Robert Wayne Holsey v. Warden, Georgia Diagonstic Prison

BARKETT, Circuit Judge,

dissenting:

For the reasons stated in my dissent in Hill v. Humphrey, 662 F.3d 1335, 1365-78 (11th Cir.2011) (en banc), I continue to believe that Georgia’s requirement that defendants prove mental retardation beyond a reasonable doubt is unconstitutional under Atkins v. Virginia, 536 U.S. 304, 122 S.Ct. 2242, 153 L.Ed.2d 335 (2002).

I also believe that the Georgia Supreme Court’s “decision” that the ineffectiveness of counsel did not prejudice Holsey is “based on an unreasonable determination of the facts,” see 28 U.S.C. § 2254(d)(2) (2006), and therefore, we must conduct a de novo review. See Cooper v. Sec’y, Dep’t of Corr., 646 F.3d 1328 (11th Cir.2011); Jones v. Walker, 540 F.3d 1277, 1288 n. 5 (11th Cir.2008) (en banc). A de novo review compels the conclusion that the omission of evidence describing the nature and extent of Holsey’s abusive childhood along with evidence of his mental retardation prejudiced him at the sentencing phase of his trial in violation of Strickland v. Washington, 466 U.S. 668, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984).

More specifically and as demonstrated below, the Georgia Supreme Court’s determination that the extensive evidence offered on collateral review was “largely cumulative” is unreasonable in light of the sparse — almost non-existent — evidence of childhood abuse and mental retardation presented in Holsey’s trial. See 28 U.S:C. § 2254(d)(2). At trial, a brief mention was made of the fact that Holsey was beaten, without any further explanation or description. The jury never learned that throughout his childhood he was subjected to abuse so severe, so frequent, and so notorious that his neighbors called his childhood home “the Torture Chamber.” Likewise, with reference to his status as borderline mentally retarded, a diagnosis that is undisputed by the state’s expert witnesses, the jury heard only that a report listed him as borderline mentally retarded without any testimony to explain the extent and consequences of his condition. Only one juror’s vote was necessary to impose a life sentence.1 The record here demonstrates that “both the nature and the extent of the abuse the petitioner suffered” would have affected the probability that “at least one juror” would have voted for a sentence less than death, see Wiggins v. Smith, 539 U.S. 510, 535, 536, 123 S.Ct. 2527, 156 L.Ed.2d 471 (2003) (emphasis added), and the mitigating evidence describing the behavioral and cognitive impact of Holsey’s borderline retardation would have influenced “the jury’s appraisal of his moral culpability,” see *1276Williams v. Taylor, 529 U.S. 362, 398, 120 S.Ct. 1495, 146 L.Ed.2d 389 (2000).

With reference to de novo review, trial counsel was so deficient that confidence in the outcome of Holsey’s sentencing is undermined and he is entitled to a new sentencing hearing. Rather than conduct an investigation that could lead to the discovery and presentation of the important mitigating evidence delineated below, Holsey’s lead defense lawyer drank a quart of vodka every night of Holsey’s trial while also preparing to be sued, criminally prosecuted, and disbarred for stealing client funds. He admitted during collateral proceedings that, at the time he was preparing for Holsey’s capital murder trial, he “probably shouldn’t have been allowed to represent anybody” due to his condition. The Georgia trial court, which is the only court to have presided over the receipt of live testimony and evidence on collateral review, held that “no one can seriously believe that [Holsey] received the constitutional guarantees of the Sixth Amendment right to effective assistance of counsel.” The mitigating evidence which should have been presented would have created a reasonable probability that “at least one juror” would have voted for a sentence less than death, see Wiggins, 539 U.S. at 535, 536, 123 S.Ct. 2527, even weighed against the aggravating factors presented. I amplify each of these points below.

1. The Mitigating Evidence of Childhood Abuse is Not “Largely Cumulative” of the Sentencing-Phase Testimony

The brief mention of abuse at the sentencing phase came from two witnesses: Holsey’s sister, Regina Reeves, and Clifford Holsey, who is not related to Wayne Holsey.2 Regina Reeves’ testimony made a brief, passing reference to Holsey while explaining her own move out of her mother’s house as she approached her majority. She explained why she left home at the age of seventeen to “live with friends.”

Q: And what caused you to leave home?
A: We were having — I felt that I was — that we were having a lot of problems. I hated it there. Things were bad. I was a mother, you know, by the time I left home. I had my daughter when I was 17. I graduated from high school when I was 17. I didn’t feel that I was, you know, treated well, that I was appreciated. And I felt that I could do better, you know, somewhere else because I just felt that things were horrible there.
Q: And what about being there was so horrible? ....
A: Okay. I was tired of taking beatings. I don’t think it was, you know, so horrible that we didn’t, you know, have anything. A lot of people didn’t have anything. But I always felt that you could be — not have more than anybody else but still get treated better than what we were.
Q: And when you say you were tired of taking beatings, who beat y’all?
A: My mother. (Witness crying)
Q: Did she beat all the children?
A: The oldest three mostly.
Q: That was you and Wayne and Angela?
A: (Witness nods affirmatively.) Yes.

Counsel asked Regina no further questions about beatings or abuse and this was the totality of testimony from any family member about abuse.3

*1277Clifford Holsey, who was called to describe a bar fight involving Holsey, stated during cross examination that Wayne Holsey’s mother “would scold ... and beat” the children, but admitted that he had never actually seen her beat them. He simply stated:

Well, I think Wayne came up the best that he could. I think he was neglected from his mother. She, you know, kind of like — came up kind of like child abuse. And she just didn’t see about them, you know, kind of walked all over them a little bit, and done everything.

Regina and Clifford’s testimony can hardly be characterized as having“highlighted” Holsey’s history of abuse, as the Georgia Supreme Court determined. See Schofield v. Holsey, 281 Ga. 809, 642 S.E.2d 56, 61-62 (2007). It hardly comports with the “Torture Chamber” described by Holsey’s neighbors and family members at the collateral review hearing. Holsey’s sister testified on collateral review, but not at trial, that Wayne Holsey was beaten with extension cords, shoes, broom handles, or whatever objects their mother could find, leaving permanent scars on his body. His mother punished him by holding his head under water while he cried and struggled to breathe. The beatings would spill out of the house and into the yard and the street, as people from the neighborhood stood by, watching. These beatings would occur for any reason or no reason at all, or, as Regina put it on collateral review, but not at trial, when her mother “felt that the house wasn’t clean,” was “just frustrated,” “jealous,” or “you know, just about anything.”

Holsey wet the bed until he was thirteen years old, and Regina testified on collateral review, but not at trial, that on occasion his mother stayed up at night so that, when she heard the urine leaking to the floor through Holsey’s thin mattress, she could beat him out of his sleep with an extension cord. She beat him when he went across the street to pick blackberries. Even eating meals, such as there were, became an opportunity for abuse, as Holsey’s mother punished him for asking for more food by forcing him to eat chicken bones. Regina told the judge on collateral review that abuse of this magnitude happened “daily [or] every other day.”

Members of Holsey’s extended family and community also came forward to testify to the notoriety of the abuse inflicted on Holsey. Bertha Ingram, Holsey’s adult cousin, testified to witnessing Holsey’s mother beat him with the pointed end of her high-heeled shoes and other household objects, burn him with a curling iron, and lock him in a closet. She corroborated Reeves’ testimony that abuse on this scale occurred on a daily basis, testifying that “it was a rare evening” when these beatings would not occur for reasons as trivial as Holsey opening the refrigerator to look for food.

*1278Essie Anderson, a Mend of Holsey’s mother who worked at the local hospital and visited the apartment regularly, stated that the abuse was well known throughout the neighborhood, and reflected that “we were all guilty for letting Mary [Holsey’s mother] get away with the way she mistreated her kids, but we were all so scared of her ourselves.” When Anderson tried to intervene, Holsey’s mother would shout at her “I’ll kill him if I want to. These are MY kids.” Neighbor Catherine Harris recalled going to Holsey’s apartment when he was a child and watching Holsey’s mother attack him and his sister with a belt or curling iron “while they tried to squat down in a corner and curl up in a little ball to fend off her blows.” Another neighbor, Sandra Francis, recounted the times she had watched out the window and shook with fear while Holsey’s mother beat him in the yard, and that after these beatings, “it wasn’t uncommon for [Holsey] to curl up in the bushes outside the Housing project office ... and sleep there until morning.”

The Georgia Supreme Court’s decision that Holsey was not prejudiced by the omission of this evidence was based on its determination that Holsey’s original sentencing hearing “highlighted” Holsey’s abusive history such that his collateral evidence was “largely cumulative” of what was introduced at trial. However, this determination is unreasonable in light of the facts contained in this record. See § 2254(d)(2). Because the Georgia Supreme Court’s “decision” that Holsey was not prejudiced at sentencing was “based on” an unreasonable factual determination, we owe it no deference. See Jones, 540 F.3d at 1288 n. 5; see also Callahan v. Campbell, 427 F.3d 897, 927 n. 26 (11th Cir.2005) (emphasizing that § 2254(d)(2) applies where “‘the adjudication of the claim ... resulted in a decision that was based on an unreasonable determination of the facts’ ”) (emphasis in original) (quoting § 2254(d)(2)); Harvey v. Warden, 629 F.3d 1228, 1252 (11th Cir.2011) (applying § 2254(d)(2) to state court’s factual determinations underpinning its conclusion that the petitioner was not prejudiced under Strickland).

In Cooper v. Secretary, Dep’t of Corrections, 646 F.3d 1328 (11th Cir.2011), we likewise concluded that a state court’s rejection of a petitioner’s Strickland, claim was based on an unreasonable factual determination under a materially indistinguishable set of facts. In Cooper, the Florida Supreme Court rejected Richard Cooper’s Strickland claim on the ground that “ ‘a substantial part of the information regarding Cooper’s disadvantaged childhood was presented at Cooper’s trial,’ ” and therefore “ ‘in large part, introduction of the evidence proffered [on collateral review] would have been repetitive.’ ” See Cooper, 646 F.3d at 1348 (quoting Cooper v. State, 856 So.2d 969, 976 (Fla.2003)). However, we held that the Florida Supreme Court’s conclusion that the collateral testimony was “in large part ... repetitive of’ the sentencing phase testimony was an unreasonable determination of the facts under § 2254(d)(2), because this testimony omitted the “specifics of the abuse” directed against Cooper and “did not begin to describe the horrible abuse testified to [on collateral review].” Id. at 1352-53.

Every one of the dispositive facts that led this Court in Cooper to find the state court’s factual determinations unreasonable under § 2254(d)(2) is present in this case. First, Holsey’s sentencing phase testimony was even less illuminating as to the nature and extent of abuse than in Cooper, where the defendant’s mother testified that Cooper’s father beat him with a belt, leaving marks on his body, was “very hard” and “authoritarian” with him, and defense counsel emphasized that Cooper’s “horrible” and “tragic” family life was “something that none of [the jurors] ha[d] *1279experienced.” Id. at 1337, 1339. Moreover, during proceedings before the sentencing judge, a defense psychologist testified that Cooper had had a “ ‘horrendous background,’ ” that his father was “ ‘exceptionally abusive, both physically and verbally,’ ” and that the impact of the “terror-filled years” of abuse his father inflicted caused Cooper to suffer lasting psychological harm. Id. at 1340.

Second, the disparity between the sentencing testimony about abuse and what was revealed on collateral review in Holsey’s case was even greater than that in Cooper. On collateral review, the “ ‘description, details, and depth of abuse’” presented by Cooper “ ‘far exceeded what the jury was told’ ” at Cooper’s sentencing. Id. at 1354 (quoting Johnson v. Sec’y, 643 F.3d 907, 936 (11th Cir.2011)). Similarly, the “description, details, and depth of abuse” presented by Holsey on collateral review “far exceeded” the sentencing testimony at Holsey’s trial, and the abuse revealed by Holsey’s collateral evidence is certainly as extensive and atrocious as the evidence omitted in Richard Cooper’s trial.

Finally, the erroneous characterizations of the sentencing phase testimony about abuse are the same in both cases. In Cooper, the state court described the “substantial part of the information” on collateral review as “in large part ... repetitive” of the sentencing testimony. See id. at 1348 (internal quotation marks omitted). Likewise, the Georgia Supreme Court characterized the information presented on collateral review as “largely cumulative” of or “highlighting]” Holsey’s sentencing phase testimony. Based on the almost identical facts of Cooper, the same legal conclusion is compelled here: the Georgia Supreme Court’s characterization of the collateral evidence as largely cumulative of the testimony at trial is an unreasonable determination of the facts under § 2254(d)(2).

The majority disparages Cooper as “an outlier.” See Majority op. at 1259. However, we have no authority to disregard a previous decision of another panel arising from facts that, like the facts of Cooper, are materially indistinguishable from those before us.4 See Anders v. Hometown Mrtg. Svcs., Inc., 346 F.3d 1024, 1031 (11th Cir.2003). That rule “is not dependent upon a subsequent panel’s appraisal of the initial decision’s correctness.” Smith v. GTE Corp., 236 F.3d 1292, 1302 (11th Cir. 2001) (internal quotation marks omitted).

Moreover, Cooper is not an outlier and is fully consistent with our cases, which, in applying § 2254(d)(2), have reviewed state appellate courts’ resolutions of factual issues which are the predicates to their legal conclusions. See, e.g., Hardy v. Comm’r, Dep’t of Corr., 684 F.3d 1066, 1087-88 (11th Cir.2012) (applying § 2254(d)(2) to factual determinations underpinning decision whether defendant invoked his Miranda right to remain silent); Harvey, 629 F.3d at 1252 (applying § 2254(d)(2) to state court’s determination that defendant was not prejudiced under Strickland by counsel’s statements in opening argument because these statements were factually similar to statements in the defendant’s confession); Bui v. Haley, 321 F.3d 1304, 1315-16 (11th Cir.2003) (applying § 2254(d)(2) to appellate court’s factual determination that was “subsidiary” to a legal conclusion). Similarly, in Cooper, as in *1280Holsey’s case, the state court’s characterization of evidence as cumulative was predicated on a subsidiary determination that the two bodies of evidence being compared are factually comparable to each other. See Harvey, 629 F.3d at 1252 (reviewing under § 2254(d)(2) a state court’s comparison of the factual content of two statements contained in the record). The Georgia Supreme Court’s factual determination was “subsidiary” to the ultimate legal decision that the petitioner had not shown prejudice under Strickland, and it involved a comparison of the collateral evidence with the trial evidence. See Bui 321 F.3d at 1316. In both Cooper and Holsey’s cases, the state appellate courts’ decisions that no prejudice was established were predicated on factual determinations that were “unreasonable ... in light of the evidence presented in the State court proceeding.” See § 2254(d)(2).5 Cooper is consistent with the cases from our circuit applying § 2254(d)(2).

Moreover, both the Supreme Court and our court have held prejudice to be established under Strickland where the testimony presented at trial failed to describe the “nature and extent” of childhood abuse suffered by the petitioner. If such evidence is capable of establishing prejudice, it cannot reasonably be dismissed as “largely cumulative” of testimony that merely mentions that the defendant was beaten. Specifically, in Wiggins, it was precisely the “nature and extent of the abuse [the] petitioner suffered” that led the Supreme Court to conclude that any reasonable counsel would introduce evidence of the petitioner’s abusive childhood due to its likely impact on the jury. See 539 U.S. at 535-37, 123 S.Ct. 2527 (emphasis added). Similarly, in Williams, the Court held that the petitioner was prejudiced by counsel’s omission of a “graphic description of [the petitioner’s] childhood” including “documents ... that dramatically described mistreatment, abuse, and neglect during his early childhood.” 529 U.S. at 370, 398, 120 S.Ct. 1495 (emphasis added). And in Porter v. McCollum, 558 U.S. 30, 130 S.Ct. 447, 175 L.Ed.2d 398 (2009) (per curiam), the Court held that “it is unreasonable to discount to irrelevance the evidence of [a defendant’s] abusive childhood.” 130 S.Ct. at 455. In each of these cases, it was not the omission of an acknowledgment of abuse that the Court found prejudicial, but the omission of testimony that described this abuse in sufficient detail for the jury to understand what actually occurred.

Similarly, in Williams v. Allen, 542 F.3d 1326, 1342 (11th Cir.2008), notwithstanding that the defendant’s mother testified at sentencing that his father beat him “with his fists,” “many times,” “whipped him more than he should,” and once beat him so severely that the defendant escaped to a neighbor’s house and called the police, see Williams, 542 F.3d at 1329, we held the defendant was prejudiced by the omission of his sister’s account of abuse because her testimony made clear that “the violence *1281experienced by Williams as a child far exceeded — in both frequency and severity — the punishments described at sentencing.” See id. at 1342 (emphasis added).

Finally, in Johnson v. Secretary, Dep’t of Corrections, 643 F.3d 907 (11th Cir.2011), we held that prejudice was established because of the “description, details, and depth of abuse in [the defendant’s] background that were brought to light” on collateral review. See Johnson, 643 F.3d at 936 (emphasis added). In Johnson, counsel deficiently failed to introduce detailed evidence about how the defendant and his siblings “huddled together in terror” when their father would beat their mother, that their parents engaged in “knockdown, drag-out fights” that sometimes led them to call the police, and that the defendant’s mother “beat him more severely than the other children — sometimes with her knuckles and sometimes with a leather strap.” Id. (internal quotation marks omitted).

Although mention was made by Clifford Holsey and Regina Reeves about beatings, it is no less true in Holsey’s case than it was in Johnson that the “description, details, and depth” of the abuse are what gives this evidence mitigating value and are what was completely omitted in Holsey’s trial. Indeed, we held in Johnson that prejudice was established in part because the evidence at trial “was not nearly as helpful to [the defendant’s] case as it could have been” and “misleadingly minimized the mitigating circumstances.” Id. Counsel’s failure to present any testimony about the “nature and extent” of the abuse Holsey suffered, Wiggins, 539 U.S. at 535, 123 S.Ct. 2527, similarly misled the jury about Holsey’s background because it “minimized the mitigating circumstances” of Holsey’s abusive history.

The cases relied upon by the majority are inapplicable to the facts of this case. In the cases cited by the majority, there either was descriptive evidence at sentencing conveying the nature and extent of the abuse suffered by the petitioner, or the mitigating evidence introduced on collateral review did not resemble the consistent, voluminous, and unrebutted evidence of pervasive and severe abuse that is at issue in this case. Thus, in Sochor v. Secretary, Dep’t of Corrections, 685 F.3d 1016 (11th Cir.2012), the testimony during the petitioner’s sentencing phase did describe, in extensive and even greater detail than the testimony on collateral review, the abuse that the petitioner regularly suffered as a child, including that he suffered multiple head injuries inflicted in an identical manner to the head injury described on collateral review. See Sochor, 685 F.3d at 1021-22.6 In Sochor, the jury was told about *1282the kind, extent, and severity of the abuse and not simply that beatings occurred that may or may not have even amounted to abuse.7

. Likewise, in Cullen v. Pinholster, —• U.S.-, 131 S.Ct. 1388, 179 L.Ed.2d 557 (2011), the testimony presented on collateral review was relatively limited in scope and did not approach a description of the scale of abuse at issue here, where not only family members but also members of Holsey’s community stepped forward to describe the severity and widespread notoriety of this abuse. See Cullen, 131 S.Ct. at 1409-10. The evidence presented at the collateral hearing provided far less support than that presented on behalf of Holsey at his collateral hearing. See Pinholster v. Ayers, 590 F.3d 651, 712 (9th Cir.2009) (en banc), (Kozinski, C.J., dissenting), rev’d, — U.S. -, 131 S.Ct. 1388, 179 L.Ed.2d 557 (“[Wjhat’s remarkable is how little support the family members provide for Pinholster’s theory of extreme abuse and deprivation .... What’s remarkable here is just how weak this testimony actually is.”); id. at 713 (“The simple fact is, there’s nothing supporting the theory of abuse or deprivation, in stark contrast to the evidence in many other cases.”) (citing Wiggins, 539 U.S. at 517, 123 S.Ct. 2527; Williams, 529 U.S. at 370, 120 S.Ct. 1495). Moreover, the ostensibly mitigating testimony at issue in Cullen was undermined by the fact that the petitioner himself denied that any abuse had actually occurred, and in fact described the purported abuse by his stepfather as “ ‘discipline’ from which he ‘benefitted.’ ” See Pinholster v. Ayers, 525 F.3d 742, 767-68 (9th Cir.2008), vacated en banc, 590 F.3d 651 (9th Cir. 2009), rev’d, — U.S. --, 131 S.Ct. 1388, 179 L.Ed.2d 557 (2011); see also id. at 767 (“Pinholster’s primary complaint about his step-father was that ‘he didn’t seem to want the kids around.’ ”).8 In contrast,' in *1283Holsey’s case, the evidence of pervasive abuse is uncontradicted.

II. The Collateral Hearing Evidence of Mental Retardation Was Not “Largely Cumulative” of That Presented at the Sentencing Hearing

The Georgia Supreme Court’s determination that trial testimony “highlighted Holsey’s limited intelligence” such that his mitigating evidence was “largely cumulative” of the collateral .evidence is also an unreasonable determination of the facts as it relates to his uncontested borderline mental retardation. The only reference to Holsey’s mental retardation during the sentencing phase was .Regina Reeves reading the words “borderline mental retardation” from a report created when Holsey was fifteen. That testimony was as follows:

Q: And I want you to read to the jury a portion of this psychological testing dated 7-25-80. And, again, just read the highlighted portion. The jury will be able to read the whole thing on this page and the next page. This is regarding Robert.
A: It says “Robert evidenced in”— I’m sorry. “Robert evidenced an inap- • propriate effect [sic] during the evaluation. ' He smiled inappropriately and had difficulty maintaining thought patterns. At times he appeared unaware of his immediate environment, and in a world of his own.” Another paragraph says, “present testing indicates Robert functions in the borderline mental retardation range of intelligence.”
Q: Borderline mental retardation range.
A: Yes.

As the Georgia trial court found on collateral review, and as the Georgia Supreme Court did not dispute, Holsey’s record was “read to the jury without any context or explanation of its significance.” Indeed, although Regina was the only witness who had the opportunity to explain the meaning of this information during Holsey’s sentencing, she testified on collateral review that she had never seen the report before she was asked to read it, had not discussed this testimony with counsel in advance of trial, had never heard her brother described as borderline mentally *1284retarded before she read those words to the jury, and had not even met the attorney who asked her to read these words before she did so. As a result, although the jury heard the words “borderline mental retardation” spoken during live testimony and again, once, by counsel during closing argument, the jury had no way of knowing what those words meant, either in general or as they relate to Holsey.

In contrast, on collateral review, Holsey presented testimony and written reports from four psychologists: Dr. Mark Cunningham, Dr. Jethro Toomer, Dr. Marc Einhorn, and Dr. Michael Shapiro. Both Dr. Cunningham and Dr. Toomer concluded that Holsey is mentally retarded. Dr. Einhorn, who was asked to evaluate Holsey by the state, and Dr. Shapiro, who had examined Holsey in preparation for trial, agreed at the very least that he is “in the borderline range” of mental retardation.

In the course of explaining his evaluation of Holsey as mentally retarded, Dr. Cunningham’s testimony related accounts given by various adults who knew Holsey and concluded that Holsey’s adaptive behavior in the area of communication was that of a six-year-old child. Dr. Cunningham related accounts of Holsey’s inability to take care of his basic living needs. Holsey had “never lived on his own ... but always lived with either a woman that he was going with or a family member.” He would give whatever money he earned to his living companion at the time, who would pay him the money as an allowance, sometimes reminding him how to count the money when she gave it to him. Holsey never had a bank account, was unable to shop for groceries or clothing by himself, was unable to use public transportation, and could not order food from a menu at a restaurant. While working at a Pizza Hut restaurant, Holsey’s supervisors had tried to promote him from dishwasher to pizza maker, but Holsey was unable to follow the recipe for making the pizzas. In this area of testing, Dr. Cunningham placed Holsey’s capacity as falling in the range of a four-year-old to six-and-a-half-year-old child.

Holsey’s social adaptation was also years behind his biological age. Dr. Cunningham summarized several accounts of Holsey’s social interaction in these terms:

The description that he’s a loner is not [that he is] a loner by choice; it’s not that he is schizoid and is uninterested in relating to others, it’s he wants the friends, he’s just not able to relate on an adult to adult basis. With the kids in the household he lives in, with the girlfriend’s kids, he got along famously with them, would play with them for hours. It’s not that he’s not interested in relating, it’s that his relationship skills and capabilities are at a child level and that puts him out of sync as he attempts to relate to adults.

Similarly, Holsey’s relationships with women in adulthood “all ultimately broke up as they identified that they were relating to an emotional child instead of an adult .... [Ultimately they just couldn’t tolerate the experience of being in an intimate relationship with a child.” Dr. Cunningham gauged Holsey’s social adaptive behavior as at the level of a five-year-old child.

In terms of self-direction, Dr. Cunningham related accounts given by other adults who knew Holsey that he was unable to direct himself toward goals as simple as cleaning the house or even cooking a basic meal. When confronted by another adult about his inability to perform these simple tasks, Holsey “would just slump his shoulders and hang his head and mumble about what he couldn’t do very well.” Testing in the area of self-direction placed Holsey at the functional level of a four-year-old child.

Summarizing his evaluation in these and other areas, Dr. Cunningham testified that *1285Holsey operates, on average, at the level of an eight-year-old child in his adaptive behavior. He testified that, whereas mental retardation requires proof of deficits in only two of ten behavioral areas, Holsey was deficient in eight. Moreover, Dr. Cunningham testified that Holsey’s IQ scores taken over a twenty-three year period by the Georgia Youth Development Center (“YDC”), by Dr. Einhorn, and by Dr. Cunningham himself, and which registered an IQ of 69, 70, and 71, amounted to “an extraordinarily reliable demonstration of his actual intellectual capability” as these scores “cluster[ed] within three points.”9 Dr. Cunningham described Holsey’s borderline mental retardation as “a catastrophic disability.”10

Dr. Toomer’s testimony also described Holsey’s intellectual and behavioral capacity as lagging far behind his biological age. Dr. Toomer reported that Holsey registered scores at a fifth- or fourth-grade level in reading, math, and spelling tests administered as part of his assessment. Dr. Toomer’s assessment was also consistent with Dr. Cunningham’s conclusion that Holsey lagged far behind his age group in multiple areas of adaptive behavior.

Dr. Einhorn examined Holsey for the state, but he agreed with Holsey’s experts that Holsey’s test results and behavioral development met all three elements of the definition of mental retardation required by Georgia law. Consistent with test results obtained by Drs. Cunningham and Toomer, Dr. Einhorn’s academic testing revealed that Holsey performs at a fourth- or fifth-grade level in reading, spelling, and arithmetic. He nevertheless opined that Holsey was not mentally retarded because he judged Holsey’s impairments to have been caused by “cultural deprivation” and alcohol abuse instead of mental retardation.11 Consistent with Holsey’s ex*1286perts’ assessments, however, Dr. Einhorn testified that Holsey’s mental functioning was between “low average” and “borderline” mental retardation. Einhorn also testified that his testing revealed no signs that Holsey was malingering, or dissembling in order to appear mentally retarded, which was also consistent with Dr. Cunningham’s assessment that Holsey’s responses to questions were the product of genuine effort.

Dr. Shapiro, who was retained by Holsey’s trial counsel but who never testified during Holsey’s trial or sentencing, testified on collateral review that Holsey is “in the borderline range” of mental retardation. Like Dr. Toomer, Dr. Shapiro testified that Holsey reads at a fourth-grade level.

Finally, even Dr. Thomas Sachy, a psychiatrist who examined Holsey on behalf of the state, offered testimony that would have been helpful to Holsey if put before a jury. Specifically, Dr. Sachy described mentally retarded individuals as often incapable of keeping jobs because they are unable to perform sequenced activities. This description was consistent with Holsey’s evidence that he is unable to follow basic directions required to shop for groceries, cook a simple meal, or perform jobs more sophisticated than dishwasher or truck driver.12 Dr. Sachy identified inability to maintain adult relationships as another characteristic feature of mental retardation, and this description was corroborative of testimony by Holsey’s experts and former girlfriends that he was unable to interact on an adult level.. Further, Dr. Sachy acknowledged that inability to operate simple machinery, such as a washing machine, without instructions and supervision may be consistent with mild mental retardation, just as Holsey’s collateral testimony revealed that Holsey was unable to operate a washing machine. Dr. Sachy emphasized his view that the intellectual element of borderline retardation is the most important element in diagnosing that condition, and he unequivocally testified that Holsey’s IQ scores place him in the borderline range of mental retardation.

Clearly, Regina Reeves’ unexplained reading of the words “borderline mental retardation range” cannot reasonably be called “largely cumulative” of this testimony. The cognitive and behavioral impairments that were painstakingly explained on collateral review by the psychologists are not a matter of everyday knowledge such that a jury will be reminded of them automatically merely by hearing the words “borderline mental retardation” spoken. Rather, the testimony by these experts would have enabled the jury to understand in concrete terms that Holsey suffers from the “cognitive and behavioral impairments” that reduce the moral culpability of mentally retarded and borderline mentally retarded offenders, including Hol*1287sey’s diminished capacity “to understand and process information, to communicate, to engage in logical reasoning, to control impulses, and to understand the reactions of others.” Atkins v. Virginia, 536 U.S. 304, 318, 122 S.Ct. 2242, 153 L.Ed.2d 335 (2002); see Brownlee v. Haley, 306 F.3d 1043, 1073 (11th Cir.2002).13

The majority refers to Regina Reeves’ sentencing phase testimony that Holsey “performed poorly in school, and was usually assigned to the next grade level instead of actually passing into that grade level,” that Holsey “dropped out of school before finishing the tenth grade,” and that Holsey was “ ‘very slow* and a ‘poor worker’ who ‘need[ed] help from home’ but never got that help.” Majority op. at 1262. However, nothing in the testimony referred to describes the depth and severity of mental impairment that distinguishes a borderline mentally retarded defendant from the general population. See Atkins, 536 U.S. at 319, 122 S.Ct. 2242 (“If the culpability of the average murderer is insufficient to justify the most extreme sanction available to the State, the lesser culpability of the mentally retarded offender surely does not merit that form of retribution.”). To say that testimony describing Holsey as having academic difficulties that are shared by non-mentally retarded individuals “highlighted” the effects of Holsey’s mental retardation described on collateral review is to erase the distinction between mentally retarded persons and those who are not. See Atkins, 536 U.S. at 319, 122 S.Ct. 2242 (“ ‘By definition, [mentally retarded] individuals have substantial limitations not shared by the general population.’ ” (quoting Atkins v. Commonwealth, 260 Va. 375, 534 S.E.2d 312, 325 (2000) (Koontz, J., dissenting))). Testimony that describes the impairments that are unique to mentally retarded individuals is necessary to distinguish the mentally retarded or borderline mentally retarded defendant from a defendant whose culpability is not decreased by this incapacity. Such testimony does not, as the majority claims, merely add “details” or elaborate on “themes” that have already been canvassed; instead, it is essential to give recognition to the “cognitive and behavioral impairments that make these defendants less morally culpable.” Id. at 320, 534 S.E.2d 312. The sentencing phase testimony wholly failed to describe the effects of Holsey’s borderline mental retardation.

III. Holsey Received Ineffective Assistance of Counsel During the Penalty Phase of His Trial

Because the Georgia Supreme Court’s decision is founded on the unreasonable factual determination that the sentencing phase testimony “highlighted” Holsey’s abuse and borderline mental retardation such that his mitigating evidence was “largely cumulative,” we must apply de novo review to the prejudice component of Holsey’s Strickland claim. See § 2254(d)(2); Cooper, 646 F.3d at 1353; Jones, 540 F.3d at 1288 n. 5. Moreover, because the Georgia Supreme Court did not adjudicate the deficiency prong of Holsey’s Strickland claim on the merits, we have no state-court adjudication of the deficiency prong to defer to. See Rompilla, 545 U.S. at 390, 125 S.Ct. 2456.

A. Deficiency

To establish that his counsel provided ineffective assistance, Holsey must show that his counsel’s performance was objectively unreasonable according to prevailing professional norms at the time of his trial. *1288See Strickland, 466 U.S. at 687-88, 104 S.Ct. 2052. It was well established throughout the time of counsel’s preparation that an attorney representing a defendant in a capital case bears “ ‘an obligation to conduct a thorough investigation of the defendant’s background.’ ” Johnson, 643 F.3d at 931 (quoting Williams, 529 U.S. at 396, 120 S.Ct. 1495). However, as in Wiggins, counsel gathered information about Holsey’s childhood from a narrow set of sources: Holsey’s school records, records from a state juvenile rehabilitation program where Holsey was sent when he was fifteen, and Holsey’s criminal history. See 539 U.S. at 523-34, 123 S.Ct. 2527. Although assistant counsel Brenda Trammel also conducted interviews of Holsey’s mother and three sisters, these interviews were limited to discussing the “guilt/innocence” phase of Holsey’s trial. At the state habeas hearing, neither lead counsel Andy Prince nor Trammel recalled asking Holsey himself about his childhood. Moreover, although the abuse suffered by Holsey was severe and the knowledge of it widespread among community members who were available and willing to testify as mitigation witnesses, most of these witnesses were never contacted by trial counsel. See Cooper, 646 F.3d at 1352 (noting list of witnesses who were willing to testify and not contacted). Those individuals who were contacted simply were never asked for information about Holsey’s upbringing. Even though counsel relied on Regina Holsey to present evidence about Holsey’s family history, trial counsel never discussed the scope or subject matter of Regina Holsey’s mitigation-phase testimony with her.

Moreover, “the information that trial counsel did acquire would have led a reasonable attorney to investigate further” and to discover the pervasive abuse in Holsey’s background. See Williams, 542 F.3d at 1340. During the course of counsel’s preparation for trial, Holsey’s four sisters and his mother provided counsel with responses to a questionnaire posing generic questions about their knowledge of Holsey and his background. Several of these questionnaires contained vivid references to the abuse and poverty in Holsey’s childhood. In particular, Regina Holsey’s questionnaire reported that Holsey “was beaten with various cords, sticks, switches, brooms, shoes,” and had been “choked” and “held under water.” Although counsel received these forms in advance of Holsey’s trial, counsel did not discuss these obvious indications of abuse in their interviews with Holsey or his sisters. See Johnson, 643 F.3d at 932 (“No reasonable attorney [having been notified of abuse] would fail to interview members of his client’s family who were readily available and could corroborate or refute the allegations of abuse.”). Nothing in the record indicates that counsel chose to cut off their investigation into Holsey’s childhood in light of signs that this inquiry would be fruitless or would lead to the discovery of adverse evidence. See Wiggins, 539 U.S. at 525, 123 S.Ct. 2527 (finding performance deficient absent indications that further research would be useless).

Like counsel’s inadequate investigation of Holsey’s childhood abuse, counsel’s investigation of Holsey’s mental condition was deficient in failing to discover readily available evidence of his borderline retardation, in ignoring prominent leads that should have sparked further inquiry, and in lacking any strategic basis to forgo this investigation. Lead counsel Andrew Prince testified at the collateral evidentiary hearing that, prior to Holsey’s sentencing, he had read the report from which Regina read at Holsey’s trial, which revealed that Holsey, as a fifteen-year-old, functioned at a third-grade level and had an IQ of 70. The report also referred to Holsey as having a “borderline mental retardation range” of intelligence and as ex*1289hibiting behavior that denoted “a prepsychotie disturbance.” Indeed, lead counsel noted in his preparation materials that Holsey was “borderline mentally retarded.” Moreover, assistant counsel Trammel testified at the evidentiary hearing that she noticed that Holsey was extremely slow from the first time she met him, and Prince documented Holsey’s demeanor as “rather psychotic.” Trammel also reviewed Holsey’s school records and shared her notes with lead counsel Prince, in which she summarized Holsey’s school performance as “pitiful.” Further, Regina Holsey’s questionnaire, which was submitted to counsel well in advance of trial, specifically referred to a paternal uncle who was mentally retarded and to mental health problems experienced by Holsey’s mother and two of his sisters. As the Georgia habeas court found, however, and as the Georgia Supreme Court did not dispute, Prince admitted that he never considered presenting any mental retardation evidence at either phase of trial. This is not surprising given his own testimony that he was drinking heavily during this time and the malpractice suit and criminal charges concerning his theft of client funds.

Based upon the mental health records that were available to Prince, “any reasonably competent attorney would have realized” that pursuing additional information about Holsey’s mental deficiency “was necessary to making an informed choice” about mitigation strategy. See Wiggins, 539 U.S. at 525, 123 S.Ct. 2527. However, Brenda Trammel, who was ostensibly in charge of mitigation, testified that by the time she joined the case, Prince had already decided that mental health issues were not going to be pursued or presented.14 This record gives no indication that Prince’s failure to pursue apparent leads into Holsey’s mental incapacity was a matter of “reasoned strategic judgment;” instead, every indication is that this defect in preparation “resulted from inattention” alone. Id. at 526, 123 S.Ct. 2527.

B. Prejudice

To satisfy the prejudice standard of Strickland, Holsey must show “a reasonable probability that ... the sentencer ... would have concluded that the balance of aggravating and mitigating circumstances did not warrant death” if mitigating evidence of his childhood abuse and mental retardation had been introduced. See Strickland, 466 U.S. at 695, 104 S.Ct. 2052. Because Georgia requires that the death penalty may be imposed only by a unanimous jury verdict,15 Holsey need only show a “reasonable probability that at least one juror would have struck a different balance” between aggravating and mitigating factors. Wiggins, 539 U.S. at 537, 123 S.Ct. 2527. “A reasonable probability is a probability sufficient to undermine confidence in the outcome” of Holsey’s original sentencing proceeding. Strickland, 466 U.S. at 694, 104 S.Ct. 2052. The likelihood of a different outcome must be “substan*1290tial,” but Holsey need not show that his counsel’s errors “more likely than not altered the outcome.” Harrington v. Richter, -U.S.-, 131 S.Ct. 770, 791, 178 L.Ed.2d 624 (2011) (internal quotation marks omitted).

As detailed above, “[t]his is not a case in which the new evidence would barely have altered the sentencing profile” presented to the sentencing jury. Porter, 130 S.Ct. at 455 (internal quotation marks omitted). Holsey’s mitigating evidence is precisely the potent combination of child abuse and borderline mental retardation that was held to establish prejudice in Williams v. Taylor. See 529 U.S. at 398, 120 S.Ct. 1495. Moreover, neither of the state’s expert witnesses, Drs. Sachy and Einhorn, disputes that Holsey is borderline mentally retarded, both witnesses agree with Holsey’s experts that he was not “malingering” or faking the effects of retardation, and no witness has called into question Holsey’s evidence that he was the victim of severe abuse throughout his childhood and adolescence. Therefore, Holsey’s mitigating evidence is especially strong because it is “consistent, unwavering, compelling, and wholly unrebutted.” Ferrell v. Hall, 640 F.3d 1199, 1234 (11th Cir.2011).

This is not a case where the aggravating factors so clearly outweighed the mitigating circumstances that there is no substantial probability that one juror would have weighed them differently. The jury’s sentence of death was based on four statutory aggravating circumstances. The first of these factors was based on Holsey’s prior conviction for armed robbery when he was eighteen years old, in which he robbed a convenience store using a brick. The remaining three were derived from the immediate circumstances of Holsey’s crime: that the shooting of Deputy Robinson occurred while Holsey fled the scene of a robbery, that it occurred in order to evade arrest, and that Deputy Robinson was a peace officer in the line of duty. See Ga. Code Ann. § 17-10-30(b)(2), (8), (10). In addition, the state presented evidence of the non-statutory aggravating circumstances that Holsey had stabbed a man in a barroom fight and that, as Holsey attempted to flee the barroom fight, a man handed him a rifle with which he shot at a pursuer.

However, these aggravating circumstances were weakened by defense evidence at trial and would have been further diminished by the evidence presented on collateral review. As to the three statutory aggravators relating to the crime itself, the Supreme Court and our circuit both have held mitigating evidence of child abuse and mental impairment like Holsey’s to create a reasonable probability of outweighing evidence of crimes that were far more aggravated than this one. In Rompilla, the Supreme Court held that mitigating evidence of abuse and mental infirmity created a reasonable probability of outweighing aggravating evidence of murder involving torture, a simultaneous felony, and a history of violent crime including rape at knifepoint. See 542 U.S. at 378, 392-93, 124 S.Ct. 2576; id. at 402, 124 S.Ct. 2576 (Kennedy, J., dissenting). In Williams, the Court held that omitted mitigating evidence of abuse and borderline retardation may have outweighed a robbery and murder that was aggravated by the defendant’s subsequent “violent assaults on elderly victims” and arson committed while in prison. See 529 U.S. at 368-69, 398, 120 S.Ct. 1495. Likewise, in Cooper, we held that mitigating evidence of child abuse alone, without mental impairment, created a reasonable probability of outweighing evidence of the crime, which was an “extremely aggravated triple homicide” involving the execution-style murder of restrained victims. See 646 F.3d at 1338-41, 1353-56. And in Johnson, we held that mitigating child abuse *1291evidence alone, without mental impairment, created a reasonable probability of outweighing evidence of the defendant’s murder of two victims and five separate aggravating circumstances, including that the murders were committed “in a cold, calculated, and premeditated manner.” See 643 F.3d at 917.

The murder involved in this case was “no more brutal than the murder[s] in th[ese] case[s].” Id. at 937. And the culpability associated with this murder would have been diminished by the collateral expert testimony describing the effect of Holsey’s borderline retardation on his limited capacity “to understand and process information, to learn from experience, to engage in logical reasoning, or to control impulses.” Atkins, 536 U.S. at 320, 122 S.Ct. 2242. Similarly, as to the fourth aggravating factor of Holsey’s prior conviction at the age of eighteen, the weight of this aggravating circumstance would have been reduced by the collateral testimony describing the extent of the abuse that Holsey suffered throughout his adolescence, as well as his borderline mental retardation. See Jackson v. Herring, 42 F.3d 1350, 1369 (11th Cir.1995) (holding that aggravating evidence of murder would have been diminished by “[e]vidence showing the genesis of [the defendant’s] irrational rage through an abusive upbringing”).

Regarding the non-statutory aggravating evidence regarding the barroom fight, Kenneth Simmons, the person who was stabbed, testified that he “had words with” Holsey about a fight between Holsey and Simmons’ cousin that occurred earlier that night. According to Simmons, Holsey, otherwise unprovoked, then stabbed him in the back with a knife. Simmons testified that he could not remember what he and Holsey had said to each other. In contrast, four witnesses who were present on the night of the bar fight testified that Holsey did not start the fight and stabbed Simmons only after Simmons had bludgeoned Holsey in the back of the head with a brick, causing Holsey to bleed profusely from the head. Lucille Kendrick testified that Holsey was dancing with her when Simmons came up from behind Holsey and hit him in the back of the head with an object in his hand, and Holsey started bleeding from the head. Holsey pushed Kendrick away from him, leaving some of his blood on her shirt, and started fighting Simmons.

Similarly, Clifford Holsey, the owner of the bar who is not a relative of Wayne Holsey, consistently described Kenneth Simmons and Simmons’ cousin, Scottie Simmons, as the aggressors in the fight. Specifically,' Clifford Holsey testified that Simmons, his cousin,' and an accomplice were coming from another club to Clifford’s bar to attack Holsey. When Simmons arrived, Clifford Holsey went outside to meet him and his companions and told them not to interfere with Holsey, who was already inside the bar with his girlfriend, because “he is not bothering anybody. Wayne is not going to bother you if you don’t bother him.” However, Simmons “slipped out from around me and went on inside the place,” and once inside he “went in there and jumped on Wayne. And Wayne had to defend himself because Scottie [Simmons] and his family were going to attack him.”

Further, Belinda Hawkins, Holsey’s girlfriend who was with him on the night of the altercation, testified that Simmons and two others had come to the club to attack Wayne because of an incident that had occurred earlier that night. She remembered Holsey coming to her in the bar with blood running from his head. She asked him what happened, and Holsey told her Simmons had hit him with a brick from behind. Like Clifford Holsey and Lucille Kendrick, she testified that Wayne *1292Holsey did not start the fight. Finally, all of this testimony was corroborated by Regina Reeves, who testified at trial that Holsey had “three holes in his head” as a result of the bar fight, one on the temple, and two on the back side of his head, all of which required stitches. Bertha Simmons, who is not related to Kenneth or Scottie Simmons, testified that she did not know who started the fight but did see Holsey stab Kenneth Simmons four times.

We cannot say that a reasonable jury would weigh the evidence of the bar fight heavily against Holsey. Rather, the weight of the testimony showed that Simmons attacked Holsey first from behind and that Simmons came to the bar with two companions for the specific purpose of assaulting Holsey. The weight of the testimony revealed that Holsey stabbed Simmons in the midst of a fight that started after Simmons inflicted a serious head wound on Holsey.

As for the evidence that Holsey fired a rifle at Kenneth’s cousin, Scottie Simmons, the aggravating weight of this testimony was diminished by Scottie Simmons’ admission that he was pursuing Holsey in order to attack him as Holsey left the bar. Moreover, Holsey’s experts testified on collateral review that Holsey is prone to follow the instructions from others as a consequence of his mental retardation, and Holsey fired the rifle at Simmons only after a third person handed the rifle to Holsey and instructed him on how to cock, aim, and fire it as Simmons approached. See Ferrell, 640 F.3d at 1234 (holding prejudice established in part where “evidence of [the petitioner’s] mental illness measurably weakens the aggravating circumstances” by providing an explanation for the petitioner’s conduct); Hardwick v. Crosby, 320 F.3d 1127, 1185 (11th Cir.2003) (“[Psychiatric mitigating evidence not only can act in mitigation, it also could significantly weaken the aggravating factors.”) (internal quotation marks omitted).

Although the majority claims that testimony from defense psychologists would have led to introduction of testimony by Dr. Shapiro and Dr. Sachy that Holsey’s conduct “evidenced an antisocial personality disorder,”16 neither doctor gave an opinion that Holsey exhibited this disorder. At most, Dr. Sachy opined, hypothetically and in the abstract, that aggravated assault, fighting, and successful armed robbery were each individually more consistent with antisocial personality than with mental retardation. However, Dr. Sachy made no diagnosis, and gave no opinion, about Holsey, and made no conclusions based on any review of Holsey’s conduct in particular. For his part, Dr. Shapiro repeatedly refused to make an antisocial personality diagnosis or to give any opinion about antisocial personality disorder based on the information available to him, despite counsel for the state’s persistent questioning.

Furthermore, many aspects of Dr. Sachy’s testimony would have been undermined on cross examination or rebutted by Holsey’s experts. Specifically, although Dr. Sachy testified that Holsey does not show signs of “gross neurological disfunctioning” or physical abnormality, and that Holsey’s cognitive functioning was “grossly intact,” he admitted on cross examination that none of these features is necessarily correlated with mental retardation. In addition, Dr. Sachy readily admitted that, because he is a psychiatrist and not a psychologist, he was unable to criticize any of the testing conducted by Drs. Cunningham and Toomer, and even Dr. Einhorn, all of which registered scores falling within the range of mental retardation.

Moreover, Holsey’s expert Dr. Cunningham was prepared to offer persua*1293sive criticism of Dr. Sachy’s assessment that would have pointed out flaws in his methodology and, overall, described it as relying on scientifically and statistically inappropriate, selective information. In particular, Dr. Sachy testified that his opinion was based on Holsey’s physical appearance17 and on Holsey’s use of words that Dr. Sachy deemed “complex,” which he explained were the “single most significant thing” that he relied upon to form his diagnosis.18 However, Dr. Sachy admitted that his opinion that these words are contraindicative of mental retardation was based on his own subjective judgment rather than on any normed survey, such as the testing that formed the basis of Holsey’s experts’ opinions. Asked to justify his reliance on vocabulary as a measure of mental retardation, Dr. Sachy stated, “what separates us from lower animals than our language skills [sic]? ... I mean, people say when you open your mouth that’s when other people find out you’re stupid.” At several other points, Dr. Sachy related his view that Holsey’s conduct during his crime was not indicative of mental retardation in part because the conduct was not “stupid” or was “non-silly.” He also stated that, in conducting his assessment, he was “looking for ... blatant, silly, juvenile, mentally retarded ... ways of behaving that make a person dysfunctional .... I think people used to call them [mentally retarded people] simple ... That was the terminology at the time. And I just did fail to find that in the record, anything significant indicating a real simplicity.” This testimony corroborates Dr. Cunningham’s criticism that Dr. Sachy relied on scientifically inappropriate anecdotes, including stereotypes about mentally retarded persons, rather than on a statistically or scientifically valid method by which to assess mental retardation. Dr. Cunningham testified that Dr. Sachy’s reliance on vocabulary is not an appropriate basis on which to diagnose mental retardation because, whereas “systematic assessment of vocabulary is one of ten or eleven subtests in the full scale IQ test, ... unsystematic subjective evaluation is not at all a reliable way of diagnosing mental retardation.” Therefore, excluding a diagnosis of mental retardation based on anecdotal evidence of vocabulary “would be an extraordinarily inappropriate application to make.” In contrast, neither Dr. Sachy nor Dr. Einhorn offered any criticism of the methodology and assessments conducted by Drs. Cunningham and Toomer. Introducing Holsey’s mitigating evidence of child abuse and borderline retardation would not have led to the introduction of significant aggravating evidence beyond what the jury heard during Holsey’s original sentencing.19

*1294Because we must presume that the jury evaluating the evidence is acting “reasonably, conscientiously, and impartially,” see Strickland, 466 U.S. at 695, 104 S.Ct. 2052, I cannot believe that one juror hearing all of the mitigating evidence would not credit Holsey’s experts and lay witnesses and find Holsey to be either fully mentally retarded or borderline mentally retarded and so diminished in his cognitive and behavioral capacity as to be either ineligible for or undeserving of the death penalty. When combined with Holsey’s evidence of his horrific child abuse, none of which was presented to his sentencing jury, there is a substantial probability that one juror would not have voted in favor of the death penalty had this evidence been introduced by competent counsel. See Wiggins, 539 U.S. at 537, 123 S.Ct. 2527; Williams, 529 U.S. at 398, 120 S.Ct. 1495. Accordingly, the Sixth Amendment requires that Holsey receive a new sentencing hearing.

. See Ga.Code Ann. § 17-10-31(c) ("If the jury is unable to reach a unanimous verdict as to sentence, the judge shall dismiss the jury and shall impose a sentence of either life imprisonment or imprisonment for life without parole.”); Humphrey v. Morrow, 289 Ga. 864, 717 S.E.2d 168, 173 (2011) ("Georgia’s death penalty laws ... provide for an automatic sentence less than death if the jury is unable to reach a unanimous sentencing verdict.....”).

. Holsey goes by his middle name, "Wayne,” but is also referred to by several witnesses by his first name, "Robert.”

. The majority refers to a two-page "home evaluation” created by a social worker at Georgia’s Youth Development Center, which Holsey attended when he was fifteen years old. See Majority op. at 1264-65. The report contains a two-sentence summary of the social worker’s "impressions” of Holsey’s *1277mother, in which the social worker states his "doubts as to her ability to cope with Angela and Robert [Holsey],” and that Holsey’s mother "loves and cares for [the children] a great deal, but she has no idea how to control them without resorting to excessive punishment.” The report also includes positive commentary on Holsey’s mother’s "parental commitment," including that "she is willing to give Robert all the encouragement that he needs” and that "[s]he foresees no problems with Robert and his sister, Angela, returning from the Youth Development Center at the same time.” Taken in the context in which it was presented to the jury and given the equivocal nature of the social worker’s assessment of Mary Holsey, this document cannot reasonably be characterized as "highlighting” Holsey’s abusive background, much less as "largely cumulative” of the evidence introduced by collateral counsel, especially when there is no indication that trial counsel ever made the jury aware of the report's assessment of Mary Holsey’s parenting ability. See Schofield v. Holsey, 281 Ga. 809, 642 S.E.2d 56, 61-62 (2007).

. As demonstrated by the facts described above, I cannot agree with the majority's characterization of Holsey’s case as different because Cooper presented “a different story — ■ not just a less detailed one” than what was presented on collateral review. See Majority op. at 1267. The difference between the sentencing evidence and collateral evidence in Cooper is no greater than the disparity between the sentencing evidence and collateral evidence at issue here.

. The majority asserts that Cooper is in conflict with Cullen v. Pinholster, -U.S. -, 131 S.Ct. 1388, 179 L.Ed.2d 557 (2011), because the majority reads Cullen as establishing that § 2254(d)(2) cannot apply to a state court’s characterization of evidence as cumulative. See Majority op. at 1259-60. However, § 2254(d)(2) applies to an appellate court’s construction and characterization of the facts in the record, including mixed questions of law and fact. See, e.g., Harvey, 629 F.3d at 1252 (applying § 2254(d)(2) to state court's comparison of factual content of two statements); Gore v. Sec’y, Dep’t of Corr., 492 F.3d 1273, 1297-98 (11th Cir.2007) (applying § 2254(d)(2) to "mixed question of fact and law” of whether defendant’s Miranda rights were "scrupulously honored”). Moreover, Cullen does not address, let alone hold, that § 2254(d)(2) does not apply to a state court’s characterization of evidence as cumulative, and thus, cannot conflict with Cooper. See United States v. Kaley, 579 F.3d 1246, 1255 (11th Cir.2009).

. Indeed, the court in Sochor noted the original sentencing testimony as follows:

Sochor’s brother, Gary, testified that his "whole childhood was memories of nothing but getting beat” by his parents, and that Sochor "took a lot of ... beatings” that were intended for Gary. Gary testified that he “remember[ed] my mother, when [his] dad would get home, yelling at my dad to spank us for this or for that, and he would start in with the belt, I mean a big belt, and just beat, beat, beat.” Sochor's sister, Cathy Cooper, testified that Sochor "had a pretty rough life” and that all of their parents’ “frustrations and everything got taken out on” Sochor. Cooper stated that their father "used to be a boxer," "knew how to hit,” and had a "very quick and violent temper.” Cooper testified that "[t]here were times where literally you’d have to pull [her father] off of” Sochor. Sochor's father “constantly hit [Sochor] with his fist. He’d hit him in the face, in the arms, and ... would just be in a rage, anyplace he could hit him.” Cooper testified that Sochor “constantly had his lips split open, black eyes, bruises all over his body.” Cooper stated that she "remember[ed] very clearly that [her] dad had [Sochor] down on the floor and was strattled over the top of him and just, you know, like plummeting with his fist.” Sochor’s father "got ahold *1282of [Sochor's] hair, and he kept banging his head against the wall.'' Sochor “just sort of slid to the floor, and then dad just started kicking him.”

Sochor, 685 F.3d at 1022-22. Sochor's parents also testified at the original sentencing hearing, describing his abuse in detail. See id. at 1022.

. Similarly, Robinson v. Moore, 300 F.3d 1320 (11th Cir.2002), is not comparable. Unlike Holsey's case but like Sochor, the actual sentencing phase testimony provided an extensive description of specific instances of abuse and the additional details only presented other instances of abuse. Compare Robinson, 300 F.3d at 1329-30 (describing sentencing phase testimony as explaining that the petitioner was exposed to "considerable physical abuse” such as being beaten with a belt or switch while bound at the wrists, being forced to sit with a broom handle between his legs while being struck by his grandfather, that neighbors observed him with bruises and told him not to go home, and that petitioner was sexually abused by relatives and while working in a migrant labor camp), with id. at 1338-39 (describing collateral child abuse evidence as consisting of evidence that the defendant was also beaten with “belts, electric cords, or whatever,” and would be beaten if he stopped picking crops for his grandfather). A holding that a petitioner was not prejudiced by counsel’s failure to present "more details" or "different examples” of abuse does not remotely suggest that it is reasonable to discount evidence of child abuse where no details or examples that would enable the jury to understand the nature and extent of this abuse were presented at sentencing.

. Nor is the remainder of the cases cited by the majority comparable to this one. In Wong v. Belmontes, 558 U.S. 15, 130 S.Ct. 383, 175 L.Ed.2d 328 (2009), the petitioner did not offer evidence that he was a victim of child abuse on collateral review, only that his family “lived in a state of ‘constant strife’ " that was not directed at him. See Wong, 130 S.Ct. at 388. And the sentencing phase testimony provided a detailed account of that violence. See Belmontes v. Ayers, 529 F.3d 834, 885 (9th Cir.2008) (O’Scannlain, J., dissenting), rev’d, 558 U.S. 15, 130 S.Ct. 383, 175 L.Ed.2d 328 (recounting that, at trial, "Belmontes's mother spoke of how Belmontes's father used to beat her, once breaking her arm, and another time stabbing her *1283and of how Belmontes suffered from the departure of her second husband and became 'difficult to control.’ ”). In Boyd v. Allen, 592 F.3d 1274 (11th Cir.2010), we distinguished the facts of Wiggins v. Smith and Williams v. Taylor and discounted the petitioner's mitigating evidence because, unlike Wiggins and Williams, this evidence did not clearly reveal that the petitioner himself was regularly exposed to abuse, and because “the evidence [was] mixed on the impoverished conditions found in [the petitioner’s] home.” See Boyd, 592 F.3d at 1300; see also id. at 1299-1300 ("There is also no evidence to suggest that Boyd’s father was ever directly violent towards him, and it is unclear how much of his stepfather’s violence was directed at Boyd himself.”). Moreover, we repeatedly emphasized throughout our decision that discounting this limited evidence of child abuse was justified because of the horrific nature of the petitioner's crime, which involved the premeditated torture and murder of two victims followed by deliberate defacement of their corpses. See id. at 1302-03, 1302 n. 7. Accordingly we stressed that “in brutal torture-murder cases like this one, this Court generally has not found Strickland prejudice, even where the petitioner’s counsel may have performed deficiently by failing to uncover and present evidence of troubled and abusive childhoods.” Id. at 1301. Finally, Rhode v. Hall, 582 F.3d 1273 (11th Cir.2009) (per curiam), did not include any claim that trial counsel was deficient in failing to introduce evidence of abuse in the petitioner's background, and so is not relevant to the reasonableness of the Georgia Supreme Court’s determination that Holsey’s child abuse evidence is "largely cumulative” of his collateral evidence. See Rhode, 582 F.3d at 1284 (summarizing evidence that petitioner alleged counsel deficiently failed to introduce).

. When Dr. Shapiro performed a partial IQ test of Holsey before his trial, he obtained an IQ score of 79, however, Dr. Shapiro nevertheless agreed with every other expert to have examined Holsey that Holsey’s intelligence falls at least in the borderline range of mental retardation. Moreover, as explained by Dr. Cunningham, the results of that test were not as reliable as the tests administered by the YDC and by Drs. Cunningham, Einhorn, and Toomer, because Dr. Shapiro administered only an incomplete subset of the tests administered on these other occasions. The only other contrary test administered to Holsey, the Culture Fair test administered by the Georgia Department of Corrections, was described by the psychologist who designed the test, Dr. Herbert Eber, as "not intended to be used for psychological diagnosis” and that those scores “have absolutely no validity for assessing IQ as defined by the DSM and are not valid for purposes of diagnosing or ruling out mental retardation.”

. Although the majority apparently discredits Dr. Cunningham’s description of Holsey’s mental retardation because those precise words are not also found in the DSM-IV, see Majority op. at 1242-43 n.7, even Dr. Sachy, the state expert witness whose testimony the majority finds convincing, stated during his deposition that he does not always follow the letter of the-DSM-IV in diagnosing patients. Specifically, Dr. Sachy testified that the DSM-IV is, at most, "a good starting point for arriving at a diagnosis,” but is not to be treated as a "diagnostic bible,” and that he disagreed with its assessments in some areas. In any event, no witness for the state questioned the scientific basis for Dr. Cunningham’s use of the term "catastrophic disability” in relation to borderline mental retardation.

.However, Dr. Cunningham explained that mental retardation is not contingent upon biological or neurological impairment, and can be the result of so-called "cultural deprivation” because "[m]ental retardation is independent of cause. And so whether or not your hardware was permanently stunted because you didn’t get proper nurturance as a baby, because you were neglected or abused or were impoverished or ate lead paint or were dropped on your head ... all of that is called mental retardation.” Dr. Cunningham described Dr. Einhorn’s assumption that a "cultural deficit” and "mental retardation” are mutually exclusive as "entirely inconsis*1286ten! ... with DSM-IV and with the American Association of Mental Retardation Standards” and is "not a professionally accepted stance or viewpoint.” Similarly, Dr. Toomer testified that "when we talk about diagnosing mental retardation in individuals, the etiology is varied. And because ... it is psychosocial or what have you does not in any way diminish the existence or the likelihood of the existence of mental retardation.”

. It is not inconsistent with Holsey’s evidence that he maintained menial jobs as a tray-loader at a chicken processing plant or as a dishwasher at a Pizza Hut restaurant because of the extraordinarily simple, repetitive nature of both jobs. Or, as Dr. Cunningham explained, it was not surprising that Holsey managed to keep these jobs despite his impairments because "mentally retarded individuals, if the task is simple and routine and they have some guidance, they're likely to be excellent employees in that area because it fully engages them and they get a sense of pride in doing that.”

. "[I]t is abundantly clear that an individual 'right on the edge’ of mental retardation suffers some of the same limitations of reasoning, understanding, and impulse control as those described by the Supreme Court in Atkins." Brownlee, 306 F.3d at 1047.

. Prince obtained funds from the court to pay for a psychological assessment by Dr. Michael Shapiro, but, as in Ferrell v. Hall, counsel's use of Dr. Shapiro was "unjustifiably and unreasonably circumscribed.” 640 F.3d 1199, 1227 (11th Cir.2011). Prince used Dr. Shapiro only to assess competency to stand trial and did not inquire about mental limitations relevant to mitigation. Shapiro testified that counsel provided him with only a two-page summary of Holsey’s personal and family history. Dr. Shapiro created no report of his findings, and did not make any diagnosis as to Holsey’s mental condition in advance of trial. Whether or not counsel ever discussed Shapiro’s examination with him, the limited evidence of counsel’s interaction with Dr. Shapiro reveals that Shapiro’s investigation of Holsey’s mental condition was, at best, a "sharply limited inquiry.” Id.

. See supra n. 1.

. Majority op. at 1270.

. In response, Dr. Cunningham testified that abnormal physical characteristics are generally indicative only of severe mental retardation and not relevant to Holsey's contention that he is mildly mentally retarded.

. The words or statements that Dr. Sachy identified were the words “forbidden'' (pronounced by Holsey as “forbidded”), "subpoena,” “animosity,” "coverup,” and the phrases "unconditional love,” “don’t know if I'm coming or going,” "things happen for a reason,” and "my life is at stake.”

.The majority references Holsey's arrests in 1982 and 1983 for simple battery and shoplifting, and in 1990 for carrying a concealed firearm, however, both of the earlier arrests occurred during Holsey’s abuse- and poverty-stricken adolescence, and the 1990 concealed-weapon charges were dropped. And although Holsey’s prison records reflect some disciplinary violations, the records also contain mitigating statements about his conduct, such as those from his activity counselor that Holsey "appears honestly interested in staying out of any trouble” and "is a quiet individual [who] appears to stay pretty much to himself.”