffi- cient investigation and thus constituted deficient performance. Second, Allred argues that Caudill performed deficiently by failing to seek out and present the testimony of additional mental health professionals, like Caddy and Geffken, whose testimony would have supported statutory mental health mitigators in the penalty phase. Had the trial court heard the testimony of a mental health expert, Allred says, there is a reasonable probability that it would have weighed the balance of aggravating and mitigating cir- cumstances in his favor and sentenced him to life imprisonment instead of death. Counsel provides ineffective assistance, warranting vacatur of a conviction or sentence, when his performance falls “below an objective standard of reasonableness,” taking into account prevail- ing professional norms, and “there is a reasonable probability that, but for counsel’s unprofessional errors, the result of the proceeding would have been different.” Strickland, 466 U.S. at 688, 694. “A rea- sonable probability is a probability sufficient to undermine confi- dence in the outcome.” Id. at 694. To determine whether this rea- sonable probability exists, “we consider the totality of the available mitigation evidence—both that adduced at trial, and the evidence adduced in the habeas proceeding—and reweigh it against the evi- dence in aggravation.” Porter v. McCollum, 558 U.S. 30, 41 (2009) (al- teration adopted) (internal quotation marks omitted). When apply- ing AEDPA to this prejudice standard, “we must decide whether USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 31 of 38 22-12331 Opinion of the Court 31 the state court’s conclusion that [counsel’s] performance . . . didn’t prejudice [the petitioner]—that there was no substantial likelihood of a different result—was so obviously wrong that its error lies be- yond any possibility for fairminded disagreement.” Pye, 50 F.4th at 1041–42 (internal quotation marks omitted). We conclude that the Florida Supreme Court’s determina- tion that Allred was not prejudiced by his counsel’s failure to pre- sent the testimony of a mental health expert during the penalty phase was not unreasonable, and thus its decision is entitled to def- erence under AEDPA.9 Because Allred failed to establish prejudice under the Strickland-AEDPA framework, we affirm the federal dis- trict court’s denial of habeas relief. In so ruling, we need not address whether the Florida Supreme Court’s conclusions as to trial coun- sel’s deficient performance were reasonable. “Because a petitioner must prove both deficient performance and prejudice, a court need not address one element if it determines that the petitioner has failed to prove the other.” Mashburn v. Comm’r, Ala. Dep’t of Corr., 80 F.4th 1292, 1301 (11th Cir. 2023) (internal quotation marks omit- ted). 9 The Florida Supreme Court said little regarding the prejudice prong of Allred’s claim. Nonetheless, it concluded that Allred “failed to establish either prong of Strickland” as to each of his ineffective assistance claims and, more particularly, ruled that Allred “demonstrated neither deficiency nor prejudice” as to the claim that Caudill rendered ineffective assistance in failing to ensure that Allred underwent a reasonably competent mental health evaluation for the penalty phase. Allred II, 186 So. 3d at 535, 539. USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 32 of 38 32 Opinion of the Court 22-12331 Allred resists our conclusion. In arguing that he was preju- diced by trial counsel’s performance, he relies on the testimony of Caddy and Geffken. He argues that his experts’ opinions on two diagnoses—dissociation and autism spectrum disorder—show that the statutory “mental health” mitigators applied to him. Appel- lant’s Br. 20; see Fla. Stat. § 921.141(7)(b), (h). Therefore, he argues, his experts would have shifted the balance of aggravating and mit- igating factors and created a reasonable probability of a life sen- tence. But at the postconviction hearing the parties introduced con- flicting evidence as to whether Allred experienced dissociation and whether he met the criteria to be diagnosed with autism spectrum disorder. The State’s expert, Danziger, refuted Caddy’s and Ge- ffken’s diagnoses. The state postconviction trial court resolved the factual disputes by crediting Danziger’s opinion over the opinions of Allred’s experts, and the Florida Supreme Court affirmed that finding. We first consider whether the credibility finding—a deter- mination of fact, Jones v. Sec’y, Fla. Dep’t of Corr., 834 F.3d 1299, 1315 (11th Cir. 2016)—was unreasonable. See Pye, 50 F.4th at 1035–36 (stating that we cannot grant habeas relief unless the state court’s “ultimate decision” was “based on an unreasonable determination of the facts in light of the evidence presented in the State court pro- ceeding” (internal quotation marks omitted)). Allred argues that the court made “an unreasonable interpretation of the facts [and] the testimony” on its way to concluding that “neither Dr. Caddy’s nor Dr. Geffken’s testimony would have changed the outcome” of his penalty phase, because “[b]oth experts were able to show that USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 33 of 38 22-12331 Opinion of the Court 33 the mental health statutory mitigator applied.” Appellant’s Br. 19– 20. Thus, Allred asks us to evaluate the prejudice prong of his inef- fective assistance claim based on the testimony of his experts and “not based on which expert the courts found more convincing.” Id. at 23. In so arguing, Allred asks us to reject the finding that Caddy and Geffken were not credible and to credit their opinions over the contrary opinion of the State’s expert, Danziger. We cannot do so. We have held that it is not unreasonable for a state habeas court to discount the testimony of one expert in favor of another’s when the experts’ testimony conflicts, so long as crediting the opposing expert’s testimony is itself not unreasonable. See Ferguson v. Sec’y, Fla. Dep’t of Corr., 716 F.3d 1315, 1340–41 (11th Cir. 2013); Jones, 834 F.3d at 1316–17; Pye, 50 F.4th at 1050. Allred does not argue that crediting Danziger’s testimony was unreasonable. So we cannot say that it was unreasonable for the Florida Supreme Court to af- firm the state postconviction trial court’s finding that Allred’s ex- perts were not credible. We next consider the Florida Supreme Court’s ruling that Allred failed to establish the prejudice prong of his ineffective assis- tance of counsel claim. The prejudice ruling is a legal conclusion that we review under § 2254(d)(1). We conclude that the court’s determination was neither “contrary to,” nor “an unreasonable ap- plication of, clearly established federal law” under the statute. 28 U.S.C. § 2254(d)(1). In deciding whether the ruling was unreasona- ble, we reweigh the evidence in mitigation against the evidence in aggravation. Porter, 558 U.S. at 41. And so we turn back to the evi- dence introduced at the evidentiary hearing. USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 34 of 38 34 Opinion of the Court 22-12331 Without Caddy’s and Geffken’s opinions, the mental health evidence adduced at the hearing yielded little in mitigation. The only other testimony indicating that Allred had significant mental health conditions came from Day, who testified that Allred had fea- tures of antisocial personality disorder and psychopathy. In a case where a habeas petitioner presented with “antisocial tendencies,” this Court said that evidence of antisocial personality disorder is “not mitigating, but damaging.” Suggs v. McNeil, 609 F.3d 1218, 1231 (11th Cir. 2010) (internal quotation marks omitted). More re- cently, we have observed that high psychopathic deviate scores on the MMPI-2 and testimony to behavior consistent with that score are damaging rather than mitigating. Puiatti v. Sec’y, Fla. Dep’t. of Corr., 732 F.3d 1255, 1288–89 (11th Cir. 2019). And so, here, as in Suggs, Day’s testimony “would have come at a steep price.” Suggs, 609 F.3d at 1231. We acknowledge that at the hearing experts on both sides noted social and familial difficulties in Allred’s background, as well as his childhood mental health diagnoses of ADHD and tic disor- der. But these facts were presented during the penalty-phase trial and considered at sentencing. In addition, at sentencing the trial court weighed Allred’s emotional distress following his breakup with Barwick. Thus, Geffken’s and Caddy’s testimony on the emo- tional trauma they believed Allred suffered post-breakup—even di- vorced from these experts’ discredited diagnoses—was of little ad- ditional mitigating value. Because the evidence presented in state habeas proceedings was “by no means clearly mitigating” or other- wise “largely duplicated” mitigation evidence at trial, we cannot USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 35 of 38 22-12331 Opinion of the Court 35 say that it was unreasonable for the Florida Supreme Court to con- clude that Allred had failed to show a substantial likelihood of a different sentence. See Cullen v. Pinholster, 563 U.S. 170, 200–01 (2011). Our conclusion that the Florida Supreme Court’s no-preju- dice determination was based a reasonable application of Strickland is bolstered by the strength of the aggravating circumstances the sentencing court identified. Where there is “substantial evidence of aggravating circumstances,” it is more difficult for a petitioner to establish prejudice under Strickland. Holsey v. Warden, Ga. Diagnos- tic Prison, 694 F.3d 1230, 1269 (11th Cir. 2012). So, too, when the mitigating evidence is scant. The new mitigating evidence in this case—that Allred had antisocial traits—would likely have been damaging. Even considering the mitigating evidence presented at trial—that Allred had social difficulties, a troubled family life, and childhood developmental disorders but was a self-sufficient young adult with no prior history of violence—we cannot conclude that the Florida Supreme Court unreasonably found no prejudice, given the aggravating circumstances. And “[t]his is not a case where the weight of the aggravating circumstances or the evidence supporting them was weak.” Sochor v. Sec’y, Dep’t of Corr., 685 F.3d 1016, 1030 (11th Cir. 2012) (internal quotation marks omitted). The trial court found that both Ru- schak’s and Barwick’s murders were cold, calculated, and USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 36 of 38 36 Opinion of the Court 22-12331 premeditated and that Barwick’s murder was heinous, atrocious, or cruel.10 For the cold, calculated, and premeditated aggravator, the evidence of premeditation in the penalty-phase case was strong. The trial court cited the timing of Allred’s purchase of the murder weapon, his threatening messages to and about the victims, and his warning to Ruschak about his arrival at the scene as evidence that the murders were preplanned. If Caddy’s testimony had been cred- ited, it is possible that his opinion of Allred’s diminished capacity might have mitigated the impact of this premeditation evidence. But Caddy’s testimony was not credited. And Allred introduced no other evidence at the evidentiary hearing to undercut the cold, cal- culated, and premeditated aggravator. To the contrary, the evi- dence adduced at the evidentiary hearing tended to show that the testimony of a mental health expert would have supported the cold, calculated, and premeditated aggravator. Based on her evalu- ation of Allred, Day opined that his behavior leading up to the 10 As we previously noted, the court found additional aggravating factors as to each murder. As to Ruschak’s murder, the court found aggravating that Allred committed the murder while engaged in a burglary, and he was previ- ously convicted of another capital felony (Barwick’s contemporaneous mur- der). As to Barwick’s murder, the court found aggravating that Allred was pre- viously convicted of another capital felony (Ruschak’s contemporaneous mur- der). Because the cold, calculated, and premeditated and heinous, atrocious, or cruel findings are sufficient to support our conclusion that the Florida Su- preme Court’s prejudice determination was not unreasonable, we do not dis- cuss the court’s findings as to the other aggravators. USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 37 of 38 22-12331 Opinion of the Court 37 murders suggested premeditation. Caddy testified that Allred had a preconceived fantasy of killing Ruschak and Barwick. To support the heinous, atrocious, or cruel aggravator, the trial court adduced from the 911 call Barwick made while hiding from Allred that she was terror-stricken, anticipating her own death, in the minutes before Allred shot and killed her. It described the 911 call “as the most horrific piece of evidence this court has heard in a homicide case in nearly twenty-three years as a trial judge.” And the heinous, atrocious, or cruel aggravator “pertains more to the nature of the killing and the surrounding circum- stances” than the petitioner’s mental state. Hardwick v. Sec’y, Fla. Dep’t of Corr., 803 F.3d 541, 561 (11th Cir. 2015) (internal quotation marks omitted). Thus, postconviction testimony about Allred’s mental state was unlikely to undermine the heinous, atrocious, or cruel aggravator. We have said that the cold, calculated, premeditated and heinous, atrocious, or cruel aggravators are “among the most seri- ous aggravating circumstances.” Id. at 559. It is improbable that the evidence adduced at Allred’s postconviction evidentiary hearing would have reduced the impact of these powerful aggravators suf- ficiently to introduce the reasonable probability of a different out- come for Allred. See Pye, 50 F.4th at 1049 (“We’ve repeatedly held that even extensive mitigating evidence wouldn’t have been rea- sonably likely to change the outcome of sentencing in light of a particularly heinous crime and significant aggravating factors.”). USCA11 Case: 22-12331 Document: 31-1 Date Filed: 04/11/2024 Page: 38 of 38 38 Opinion of the Court 22-12331 For these reasons, the Florida Supreme Court’s determina- tion that Allred could not show prejudice from his trial counsel’s performance withstands our highly deferential review under AEDPA. The court’s ruling that there was no substantial likelihood of a life sentence in Allred’s case was not “so obviously wrong that its error lies beyond any possibility for fairminded disagreement,” so it is entitled to deference. See id. at 1041–42 (internal quotation marks omitted). Our reweighing of the totality of the evidence in mitigation against the evidence in aggravation shows that Allred was unable to shift the balance of the sentencing factors. Although Allred argues that evidence of his mental state should have been presented at trial, the mental health evidence adduced at the post- conviction hearing—notwithstanding the testimony found not credible—would have been of little value. The mental health ex- pert testimony presented was either supportive of existing aggra- vating factors like premeditation, supportive of new and poten- tially aggravating mental health diagnoses like antisocial personal- ity disorder, or cumulative of mitigating evidence the sentencing court considered. We therefore cannot say that the Florida Su- preme Court’s prejudice ruling was “contrary to” or “an unreason- able application of[] clearly established Federal law.” 28 U.S.C. § 2254(d)(1). The court’s decision withstands our highly deferential review under AEDPA, and we affirm the denial of relief on Allred’s penalty phase ineffective assistance of counsel claim. IV. CONCLUSION The district court’s denial of Allred’s petition for a writ of habeas corpus is AFFIRMED.