Sammie Stokes v. Bryan Stirling

PUBLISHED

UNITED STATES COURT OF APPEALS
FOR THE FOURTH CIRCUIT

No. 18-6

SAMMIE LOUIS STOKES,

Petitioner – Appellant,

v.

BRYAN P. STIRLING, Director, South Carolina Department of Corrections;
MICHAEL STEPHAN, Warden of Broad River Correctional Institution,

Respondents – Appellees.

Appeal from the United States District Court for the District of South Carolina, at Aiken.
R. Bryan Harwell, Chief District Judge. (1:16-cv-00845-RBH)

Argued: May 6, 2021 Decided: August 19, 2021

Before GREGORY, Chief Judge, HARRIS, and QUATTLEBAUM, Circuit Judges.

Reversed and remanded by published opinion. Chief Judge Gregory wrote the opinion, in
which Judge Harris joined. Judge Quattlebaum wrote a dissenting opinion.

ARGUED: Paul Alessio Mezzina, KING & SPALDING LLP, Washington, D.C., for
Appellant. Michael Douglas Ross, OFFICE OF THE ATTORNEY GENERAL OF SOUTH
CAROLINA, Columbia, South Carolina, for Appellees. ON BRIEF: Diana L. Holt, DIANA
L. HOLT, LLC, Columbia, South Carolina; Michele J. Brace, VIRGINIA CAPITAL
REPRESENTATION RESOURCE CENTER, Charlottesville, Virginia; Ashley C. Parrish,
Joshua C. Toll, Isra J. Bhatty, Edward A. Benoit, KING & SPALDING LLP, Washington,
D.C., for Appellant. Alan Wilson, Attorney General, Donald J. Zelenka, Deputy Attorney
General, Melody J. Brown, Senior Assistant Deputy Attorney General, OFFICE OF THE
ATTORNEY GENERAL OF SOUTH CAROLINA, Columbia, South Carolina, for
Appellees.

2
GREGORY, Chief Judge:

Sammie Louis Stokes confessed to capital murder, putting mitigation of the death

penalty at the heart of his defense. His trial counsel prepared some personal mitigation

evidence but, at the last minute, withheld it. Instead, counsel presented a single witness at

sentencing: a retired prison warden who was unprepared and counterproductive. The jury

returned a death sentence without hearing a word from the defense about Stokes as an

individual. In postconviction proceedings, new counsel found more information about

Stokes’s traumatic upbringing, but failed to pursue a mitigation-based ineffective

assistance of counsel claim. We conclude that postconviction counsel were ineffective,

providing good cause for Stokes’s procedural default of such a claim. On the merits, we

find that trial counsel’s failure to adequately investigate and present personal evidence was

objectively unreasonable and prejudicial. We reverse the district court order dismissing

Stokes’s petition and remand for issuance of the writ unless the State grants resentencing.

I.

A.

Stokes’s childhood in Branchville, South Carolina was marked by extreme abuse

and neglect. His parents were serious alcoholics. Stokes initially lived with his father. He

met his mother, Pearl, for the first time when he was four years old. When he was five,

Stokes went to live with Pearl and met his sister Sara for the first time.

Pearl lived with a man, Richard, whom the kids regarded as a stepfather. As one

relative put it, the family lived in a “run-down wooden shack” without running water or

3
indoor plumbing. Richard and Pearl sometimes took the children along to clubs and bars,

and other times the children were left unsupervised. Stokes and Sara often skipped school

and sometimes stole food from neighbors to eat. On some weekends, they stayed with their

paternal grandmother, who ran a liquor house and brothel out of her home. When Stokes

was nine years old, his father died suddenly on the front lawn, where Stokes saw his body.

Afterwards, the kids lived with Pearl and Richard permanently.

The children witnessed and suffered physical and sexual abuse. Richard sexually

abused Sara, regularly and openly. When Sara was as young as 13 years old, Pearl

sometimes “gave” Sara to men as “payment” in exchange for car rides. Stokes was

disciplined by whippings with an extension cord. Pearl and Richard fought explosively.

One witness recalled Pearl being hospitalized after Richard broke a bottle over her head;

Stokes recalled Richard breaking her jaw. When Stokes was 13, he witnessed his mother,

on the couch, intoxicated, as she fell into a coma and then died, leaving Stokes parentless.

Stokes remembers his mother’s death as a point when his life turned for the worse.

Stokes and Sara briefly lived with an aunt but ultimately chose to live unsupervised with

Richard. Stokes began using drugs and alcohol. He attended under-resourced schools

where his failure to progress was ignored. Stokes repeated the eighth grade three times,

yet only stopped attending school at age 18, when he was in the ninth grade. Around age

11 or 12, Stokes had been sexually abused by a babysitter. Thereafter, he had many sexual

encounters, and at age 15, impregnated two partners. At that same age, a “relationship”

began between Stokes and Audrey Smith, a friend of his mother’s, who was almost ten

4
years older than him. Stokes was “obsessed” with Smith, and their relationship was often

tumultuous. When Stokes was 18 and Smith was 27, they married.

According to the child development expert retained by Stokes’s federal counsel,

these facts amount to an extraordinarily traumatic childhood that impaired Stokes’s future

emotional regulation and social adaptation. 1

B.

In 1988, Stokes was convicted of assaulting Smith with a knife. Soon after his

release in 1990, the couple became involved again. Before long, Stokes assaulted Smith

for a second time, choking her in a park and leaving her unconscious. He was convicted

of that assault in 1991 and sentenced to ten years. While serving the sentence, in 1998,

Stokes was cellmates with Roy Toothe. Toothe’s mother, Pattie Syphrette, lived with his

children and their mother, Connie Snipes. Syphrette wanted to gain custody of her

grandchildren by having Snipes killed. Stokes agreed to carry out the murder for $2,000.

Stokes was released from prison several months later. Within weeks, he and

Syphrette met to make plans. Syphrette falsely told Snipes that she had kidnapped and

planned to murder Doug Ferguson, a man who sometimes lived with them. Syphrette

invited Snipes to join, and Snipes agreed. Stokes also invited Norris Martin, a longtime

1
The expert, Dr. James Garbarino, applied the Centers for Disease Control and
Prevention’s well-known “Adverse Childhood Experiences” (ACE) standard to quantify
the adversity Stokes experienced on a ten-point scale. According to Dr. Garbarino, only
13 percent of the population have an ACE score of four or greater, with less than one
percent scoring seven or greater. Based on his evaluation, Dr. Garbarino concluded Stokes
has an ACE score of nine, meaning he was exposed to more childhood adversity than 999
out of 1,000 individuals on average. See generally J.A. 2173–2201.
5
friend from childhood. Martin, who has an intellectual disability, was a “follower” of

Stokes growing up. Snipes went with Syphrette, Stokes, and Martin on a drive to an

isolated area. While Syphrette waited by the car, Snipes walked with Stokes and Martin

into the woods. Though the factual accounts differ about exactly what happened next, it is

undisputed that Stokes drew a gun and held Snipes at gunpoint; Martin raped Snipes,

followed by Stokes; and Martin and Stokes each shot Snipes once in the head, killing her.

When a car passed nearby, they fled, leaving the body in the woods where it was found

days later. At the scene, police found a hat, knife, and wallet belonging to Martin. The

group was arrested soon after, and Stokes penned a detailed confession in county jail.

Martin later testified to further details about the crime. According to Martin, Stokes

instigated the rape and murder of Snipes; Stokes was especially abusive, anally raping

Snipes and using Martin’s knife to mutilate her breasts; Stokes pushed the gun into

Martin’s hand and forced him to pull the trigger; and Stokes mutilated the corpse, cutting

off a portion of the scalp and cutting off the genitals.

The jury also heard about the subsequent murder of Doug Ferguson. In the days

after the Snipes murder, Syphrette feared Ferguson’s knowledge of her plans to murder

Snipes. Syphrette enlisted Stokes and a friend, Faith Lapp, to kidnap Ferguson. The group

had bound Ferguson with duct tape when police arrived at Syphrette’s home. Ferguson

died of suffocation from his bindings. In a subsequent prosecution, Stokes pleaded guilty

to Ferguson’s murder.

6
 C.

In 1998, the trial court appointed Thomas Sims as Stokes’s lead counsel and Virgin

Johnson as second chair. The lawyers were former prosecutors with several years of

experience in private practice. They had some limited death penalty experience, but little

to no experience preparing a mitigation defense. Trial preparation spanned nine months.

Sims and Johnson began preparing mitigation evidence six months in. Trial counsel’s

efforts on the mitigation investigation totaled around 45 hours out of the hundreds they

billed. They hired their fact investigator’s receptionist as the mitigation investigator,

though she had no prior experience with mitigation investigations. She was conducting an

interview the same day that the jury reached its initial verdict.

The guilt phase of the trial began on October 25, 1999. The parties agreed to restrict

the guilt phase to a bare recitation of the facts and reserve any aggravating facts about the

crime for the penalty phase. Four days later, the jury returned a guilty verdict after an

hour’s deliberation. For the penalty phase, trial counsel decided to focus on prison

adaptability: they planned to argue that Stokes’s health condition made him especially

suitable for a life sentence. Stokes was HIV-positive. His health was declining

significantly around this time; he even needed an emergency blood transfusion in the days

before trial. Trial counsel retained a neurologist who was prepared to testify to evidence

of brain damage possibly caused by AIDS. They also prepared a forensic psychiatrist to

testify that Stokes was likely to imminently die of AIDS in prison.

Stokes was hesitant about this strategy, and trial counsel knew it. They repeatedly

intervened to secure his consent. For example, a memorandum by the mitigation

7
investigator describes a “very tense meeting” with Stokes five days before trial, in which

Stokes opposed disclosing his HIV status. J.A. 2544. She secured his approval on the

condition that the courtroom be cleared when the issue would be discussed. Ultimately,

though, on the eve of sentencing, Stokes withdrew his consent, refusing to allow his

counsel to mention his HIV status under any circumstances.

Still, the defense team had other evidence ready. Their investigation had not

uncovered the full extent of Stokes’s life story, but they knew the broad outlines. At the

start of sentencing, Stokes’s sister and aunt were on the witness list and prepared to testify.

The same was true of the psychiatrist and neurologist, as well as a social worker who was

prepared to testify about Stokes’s psychological profile. However, in another last-minute

decision, trial counsel decided not to present any personal evidence about Stokes.

In post-conviction testimony, Sims and Johnson explained that they reached this

decision based on their impressions of the jury from the guilt phase, knowing that the worst

details of the crimes were yet to come. They believed that “an Orangeburg County jury”

“back in ’97, ’98, ’99, when this was going on,” would not be receptive to evidence about

“the background of the individual and the kind of life that they had had as a child” after

hearing the prosecution’s case in aggravation. J.A. 3469–73; J.A. 3423–25. Trial counsel

was also mindful of the jury’s racial composition and “certain inner biases” that they

believed would follow. J.A. 3523–24. They believed white jurors might react especially

to Stokes, a Black man, raping Snipes, a white woman. And for Black jurors, especially

“with the older Black females during that time,” there was “this whole idea of

homosexuality.” Id.; J.A. 3520–21 (“AIDS, homosexuality . . . during that time there was

8
a prejudice and a bias against it.”). On this basis, trial counsel declined to present any

background evidence about Stokes and proceeded with their prison adaptability approach.

But because Stokes withdrew consent to present evidence related to his HIV status, this

approach amounted to a single witness, James Aiken, offered as an adaptability expert.

Aiken was a retired prison warden. He opined that Stokes “does not demonstrate

the behaviors of being a predator” and does not demonstrate an “unusual” risk of harm in

prison. J.A. 1309–14. Aiken explained that his opinion was based on the prison facility

more than Stokes as an individual. Indeed, he refused to meet Stokes, or interview anyone

who knew Stokes, explaining that his analysis simulated how a prison official would

evaluate an inmate from their case file alone. Yet the record Aiken reviewed was

apparently incomplete, and he often could not recall the details. He emphasized the prison

system’s punitive nature, stating that if Stokes acted out, he would be punished, including

by lethal force if necessary. 2 In total, direct examination of Aiken—the entirety of the

defense’s case at sentencing—spans around four pages of trial transcript, while the State’s

cross-examination spans over 25. J.A. 1310–14; J.A. 1315–41. In closing arguments, the

State emphasized that Aiken effectively agreed that Stokes may commit violence while in

prison: “So it’s sort of like an Alice in Wonderland thing, where he’s using the punishment

2
Aiken stated: “[Prison guards] have the ability, have the technique, have the
training and have the equipment to effect lethal force if that person does not adequately
follow certain rules and regulations,” and “I have ordered inmates killed because they did
not follow rules and regulations and that inmate has been killed.” J.A. 1321. He said the
facility could “put[] [Stokes] in a prison within a prison . . . [using] lethal force and taking
his life if required.” J.A. 1327.
9
for infractions, the ability to deal with that, to say that the man is adaptable. It’s just the

opposite. If it’s adapting to prison, you don’t have to punish him.” J.A. 1365–66.

Meanwhile, the State presented robust aggravating evidence, calling 12 witnesses.

Martin and the state pathologist added further details about the violence of the Snipes

murder, and Faith Lapp testified about Stokes’s role in the Ferguson murder. The State

also presented evidence related to Stokes’s prior criminal convictions. Early in his second

prison stint, Stokes assaulted an inmate with a box cutter, and the State presented graphic

pictures of the victim and the crime scene. Smith, Stokes’s ex-wife, testified to the facts

underlying Stokes’s 1988 and 1991 assault convictions.

The evidence of Stokes’s 1991 assault of Smith is especially relevant on appeal.

Stokes’s lead trial counsel, Sims, personally prosecuted that case against Stokes. In so

doing, he developed and presented extensive testimony from Smith. Stokes had refused to

be present in the courtroom for the 1991 trial, so he had not personally witnessed Sims’s

arguments and presentation. Sims never disclosed this issue to the court. 3 He later

explained, “It was never asked, and I did not—it just didn’t come up.” J.A. 1612. It is

undisputed that Sims told Stokes about his prior role; he recalled, “[W]e did discuss with

Mr. Stokes, my role, who I was, and what my role had been in the previous matter with

him. . . . He never expressed any desire not to have me as his attorney.” J.A. 1612; see J.A.

3
Sims did point out that his name appeared on the 1991 indictment when it was
entered into evidence, requesting that it be redacted. J.A. 1426–27. However, in the brief
exchange, he states that his name was on every indictment the office issued at that time; it
does not appear that the request put the court on notice that Sims had personally prosecuted
the case against Stokes and presented extensive testimony from Smith. See id.
10
1640–56. Despite his prior prosecution of the crime now being presented by the State as

aggravating evidence against his client, Sims elected to personally cross-examine Smith.

Indeed, the second chair, Johnson, never spoke a word on the record.

After the close of evidence, the jury deliberated for around three hours before

returning a death sentence.

D.

Stokes filed an application for postconviction relief (“PCR”) in October 2001,

claiming ineffective assistance based on trial counsel’s mitigation presentation. The court

appointed PCR counsel—Keir Weyble and Robert Lominack—who filed an amended

application in May 2002, adding some additional claims. PCR counsel deposed trial

counsel, hired new experts, and hired a mitigation investigator.

In August 2004, PCR counsel filed another amended application, this time dropping

the mitigation claim while adding an Eighth Amendment intellectual disability claim and

a Sixth Amendment conflict-of-interest claim. As a result, the State initiated a formal

competency assessment process that stretched on for years. Stokes was eventually found

competent, and PCR counsel dropped the disability claim. They proceeded on their

remaining claims, including the conflict-of-interest claim and other ineffective assistance

claims but not the mitigation theory.

The PCR court denied Stokes’s application in October 2010. The parties litigated

issues related to the court’s order into 2013. Stokes petitioned for South Carolina Supreme

Court review in November 2014, which was denied in February 2016. The United States

Supreme Court also denied Stokes’s petition for review, and the State set an execution date.

11
 Stokes then filed a petition for habeas corpus, asserting two mitigation-based

ineffective assistance claims and the conflict-of-interest claim. Because the ineffectiveness

claims were not exhausted in state proceedings, a federal magistrate judge held an

evidentiary hearing to determine whether there was good cause for the default under

Martinez v. Ryan, 566 U.S. 1 (2012). After hearing testimony from both sets of counsel,

the magistrate recommended denying all relief. Stokes filed objections to the magistrate’s

report and the district court overruled them. The court adopted the report and concluded

that PCR counsel were not ineffective, meaning the unexhausted mitigation-based claims

were defaulted. Alternatively, the court found that trial counsel’s alleged ineffectiveness

was not prejudicial. Finally, the court found no actual conflict of interest, and, even if there

had been a conflict, that Stokes waived any objection.

Stokes filed a timely appeal. Under 28 U.S.C. § 2253, we may consider whether

Stokes is entitled to a certificate of appealability on his exhausted claims, and under

Martinez, 566 U.S. 1, we may determine whether Stokes is entitled to appellate review of

his defaulted claims. See Owens v. Stirling, 967 F.3d 396, 423–26 (4th Cir. 2020).

II.

In general, a state prisoner must exhaust all state court remedies before filing a 28

U.S.C. § 2254 petition. Williams v. Stirling, 914 F.3d 302, 311 (4th Cir. 2019). We then

apply the Antiterrorism and Effective Death Penalty Act of 1996 (“AEDPA”) standard of

review, under which a petitioner is entitled to relief only if the state court adjudication of

their claim was 1) “contrary to, or involved an unreasonable application of, clearly

12
established Federal law, as determined by the Supreme Court”; or 2) “based on an

unreasonable determination of the facts in light of the evidence presented.” Long v. Hooks,

972 F.3d 442, 457–58 (4th Cir. 2020) (en banc) (quoting 28 U.S.C. § 2254(d)).

Under this framework, a federal habeas court may not hear a claim that was

procedurally defaulted in state proceedings unless the petitioner can show cause for the

default. Coleman v. Thompson, 501 U.S. 722, 750 (1991). Ordinarily, an attorney’s error

is not valid cause for such a default because “[t]here is no constitutional right to an attorney

in state post-conviction proceedings.” See id. at 752–57. Ineffective assistance claims

complicate matters, however, because state law sometimes dictates that collateral post-

conviction proceedings are a defendant’s first opportunity to challenge their trial counsel’s

effectiveness. See id. at 755–57. That is the case here under South Carolina law. Sigmon

v. Stirling, 956 F.3d 183, 198 (4th Cir. 2020). In Martinez v. Ryan, the Supreme Court

adopted a narrow exception to address this gap. See 566 U.S. at 9. The Court held that—

if state law restricts ineffective assistance claims to initial-review collateral proceedings—

the ineffectiveness of a petitioner’s state PCR counsel may provide cause in a federal

habeas proceeding to excuse the petitioner’s failure to challenge the ineffectiveness of his

trial counsel. See id. “[B]ecause a petitioner raising a Martinez claim never presented the

claim in state court, a federal court considers it de novo, rather than under AEDPA’s

deferential standard of review.” Gray v. Zook, 806 F.3d 783, 789 (4th Cir. 2015).

Stokes argues the district court erred in concluding that his PCR counsel provided

constitutionally effective representation. Therefore, Stokes argues, we may reach his

underlying claim against his trial counsel. He asserts two distinct theories of trial counsel’s

13
ineffectiveness: that trial counsel unreasonably failed to investigate and present personal

mitigation evidence, and trial counsel unreasonably presented Aiken as their sole

mitigation witness. Finally, Stokes argues the state court’s conclusions as to the conflict-

of-interest claim were an unreasonable application of clearly established federal law.

III.

We first address PCR counsel’s effectiveness to determine whether Stokes has

shown good cause for defaulting his ineffectiveness claims. We conclude that PCR

counsel’s failure to develop and present a claim based on trial counsel’s mitigation efforts

amounts to ineffective assistance. Proceeding to the underlying claim, we conclude that

trial counsel’s mitigation investigation and choice not to present any personal mitigation

evidence was unreasonable and prejudicial, establishing ineffectiveness. Because these

conclusions alone require resentencing, we do not reach Stokes’s remaining claims.

A.

The Martinez exception applies when, first, the petitioner shows that “appointed

counsel in the initial-review . . . was ineffective under the standards of Strickland [v.

Washington, 466 U.S. 668, 687 (1984)].” Martinez, 566 U.S. at 14. This means that PCR

counsel “performed deficiently[] under the first prong of Strickland, . . . but not that said

counsel’s deficient performance was prejudicial[] under the second prong of Strickland.” 4

4
Asking the petitioner to show that PCR counsel’s ineffectiveness prejudiced the
state proceedings would effectively require the petitioner “to show that the defaulted claim
is itself meritorious.” Owens, 967 F.3d at 423 (explaining this “apparent incongruity”).
Circularly, the petitioner would have to “prevail on the merits of [their] underlying claim
(Continued)
14
Owens, 967 F.3d at 423. Second, the petitioner must show that the underlying

ineffectiveness claim against trial counsel “is a substantial one,” meaning that it “has some

merit.” 5 Martinez, 566 U.S. at 14.

1.

To establish ineffectiveness under Strickland, a petitioner must show counsel’s

performance was constitutionally deficient, meaning it fell below an objective standard of

reasonableness. 466 U.S. at 687. Counsel’s performance is evaluated based on “prevailing

professional norms” at the time of the representation and in light of “all the circumstances.”

Id. at 688. Professional norms may be reflected in American Bar Association (“ABA”)

standards, or other comparable guides, though such guides are not dispositive of what

constitutes reasonable representation in any given case. Owens, 967 F.3d at 412 (“[N]o

fixed set of rules may ‘take account of the variety of circumstances faced by defense

counsel.’”) (quoting Strickland, 466 U.S. at 688–89). Our assessment of counsel’s

performance is “highly deferential.” Id. “[A] fair assessment of attorney performance

merely to excuse the procedural default and obtain consideration on the merits.” Id.
(internal quotations omitted). Therefore, we have joined our sister circuits in reading
Martinez’s use of the phrase “the standards of Strickland” to refer only to performance, not
prejudice. See id.
5
In imposing this requirement, the Supreme Court cited the standard that governs
certificates of appealability under 28 U.S.C. § 2253(c)(2). See Martinez, 566 U.S. at 14
(citing Miller-El v. Cockrell, 537 U.S. 322, 327 (2003)); Owens, 967 F.3d at 423. Under
that standard, “a petitioner must ‘sho[w] that reasonable jurists could debate whether (or,
for that matter, agree that) the petition should have been resolved in a different manner or
that the issues presented were adequate to deserve encouragement to proceed further.’”
Miller-El, 537 U.S. at 327 (quoting Slack v. McDaniel, 529 U.S. 473, 484 (2000) (internal
quotations omitted)); Owens, 967 F.3d at 423.
15
requires that every effort be made to eliminate the distorting effects of hindsight . . . and to

evaluate the conduct from counsel’s perspective at the time.” Strickland, 466 U.S. at 689.

Capital defense counsel have a duty to investigate and present substantial mitigating

evidence, which includes the “obligation to conduct a thorough investigation of the

defendant’s background.” See Williams v. Taylor, 529 U.S. 362, 391–99 (2000). The

Supreme Court reaffirmed this duty several times before and during PCR counsel’s

representation of Stokes. See, e.g., id.; Wiggins v. Smith, 539 U.S. 510 (2003); Rompilla

v. Beard, 545 U.S. 374 (2005); Porter v. McCollum, 558 U.S. 30 (2009) (per curiam); Sears

v. Upton, 561 U.S. 945 (2010). Counsel’s investigation should cover the defendant’s

“psychological history,” which “could explain or lessen the client’s culpability for the

underlying offense.” See Williams v. Stirling, 914 F.3d at 313. A reviewing court considers

not only the “quantum of evidence already known to counsel,” but also whether that

evidence “would lead a reasonable attorney to investigate further.” Id. (quoting Wiggins,

539 U.S. at 527). If counsel declined to present their findings, the inquiry focuses on

“whether the investigation supporting counsel’s decision . . . was itself reasonable.”

Wiggins, 539 U.S. at 523.

Here, the district court found that PCR counsel performed reasonably because they

investigated Stokes’s background “to some extent” and then “intentional[ly]” withdrew the

mitigation claim. J.A. 3839. True, “[w]hen counsel focuses on some issues to the

exclusion of others, there is a strong presumption that [they] did so for tactical reasons

rather than through sheer neglect.” Yarborough v. Gentry, 540 U.S. 1, 5 (2003).

16
 But intentionality does not guarantee reasonableness, and presumptions are

rebuttable. The adequacy of counsel’s investigation informs the strength of the

presumption of strategy. See Strickland, 466 U.S. at 690–91 (“[S]trategic choices made

after less than complete investigations are reasonable precisely to the extent that reasonable

professional judgments support the limitations on investigation.”); Wiggins, 539 U.S. at

527–28 (“[C]ounsel were not in a position to make a reasonable strategic choice . . .

because the investigation supporting their choice was unreasonable.”). For example, in

Williams v. Stirling, counsel’s investigation uncovered evidence of the defendant’s brain

damage and his mother’s alcoholism, but counsel “did not even consider whether [the

defendant] had [fetal alcohol syndrome]” and “whether to pursue that evidence.” 914 F.3d

at 314. That counsel conducted some investigation in general was not enough; they were

ineffective because they “failed to conduct any investigation” into a potentially mitigating

condition “despite the red flags.” Id. at 315.

Similarly, in Wiggins, counsel learned of the defendant’s “alcoholic, absentee

mother” and his “physical torment [and] sexual molestation” in foster care, but they did

not pursue either discovery further. See id. at 523–25. “Counsel’s decision not to expand

their investigation” violated professional standards because they did not attempt to discover

“all reasonably available mitigating evidence.” Id. (“[A]ny reasonably competent attorney

would have realized that pursuing these leads was necessary to making an informed choice

among possible defenses, . . . .”). Thus, Wiggins rejected the “presumption” of strategy:

“[T]he ‘strategic decision’ the state courts and respondents all invoke to justify counsel’s

17
limited pursuit of mitigating evidence resembles more a post hoc rationalization of

counsel’s conduct than an accurate description of their deliberations.” Id.

The same is true here of PCR counsel. Though they investigated “to some extent,”

their investigation was nevertheless inadequate because they ignored the valuable leads

they uncovered. They knew about adversity in Stokes’s background from trial counsel’s

cursory investigation. They hired their own investigator, whose additional interviews

generated rich leads about Stokes’s psychological, educational, and familial history. At

that point, an objectively reasonable attorney would be prompted to investigate further.

See Williams v. Stirling, 914 F.3d at 313–15 (holding “there was necessarily no opportunity

for counsel to make a strategic decision” after counsel failed to “further explor[e]”

significant “red flags”). Yet PCR counsel conducted essentially no investigation beyond

the investigator’s interviews. They did not re-interview any of the witnesses themselves

or seek out corroborating documentary evidence. They did not speak to or request files

from the social worker retained by trial counsel. While they interviewed the neurologist,

who had found indicators of brain damage, they did not obtain his files or his testing results.

And, perhaps most consequentially, they did not retain an expert capable of applying

their investigator’s findings. Because “psychological and social history” and “emotional

and mental health” are often of “vital importance” to a mitigation defense, “the defense

team should include at least one person qualified to screen for mental or psychological

defects.” Id. That duty was certainly implicated in this case when the investigator found

reports of Stokes’s childhood experiences of physical and sexual abuse and neglect,

domestic violence, and substance abuse. Without consulting an expert capable of

18
analyzing these significant “red flags,” counsel did not make “efforts to discover all

reasonably available mitigating evidence.” See Wiggins, 539 U.S. at 524 (quoting ABA

Guideline § 11.4.1(C) (1989)); see, e.g., Gray v. Branker, 529 F.3d 220, 229–32 (4th Cir.

2008) (holding counsel ineffective where they “simply missed or ignored—and failed to

act on—the many signs that [the defendant] was mentally and emotionally unstable” and

failed to “explor[e] the need for mental health testimony from an expert”); Hamilton v.

Ayers, 583 F.3d 1100, 1114–17 (9th Cir. 2009) (finding ineffectiveness where indicators

of mental illness meant that counsel “should have retained a mental health expert and

provided the expert with the information needed to form an accurate profile of [the

defendant’s] mental health”).

PCR counsel themselves testified that this error explained their unreasoned

approach to the mitigation issues. When asked whether they “consult[ed] an expert to

assess [the] rather wealth of mitigation information,” Weyble responded, “No, we didn’t.”

J.A. 2718. And when asked “Did you have a strategic reason for failing to do that?” he

said, “No.” Id. Similarly, Lominack testified, “We did not hire a social worker. And I do

not think that I worked on a case before or since . . . in which I did not hire a social worker.”

J.A. 2998 (“I absolutely need someone who has the knowledge and the expertise to take

this evidence and characterize it and put it in the right boxes.”); see also J.A. 3377 (“[W]e

stopped short of putting [the findings] in front of people who could help us understand it

and generate that plausible explanation for behaviors.”). He explained that this error was

due to a lack of experience, as opposed to strategy:

19
 I look back at the cases I handled early in my career with some degree of
embarrassment. I think the most specific example is ever working on a case
without a social worker. I can’t imagine doing that at the end of my career.
And when I encountered that in other cases, it was shocking because it’s not
the standard of care and wasn’t when I worked on this case.

J.A. 2621. And while it is true that a decision not to investigate may itself be strategic—

for example, if the findings would be more harmful than helpful, see Wiggins, 539 U.S. at

525 (collecting cases)—there is no evidence that informed PCR counsel’s decisions here.

Instead, PCR counsel cited non-strategic reasons for their investigatory decisions

and abandonment of the claim. See, e.g. J.A. 2983–85 (explaining that the failure to

“personally interview[] any of those [mitigation witnesses]” was either “lazy or not

knowing that should have been done or not knowing I should have done it.”); J.A. 3018

(“We didn’t have a social worker. . . . I don’t think we had enough information at the time

to decide that it was a claim that needed to be thrown out, and I don’t think that’s what we

did, certainly not with any intentionality.”); J.A. 3031 (“We weren’t thoughtful enough to

have personally interviewed any of the people that were relevant to the claim that we

dropped.”). Because PCR counsel’s investigation fell short of professional standards, it

cannot support the presumption that their subsequent abandonment of the claim was

strategic. See Wiggins, 539 U.S. at 527–28, 536 (“[C]ounsel chose to abandon their

investigation at an unreasonable juncture, making a fully informed decision with respect to

sentencing strategy impossible.”); Williams v. Stirling, 914 F.3d at 316–17 (“[T]he PCR

court relied on the factual assumption that trial counsel made a strategic choice not to

present” mitigation evidence, but “it was impossible for trial counsel to have made a

strategic choice because there was no investigation into” that issue).

20
 Beyond the investigation’s shortcomings, PCR counsel’s testimony directly rebuts

the presumption that they dropped the mitigation claim strategically. PCR counsel testified

that they neglected the mitigation theory after becoming preoccupied with other claims.

When asked if he recalled “why [he] would have omitted that claim,” Weyble explained:

“At that same time, what was then called mental retardation [and] is now called intellectual

disability claim arose, somewhat to our surprise, frankly. And I think we became

somewhat distracted by the shiny object, if you will, and thought we—that we were really

on to something there.” J.A. 3259. Likewise, when asked “why did [he] withdraw the

mitigation claim,” Lominack responded:

[W]e were so focused at the time on the intellectual disability claim, . . .
somewhat to the detriment of . . . [a] mitigation claim. I don’t recall having
a specific reason why we would abandon a general mitigation claim,
especially when at trial nothing was presented, which is quite rare, actually,
for there not to be any mitigation presented about a client’s childhood. I
don’t recall in Mr. Stokes’s case, and frankly, in any case, how I could have
thought that that was going to be a wise or reasonable choice, and the only
thought that I have now is that we were so focused on the I.D. claim that we
dropped it and focused on that instead.

J.A. 2918. This testimony does not describe strategic prioritization among multiple claims.

Rather, counsel testified that their prioritization of other claims was uninformed and

happenstance, the product of distraction, inexperience, and carelessness. See, e.g., J.A.

3018 (“I’d love to, in hindsight, say, yeah, we knew what we were doing. We didn’t. We

were focusing on the [disability], adaptability, and IQ investigation.”); J.A. 3031 (“I think

one of the reasons I’m not remembering a real strategic reason . . . is that, to be blunt, I’m

not sure that we were that thoughtful about it.”); cf. Wood v. Allen, 558 U.S. 290, 309

(2010) (Stevens, J., dissenting) (distinguishing a decision that is “the product of a deliberate

21
choice between two permissible alternatives” from one that is “the product of inattention

and neglect by attorneys preoccupied with other concerns”).

Also, Stokes was ultimately found competent years later, and PCR counsel dropped

the disability claim that had been hogging their attention. PCR counsel did not testify to

any strategic basis for declining to pursue the mitigation theory at that point. See, e.g., J.A.

2993 (“[T]o me, it’s nonsensical to be so hyper-focused [on intellectual disability], . . .

especially when th[at claim] went by the wayside after the [state] evaluation.”); J.A. 3262

(responding, after being asked why they did not revive the mitigation claim after the

intellectual disability claim failed, “I don’t have a good answer to that question”). 6 Nothing

suggests that counsel was strategically winnowing the claims, selecting only those few

deemed most meritorious: In their original petition, PCR counsel raised six claims in

6
The district court largely ignored PCR counsel’s testimony. It relied on one
exchange from Weyble’s cross-examination, where Weyble responded “yes” to the
prosecutor’s statement that “there had to be a reason that [he] withdrew [the claim]”
because if he “thought it was a strong claim,” he would have presented it. See J.A. 3840–
41. The court discredited the countervailing portions of PCR counsel’s testimony as
“fall[ing] on their sword for their former client.” J.A. 3840. But nothing in the record
justifies selectively crediting this exchange while disregarding extensive contrary evidence
from the same witnesses. An evidentiary ruling may not rest solely on the court’s
presumption of the witness’s sympathies and intentions. For example, the dissent the
district court cited as its sole authority called out perceived “sword falling” in a footnote,
but it supported that charge by arguing the witness’s testimony about “the particulars of his
investigation” contradicted his “self-denigrating” characterizations. Dugas v. Coplan, 428
F.3d 317, 346 n.39 (1st Cir. 2005) (Howard, J., dissenting). Here, the district court pointed
to no factual contradictions in Weyble and Lominack’s testimony that undermined their
generally negative portrayal of their performance. They gave reasonable explanations for
why they made decisions they now believed to be improper, such as a lack of experience
or being distracted by new developments. The district court did not identify, nor do we
find, anything in the record making PCR counsel any less credible than trial counsel, whose
testimony the district court relied on extensively.
22
addition to the intellectual disability claim, three of which were so weak that they later

abandoned them and two of which, alleging ineffective assistance of appellate counsel, had

obvious and dispositive weaknesses. 7 Considering the dramatic lack of mitigation

evidence presented at trial despite Stokes’s background, PCR counsel’s failure to consider

adding the mitigation claim back into the petition is further evidence of unreasonableness,

not strategy. See McKee v. United States, 167 F.3d 103, 106 (2d Cir. 1999) (“A petitioner

may rebut the suggestion that the challenged conduct reflected merely a [tactical] choice

. . . by showing that counsel omitted significant and obvious issues while pursuing issues

that were clearly and significantly weaker.”). Here, like in Wiggins, the purportedly

strategic decision “resembles more a post hoc rationalization of counsel’s conduct than an

accurate description of their deliberations.” See 539 U.S. at 523–25.

Absent strategic justifications, PCR counsel’s abandonment of the mitigation claim

was objectively unreasonable. PCR counsel knew that trial counsel’s investigation was

paltry. Further, trial counsel presented no background mitigation evidence at all, and the

Supreme Court had recently deemed trial counsel ineffective for even more robust

presentations. See, e.g., Williams v. Taylor, 529 U.S. at 368, 395 (finding mitigation

presentation ineffective, despite testimony from the defendant’s “mother, two neighbors,

7
As the PCR court explained, the two claims alleging ineffective assistance by
appellate counsel were without merit because they improperly faulted appellate counsel for
failing to argue points that had not been preserved at trial. Additionally, the PCR court
noted that one of the claims—that appellate counsel should have argued that the trial court
erred by failing to instruct on a mitigating factor for the victim’s participation or consent
in the act—was “groundless” in any event, because “Snipes did not consent to the violence
that was about to strike her.” J.A. 1775–76 n.6.
23
and . . . a psychiatrist,” because counsel failed to relay “[the defendant’s] nightmarish

childhood”); Porter, 558 U.S. at 32, (finding ineffectiveness, despite testimony about the

defendant’s relationship with his son, because counsel “failed to . . . present any evidence

of [the defendant’s] mental health . . ., his family background, or his military service”).

PCR counsel’s own failure to pursue and present a mitigation-based claim arising

from trial counsel’s performance constitutes ineffective assistance. See, e.g., Blake v.

Baker, 745 F.3d 977, 982–83 (9th Cir. 2014) (finding PCR counsel ineffective where

counsel did nothing with witness statements describing “the abhorrent conditions of [the

petitioner’s] upbringing and family history” and “failed to . . . retain experts” to review the

findings); Trevino v. Davis, 829 F.3d 328, 348–49 (5th Cir. 2016) (holding that PCR

counsel were ineffective in failing to pursue a mitigation claim where trial counsel

“presented only one mitigation witness and no other evidence,” and “[t]he deficiency in

that investigation would have been evident to any reasonably competent habeas attorney”).

2.

Given the extraordinary facts of this case, Stokes’s underlying ineffectiveness claim

against his trial counsel is “substantial” for Martinez purposes. See 566 U.S. at 14. The

basis for questioning trial counsel’s effectiveness is plain enough that PCR counsel’s

failure to adequately pursue it was objectively unreasonable. It follows that the underlying

claim has “some merit”—meaning, at the very least, reasonable jurists could debate its

24
viability. 8 See id.; Miller-El, 537 U.S. at 327. Therefore, under Martinez, Stokes has

established cause for procedurally defaulting his mitigation-based ineffective assistance

claim against trial counsel in state proceedings. Accordingly, we proceed to the merits.

B.

To establish trial counsel’s ineffectiveness, the petitioner must show that 1) their

performance fell below an objective standard of reasonableness, and 2) the deficient

performance was prejudicial. See Strickland, 466 U.S. at 687. To establish prejudice, the

petitioner must show “a reasonable probability that, but for counsel’s unprofessional errors,

the result of the proceeding would have been different.” See id. at 694.

1.

Trial counsel were constitutionally deficient for much the same reasons as PCR

counsel. Their investigation was inadequate, and their decision to withhold all personal

mitigation evidence was unreasonable.

Trial counsel had little-to-no experience preparing a mitigation defense, yet they

consulted with no experienced attorneys or mitigation experts. Their mitigation efforts,

totaling around 45 hours, began six months into their nine-month representation, though

the ABA Guidelines stated that sentencing investigation should “begin immediately upon

8
The district court also found, alternatively, that Stokes could not show prejudice
from trial counsel’s alleged ineffectiveness, and therefore “his underlying claim . . . is not
substantial.” J.A. 3849 (citing Martinez, 566 U.S. at 14). Though inconsequential, this
was an improper application of Martinez’s substantiality requirement. A claim that fails
on the merits may very well still be “substantial” for purposes of showing cause for a
procedural default. See Owens, 967 F.3d at 423. The question is whether the claim has
“some merit,” meaning that reasonable jurists could at least debate its viability. See
Martinez, 566 U.S. at 14 (citing Miller-El, 537 U.S. at 327); Owens, 967 F.3d at 423.
25
counsel’s entry into the case and should be pursued expeditiously.” ABA Guidelines

§ 11.4.1 (1989). They hired an inexperienced mitigation investigator who was still

conducting interviews through the guilt phase of the trial. They did not personally conduct

any follow-up interviews or otherwise develop the investigator’s findings. They retained

experts to testify at sentencing, but those experts were apparently not consulted about the

personal mitigation evidence. In a capital murder trial where mitigating the death penalty

was the central issue in the defense, such an investigation is objectively unreasonable. 9 See

Wiggins, 539 U.S. at 523–25 (“Despite these well-defined norms, . . . counsel abandoned

their investigation of petitioner’s background after having acquired only rudimentary

knowledge of his history from a narrow set of sources.”); see also Earp v. Ornoski, 431

F.3d 1158, 1175–76 (9th Cir. 2005) (“Wiggins . . . establishes that the presence of certain

elements in a capital defendant’s background, such as a family history of alcoholism,

abuse, and emotional problems, triggers a duty to conduct further inquiry before choosing

to cease investigating.”).

9
Trial counsel did conduct some investigation. Their investigator subpoenaed
records and interviewed several friends and family members, going beyond some
investigations that have been deemed unreasonable. See, e.g., Wiggins, 539 U.S. at 523–
24 (counsel only obtained the presentence report and one set of social services records);
Porter, 558 U.S. at 39 (counsel obtained no records and conducted no interviews of the
defendant’s family). Nevertheless, trial counsel unreasonably failed to pursue the
indications of extreme childhood trauma, neglect, and abuse. See Gray v. Branker, 529
F.3d at 229–32 (finding unreasonableness, even where counsel interviewed “friends and
associates” and retained experts, because they “failed to investigate for mental health
evidence”); Williams v. Stirling, 914 F.3d at 313–16 (faulting counsel for ignoring “red
flags,” though their investigation otherwise “did bear the hallmarks of effective
assistance”).
26
 Trial counsel’s subsequent decision to withhold the personal mitigation evidence

they did have was also objectively unreasonable. At the outset of sentencing, members of

Stokes’s family and a social worker were prepared to testify. A neurologist and psychiatrist

were prepared to testify about Stokes’s HIV status, but presumably could have offered

other personal testimony about Stokes instead. Yet counsel abandoned this evidence based

on their impressions of the jury, deciding the jurors—the Black jurors in particular—would

react negatively to evidence of Stokes’s life story after hearing the prosecution’s

aggravating case. As Johnson put it:

[H]ow do you go to a jury and say, look, we want you to look at the fact that
he had a poor upbringing, particularly African-American, which a lot of us
had struggles coming up, how do you say, well, just because he had a poor
upbringing, you need to overlook the fact that he raped this woman, you need
to overlook the fact that he cut her vagina out, you need to overlook the fact
that he cut her nipples off, you need to overlook the fact that he killed
somebody else.

J.A. 3424–25.

This concern reflects a misunderstanding of the duty to mitigate. Trial counsel were

not obliged to ask the jury to excuse Stokes’s actions. Instead, their duty was to mitigate

Stokes’s “moral culpability.” See, e.g., Williams v. Taylor, 529 U.S. at 398 (explaining

that “the graphic description of [the defendant’s] childhood, filled with abuse and privation,

. . . might well have influenced the jury’s appraisal of his moral culpability” by showing

that his violence was “compulsive” as opposed to “cold-blooded premeditation”); Caro v.

Woodford, 280 F.3d 1247, 1258 (9th Cir. 2002) (“By explaining that [the defendant’s]

behavior was physically compelled, . . . or even due to a lack of emotional control, his

moral culpability would have been reduced.”). Counsel can carry out this duty without

27
diminishing the defendant’s responsibility for their actions or the seriousness of their

crimes. See Eddings v. Oklahoma, 455 U.S. 104, 115–16 (1982) (qualifying, after

explaining why youth mitigates moral culpability, that “[a]ll of this does not suggest an

absence of responsibility for the crime of murder, deliberately committed in this case”).

As one mitigation specialist has put it:

Mitigation is not a defense to prosecution. It is not an excuse for the crime.
It is not a reason the client should “get away with it.” Instead, mitigation is
a means of introducing evidence of a disability or condition which inspires
compassion, but which offers neither justification nor excuse for the capital
crime. . . . It explains the influences that converged in the years, days, hours,
minutes, and seconds leading up to the capital crime, and how information
was processed in a damaged brain. It is a basis for compassion—not an
excuse.

Russell Stetler, The Mystery of Mitigation, 11 U. Pa. J. L. & Soc. Change 237, 261 (2008).

Thus, trial counsel did not have to ask the jury to “overlook” the graphic details of the

prosecution’s case. Instead, their personal evidence could have provided humanizing

context, allowing the jury to reach a more sympathetic understanding of the individual

behind the aggravating evidence. See Allen v. Woodford, 395 F.3d 979, 1000 (9th Cir.

2005) (“[Using] mitigation evidence to complete, deepen, or contextualize the picture of

the defendant presented by the prosecution can be crucial to persuading jurors that the life

of a capital defendant is worth saving.”).

But because trial counsel had no experience or formal training in mitigation,

conducted a shallow investigation, and failed to consult with experts, they underestimated

the value of their evidence. Stokes’s life story contains far more than a merely “difficult

upbringing” and “struggles coming up”; the evidence shows profound and chronic trauma

28
that was about as extreme as any child can experience. Such evidence is prototypical for a

personal mitigation narrative. Wiggins, 539 U.S. at 535 (referring to “severe privation and

abuse in the first six years of [the defendant’s] life” as “the kind of troubled history we

have declared relevant to assessing a defendant’s moral culpability”); Hooks v. Workman,

689 F.3d 1148, 1203–04 (10th Cir. 2012) (“[E]ven the most minimal investigation would

have uncovered a life story worth telling, . . . [which is] exactly the sort of evidence that

garners the most sympathy from jurors.”). If trial counsel believed the jury would not have

responded well to a presentation that minimized Stokes’s conduct, their duty was to find a

way to convey this highly significant evidence without doing so.

That is especially true when considered against their alternative decision: to offer

almost no mitigation presentation at all. Having declined to present their personal

mitigation evidence, trial counsel put forward one witness, the former prison warden, who

never met Stokes and was repeatedly unfamiliar with the details of Stokes’s records. Trial

counsel’s direct examination produced about four pages of trial transcript, the entirety of

their sentencing presentation. Cf. Wiggins, 539 U.S. at 526–27 (describing counsel’s

presentation, which failed to provide details of the defendant’s history, but did provide one

expert’s testimony about prison adaptability, a “halfhearted mitigation case” taking a

“shotgun approach”). Meanwhile, as anticipated, the government put forward 12 witnesses

detailing Stokes’s violence in the Snipes murder and in previous incidents. Thus, the only

evidence the jury heard about Stokes detailed his crimes and violence, distorting the jury’s

perception of “the uniqueness of the individual” facing execution. See Lockett v. Ohio,

438 U.S. 586, 604–05 (1978); see, e.g., Ferrell v. Hall, 640 F.3d 1199, 1234, 1236 (11th

29
Cir. 2011) (concluding that, because “the jury heard absolutely nothing about the

substantial mitigating evidence,” including childhood abuse, “[t]he jury labored under a

profoundly misleading picture of [the defendant’s] moral culpability”).

The basis for trial counsel’s decision—that a South Carolina jury in the 1990s, and

particularly Black people, would not be open to a personal mitigation narrative—was

objectively unreasonable. In 1989, ten years before Stokes’s trial, the Supreme Court

referred to “the belief, long held by this society, that defendants who commit criminal acts

that are attributable to a disadvantaged background, or to emotional and mental problems,

may be less culpable than defendants who have no such excuse.” Penry v. Lynaugh, 492

U.S. 302, 319 (1989) (emphasis added) (quoting California v. Brown, 479 U.S. 538, 545

(1987) (O’Connor, J., concurring)), overruled on other grounds by Atkins v. Virginia, 536

U.S. 304 (2002). Even earlier, in 1976, the Court explained: “A process that accords no

significance to relevant facets of the character and record of the individual offender . . .

excludes from consideration in fixing the ultimate punishment of death the possibility of

compassionate or mitigating factors stemming from the diverse frailties of humankind.”

Woodson v. North Carolina, 428 U.S. 280, 304 (1976). There is no reason to believe that

South Carolinians in 1999—no matter their race—would have been indifferent to the

“frailties of humankind” and these bedrock principles of mercy and morality.

Therefore, even when giving appropriate deference to trial counsel’s strategic

judgment, trial counsel’s failure to investigate and present personal mitigation evidence

was objectively unreasonable under professional standards at the time of the representation.

30
 2.

As to prejudice, “the question is whether there is a reasonable probability that,

absent the errors, the sentencer . . . would have concluded that the balance of aggravating

and mitigating circumstances did not warrant death.” Strickland, 466 U.S. at 694–95. “A

reasonable probability is a probability sufficient to undermine confidence in the outcome.”

Id. Because the sentencer in this case was a unanimous jury, prejudice “requires only ‘a

reasonable probability that at least one juror would have struck a different balance’” when

appraising Stokes’s moral culpability and deciding on death. See Andrus v. Texas, 140 S.

Ct. 1875, 1886 (2020) (quoting Wiggins, 539 U.S. at 537). To determine whether the

petitioner has made that showing, “the reviewing court must consider ‘the totality of the

available mitigation evidence,’” both at trial and from postconviction proceedings, “and

‘reweig[h] it against the evidence in aggravation.’” Id. (quoting Williams v. Taylor, 529

U.S. at 397–98).

The district court summarized the evidence of Stokes’s life story in a sentence,

referring to the death of his parents, his “abusive[] and neglectful” childhood, and “that he

struggled in school with no intervention.” J.A. 3843. It also considered the report of

Stokes’s new expert, who concluded that these “aspects of [Stokes’s] background” likely

impacted his adult behavior. J.A. 3843–44. The court then summarized the State’s case in

aggravation and concluded that Stokes failed to establish prejudice because the aggravating

evidence “was overwhelming.” J.A. 3844–48. The court especially emphasized the

“horrific circumstances” of the Snipes and Ferguson murders, noting that an additional

31
murder is recognized as “the most powerful imaginable aggravating evidence.” 10 Id.

(citing Wong v. Belmontes, 558 U.S. 1, 27–28 (2009); Morva v. Zook, 821 F.3d 517, 532

(4th Cir. 2016)). Thus, the court concluded that “all the mitigating evidence does not

outweigh all the aggravating evidence presented at trial,” and so Stokes did not show “a

reasonable probability that at least one juror would have voted against the death penalty

had it heard the additional mitigating evidence in question.” J.A. 3848–49.

We disagree. While there is no doubt that the State’s aggravation case was

extensive, the analysis is not as simple as comparing two piles of evidence and asking

which is greater. The addition of just some meaningful mitigating evidence could be

enough to sway one juror against death, even in the face of plentiful aggravating evidence.

See, e.g., Williams v. Taylor, 529 U.S. at 398 (“Mitigating evidence . . . may alter the jury’s

selection of penalty, even if it does not undermine or rebut the prosecution’s

death-eligibility case.”). And when a jury heard virtually no mitigation evidence at trial,

and nothing about the defendant as an individual, the unheard personal evidence is

especially impactful on the prejudice calculus. See Porter, 558 U.S. at 41–44.

10
In emphasizing these aspects of the Snipes murder, and the commission of the
Ferguson murder, the district court considered evidence going to the aggravating factors
that the jury did not find. The jury found the evidence established: 1) criminal sexual
conduct; 2) kidnapping; 3) murder for the purpose of receiving money; and 4) that the
defendant caused another person to participate. But the jury did not find that the State
established: 1) physical torture; or 2) that two or more persons were murdered by the
defendant. While there were surely other “horrific” elements of the Snipes murder that the
court may have been referring to, the court should not have given weight to Stokes’s alleged
torture of Snipes or his role in the Ferguson murder. Cf. Porter v. McCollum, 558 U.S. 30,
41–42 (2009) (explaining that “the weight of evidence in aggravation” was “not as
substantial as the sentencing judge thought” where one of two aggravating factors was
reversed on appeal).
32
 In Porter, the jury “heard almost nothing that would humanize [the defendant] or

allow them to accurately gauge his moral culpability.” 558 U.S. at 41–44. The jury could

have heard about the defendant’s PTSD from military service, a “childhood history of

physical abuse,” and a “brain abnormality” that caused learning disabilities. Id. “Instead,

they heard absolutely none of that evidence, evidence which ‘might well have influenced

the jury’s appraisal of [the defendant’s] moral culpability.’” Id. (quoting Williams v.

Taylor, 529 U.S. at 398). Some negative facts in the mitigation evidence were not enough

to undermine its impact because the negative aspects were still “consistent with th[e] theory

of mitigation and d[id] not impeach or diminish the evidence.” Id. The Court found the

prejudice requirement satisfied, especially because “[w]e do not require a defendant to

show ‘that counsel’s deficient conduct more likely than not altered the outcome, . . . but

rather that he establish ‘a probability sufficient to undermine confidence in [that]

outcome.’” Id. at 44 (quoting Strickland, 466 U.S. at 693–94).

The same reasoning applies here. The unheard mitigation evidence would have

shown Stokes experienced an “extraordinarily high” degree of childhood adversity, likely

surpassing 99.9% of the population. J.A. 2181–82; see generally J.A. 2173–2201. And,

through expert evidence, the defense could have explained the likely consequences of such

a childhood, connecting “chronic trauma” to brain development and adult behavior.

Instead, by presenting no personal evidence at all, counsel failed to “explain to the jury

why [Stokes] may have acted as he did . . . connect[ing] the dots between, on the one hand,

[his] mental problems, life circumstances, and personal history and, on the other, his

commission of the crime in question.” See Hooks v. Workman, 689 F.3d at 1204. As a

33
result, “jurors faced with an especially brutal crime were left with almost nothing to weigh

in the balance.” See id.

Thus, the prejudice analysis turns on the likely influence of dramatic mitigation

evidence on a jury that heard dramatically little about the defendant. In that context, the

weight of the unpresented mitigation evidence is significantly increased, enough to

outweigh even the upsetting and extensive aggravating evidence. See, e.g., Williams v.

Taylor, 539 U.S. at 537 (finding prejudice in failure to present personal evidence even

where the petitioner “had savagely beaten an elderly woman, stolen two cars, set fire to a

home, stabbed a man during a robbery, and confessed to choking two inmates and breaking

a fellow prisoner’s jaw”); see also Smith v. Stewart, 189 F.3d 1004, 1013–14 (9th Cir.

1999) (explaining that “[t]he horrific nature of the crimes . . . does not cause us to find an

absence of prejudice,” especially “where the presentation of mitigating evidence was

wholly inadequate”). If the jury could have placed Stokes’s “excruciating life history on

the mitigating side of the scale,” then “there is a reasonable probability that at least one

juror would have struck a different balance.” See Wiggins, 539 U.S. at 537; Porter, 558

U.S. at 42. Simply put, “there exists too much mitigating evidence that was not presented

to now be ignored.” Porter, 558 U.S. at 44 (quoting Porter v. Florida, 788 So.2d 917, 937

(Fla. 2001) (Anstead, J., concurring in part and dissenting in part)).

The district court failed to consider that trial counsel’s decisions meant they

presented almost no mitigating evidence. It cited cases where additional personal evidence

was insufficient to show prejudice after a jury heard a significant quantity of such evidence

at trial. See Wong, 558 U.S. at 20–21 (noting that “the mitigating evidence [the defendant]

34
did present” at sentencing “was substantial,” including nine witnesses who “highlighted

[his] terrible childhood”); Morva, 821 F.3d at 522–23 (listing mitigation efforts including

two court-appointed mental health experts, a specialist investigator, and thirteen

witnesses). The hypothetical impact of Stokes’s life story and expert evidence on a jury

that heard nothing about Stokes at all is quite distinct from the impact of a few more

witnesses on a jury that heard an already-robust mitigation presentation. Cf. Wong, 558

U.S. at 22 (finding that the proposed evidence “was merely cumulative of the humanizing

evidence [the defendant] actually presented” and “would have offered an insignificant

benefit”); Morva, 821 F.3d at 522–23, 529 (stating that counsel interviewed “many” family

members, and several testified or submitted affidavits, but the defendant “complains that

counsel could have interviewed other[s]”).

In sum, trial counsel’s unreasonable mitigation efforts prejudiced Stokes. Given

Stokes’s immediate confession, his defense turned almost exclusively on mitigation from

its very outset. Yet trial counsel spent too little time on their investigation and failed to

appreciate their findings. Then, despite the wealth of reasonably available, highly

compelling mitigation evidence, counsel told the jury effectively nothing about Stokes as

an individual. Had the jury heard the unpresented evidence, the probability that at least

one juror would have voted against death is great enough to undermine our confidence in

the outcome. See Porter, 558 U.S. at 44; Andrus, 140 S. Ct. at 1886.

We conclude that Stokes has satisfied both the deficient performance and prejudice

prongs of Strickland on his mitigation-based claim, establishing that he received ineffective

35
assistance of counsel at sentencing. Because this conclusion alone requires resentencing,

we do not reach the remaining claims.

IV.

For the foregoing reasons, we reverse the judgment of the district court and remand

with instructions that the district court issue the writ of habeas corpus unless the State of

South Carolina grants Stokes a new sentencing hearing within a reasonable time.

REVERSED AND REMANDED

36
QUATTLEBAUM, Circuit Judge, dissenting:

When considering the important skills of a trial attorney, those that might first come

to mind are the skills we see in action—opening statements, examining witnesses and

closing arguments. But as important as those skills are, just as important is a skill we don’t

see—strategic decision-making. A lawyer must make many decisions before and during

the course of a trial. And what often makes those decisions so difficult is that many cut

both ways. The decision to advance an argument, introduce certain evidence, call a witness,

cross-examine a witness aggressively or lightly and so many other decisions can be—and

often are—double-edged swords. There are pros and cons each way.

Evaluating these decisions, a lawyer must consider a litany of questions. How much

benefit can be gained? How much harm can be caused? Can the harm be mitigated? Is the

benefit that can be gained worth the harm that can result? And so on.

Oftentimes, the answers to these questions are not obvious. In fact, answering them

is more a matter of art than science because intangible factors come into play. A lawyer

must rely on experience, intuition and even gut instinct.

Making things even harder, trials are fluid. What might have made sense before trial

may become a bad idea based on events that transpire during trial. To address that fluidity,

a lawyer must constantly consider how the trial is progressing, and how the judge and jury

are responding to the evidence and arguments.

In these difficult decisions, different lawyers can, and do, come to different

conclusions. For some issues, you could ask ten lawyers and often get ten different

decisions. Nevertheless, decisions must be made.

37
 In this appeal, we review the decisions made by lawyers in the weightiest of

circumstances—the defense of a capital defendant. Although the decisions were

quintessentially strategic and informed by a thorough investigation, the majority

determines that they amounted to ineffective assistance of counsel. I disagree. These

decisions, according to our precedent, merit our highest deference. In my view, the record

does not come close to overcoming that required deference. Accordingly, I dissent.

I.

Over twenty years ago, Thomas Sims and Virgin Johnson’s client, Sammie Stokes,

was charged with the gruesome murder of Connie Snipes. The State sought the death

penalty. The evidence against Stokes was overwhelming and horrific. Stokes even

confessed to the murder. Because Sims and Johnson suspected Stokes would be convicted,

they developed a strategy they felt would be most effective during the probable sentencing

phase of the case. They wanted to emphasize Stokes’ remorse and highlight the conduct

and motivation of Norris Martin, who participated in the murder with Stokes, with the hope

that the jury would view him as “the bad guy.” J.A. 1543.

Trial counsel also planned to focus on the fact that Stokes had AIDS. Counsel

planned for experts, including a forensic psychiatrist, to talk about Stokes’ mental health

and related medical issues. Sims and Johnson felt the best way to move the jury to spare

Stokes’ life would be to point out how AIDS would debilitate Stokes and given the

knowledge about that disease at the time, that AIDS effectively was its own death sentence.

38
 But trial counsel’s preparation did not stop there. Sims and Johnson also

investigated mitigating evidence. They hired a mitigation investigator and interviewed

witnesses including Stokes’ family and friends. From that work, trial counsel learned that

Stokes’ childhood included significant trauma and abuse which might be useful in

providing some context for Stokes’ conduct.

For example, the trial defense team’s interview with Stokes’ sister revealed that

Stokes’ parents, who did not live together, both drank excessively. After their father died,

Stokes’ mother remarried. The stepfather also drank a lot but was violent as well. Stokes’

sister reported that he abused their mother. Stokes’ sister told the team that she and Stokes

had to fight their stepfather to keep him from beating their mother.

Other family members revealed similar information to the trial team. A cousin told

the trial team Stokes’ mother was “a drinker, fighter and was wild” and that her children

“grew to be the same.” J.A. 2535. Another relative told the team that Stokes’ mother and

stepfather “liked to drink, party, go to clubs and often took [Stokes and his sister].” J.A.

2536. She added that Stokes’ mother and stepfather both assaulted each other.

In addition to uncovering this evidence through their mitigation investigator, trial

counsel assembled experts to testify about the mitigation evidence if they decided to use it.

They engaged and worked with a forensic psychiatrist, a jury consultant and an expert in

social work.

But Sims and Johnson’s investigation revealed risks of utilizing this mitigation

evidence. The witnesses who would, if asked, be able to provide mitigating evidence, also

had information that was damaging to their strategy of portraying Martin as the main

39
culprit. Stokes’ sister, during the trial team’s interview of her, described Stokes as “very

moody” and that his personality would “flip.” J.A. 2529. She also talked about Stokes’

relationship with Martin. She said Stokes frequently made Martin, who had been his friend

since childhood, “hustle for him.” J.A. 2529. And Stokes was violent toward Martin if he

was not successful in those activities. More specifically, Stokes “beat [Martin] up when he

did not get his money on time.” J.A. 2529. Stokes’ sister told the trial team that Martin was

“very afraid of [Stokes].” J.A. 2529.

An ex-girlfriend provided additional information that Sims and Johnson had to

consider. She told the team that they dated when she was fifteen but broke up after she

became pregnant with Stokes’ child. She said Stokes never supported the child in any way.

She also revealed that it was rumored that Stokes was a bisexual and sexually abused

Martin in the past.

With knowledge of the background of Stokes’ relationship with Martin from these

interviews and of the fact that there was even “some question as to whether or not there

was even a homosexual relationship between [Stokes and Martin],” trial counsel “didn’t

want it to come out that [Stokes] had, you know, used him and had him doing everything

. . . .” J.A. 1537.

Trial counsel’s predictions about guilt proved to be correct. An Orangeburg County,

South Carolina jury found Stokes guilty of murder, kidnapping, first degree criminal sexual

conduct and criminal conspiracy arising from the murder of Connie Snipes. J.A. 990–91.

That set the stage for the penalty phase where the jury would determine whether Stokes

would be sentenced to death or life in prison. J.A. 1013. South Carolina sought to prove

40
several statutory aggravating circumstances in seeking the death penalty. With Stokes’ life

on the line, Sims and Johnson had to decide whether any statutory mitigating circumstances

or other evidence might help save his life. J.A. 997–98.

Complicating their decision-making, about five days before trial, Stokes developed

cold feet about introducing information about his AIDS prognosis to the jury. Even so, he

decided that theme could only be pursued if certain people, particularly family and friends,

were removed from the courtroom. J.A. 2544. And that was the plan—to clear the

courtroom prior to offering the evidence. J.A. 1548–49, 3480.

Then Stokes changed his mind. As the sentencing phase began, “[j]ust as [they] got

ready to start presenting that evidence, [Stokes] then said, no, I don’t want it coming in,”

forcing the court into a recess as the defense team tried to persuade Stokes to allow the

prepared defense to go forward. J.A. 1546. At the eleventh hour, Stokes decided he did not

want his children and family, and the jury for that matter, to hear he had AIDS. While that

was his choice, it substantially gutted his mitigation defense.

Based on this, Sims and Johnson had to adapt and decide what to do instead.

Presenting mitigating evidence about Stokes’ traumatic and abuse-ridden childhood was

an option. The benefit of that mitigating evidence was that it might give jurors some

understanding of how and why Stokes came to the point of committing the horrific events

the jury had just learned about. If jurors understood how such an upbringing could cause

real psychological harm and lead to a propensity toward violence, they might be willing to

spare Stokes’ life. Sims and Johnson had gathered this evidence. The witnesses were

available.

41
 But they also knew that many of the witnesses who would testify about Stokes’

childhood and background in a manner that provided some explanation or context for

Stokes’ criminal conduct—like Stokes’ sister and aunt—also knew about Stokes’

temperament and his relationship with Martin. They knew that along with the potential

mitigating evidence, there was damaging information that would likely come out on cross-

examination, information that would be harmful to Stokes’ defense.

And they also felt the mitigating evidence might not, in this context, be helpful at

all. They felt the jury might perceive the evidence not as mitigating evidence, but as an

attempt to avoid responsibility for the crime. So, after considering the issue, they decided

not to introduce it. Sims described how they came to that decision:

Q. So did -- ultimately, did Sara Stokes, Ruth Davis, or Dr. Rodgers testify
on Mr. Stokes’ behalf at sentencing?
A. No, they didn’t.
Q. And why was that?
A. We made a strategic decision -- after having the opportunity to get
together, we made a strategic decision that certain -- that mitigation kind of
evidence that was the ongoing way that things were done at that time was not
going to work in Orangeburg County.
Having had the opportunity to look at the jury, having had the
opportunity to see how they reacted to a number of things that were going on
in the courtroom, we made a decision that we were going to take another
avenue in order to try to save his life.

J.A. 3469. He continued:

Q. Okay. At the -- what made you decide that the jury would not be receptive
to that testimony?
A. In trial work, and having been in trial work for a period of time up to that
time -- this is going on my 40th year of being a trial lawyer -- you get the
opportunity to look at the jury, you see their reaction to what is happening in
the courtroom, you look at those who are leaning forward at certain times to
certain testimony, you look at those who are leaning back and closing their
arms at certain testimony, you try to look to see if anybody’s shaking their

42
 head with where you want to go, and you try to -- and during the trial you
look at things that are developing.
I always say that a trial has a life of its own. You may start out with a
theory and a process that you want to go through, but in the middle of the
trial, as it begins to progress, you may have to change the way that you are
actually going to go, and that’s what happened in this case.
And take into consideration also you’re in Orangeburg County and
there were things that, back in ‘97, ‘98, ‘99 when this was going on, that we
have to take into consideration too. When you -- the way I looked at it, and I
believe Mr. Johnson will verify this also, when we looked at it and we talked
about the kind of crime that had been committed, we talked about some of
the things that had happened to the young lady who was killed, how her body
was mutilated, and those kind of things in Orangeburg County, we have to
take that into consideration too.
Q. Can you help me understand what about the jury made you think that the
type of evidence that you had intended to introduce originally through Sara
Stokes, Ruth Davis, Dr. Rodgers, would not be persuasive to them?
A. I would have looked at that jury, I would have looked at the composition
of the jury, probably during the trial would have gone home and reviewed
the jurors’ background information again to determine the best way that you
could probably get that juror on your side. I would have -- would have --
there were African-Americans and there were white people on the jury too.
Looking at that, taking it into consideration as a whole and how the
trial had been going and what was being brought out, the question at that
point is whether or not putting out the background of the individual and the
kind of life that they had had as a child would be effective in light of the facts
of the case and the people that you had on the jury.

J.A. 3470–3472.

Johnson also testified about their decision:

Q. So in terms of deciding how to or whether to investigate [Stokes’]
childhood and background, did the aggravated nature of the case affect how
you approached that?
A. It didn’t affect how I approached it, but it sure enough affected how we
presented it.
Q. In what way?
A. When you had to present it -- in your presentation at trial, you had to try
to find a way to present it, if you had to present it, in the preparation. If you
had to present it, you had to try to find a way to present it where it didn’t
seem so offensive, yet you can’t play it down. It’s just a fine balance, because
how do you -- how do you tell somebody, how do you go to a jury and say,

43
 look, we want you to look at the fact that he had a poor upbringing,
particularly African-American, which a lot of us had struggles coming up,
how do you say, well, just because he had a poor upbringing, you need to
overlook the fact that he raped this woman, you need to overlook the fact that
he cut her vagina out, you need to overlook the fact that he cut her nipples
off, you need to overlook the fact that he killed somebody else.
And my job was to defend [Stokes]. So what I did was I looked at
every aspect of the case. If I was trying that case now, it would be a heck of
a lot different because the tolerance we have now. But I would change
nothing because they just wasn’t tolerant of that.

J.A. 3524– 3525.

This testimony reveals that Sims and Johnson, in the exercise of their reasonable

professional judgment, undertook a contemplative thought process about the pros and cons

of using mitigation evidence of Stokes’ childhood and upbringing. After doing so, they felt

the circumstances of Stokes’ crime were just too horrific for such evidence to be helpful.

They decided not to introduce it.

They did, however, offer a different kind of mitigating evidence. They presented

James Aiken, a prison adaptability expert, to testify about Stokes’ ability to adapt to prison.

Aiken’s testimony was intended to dovetail with evidence concerning Stokes’ AIDS

diagnosis. Of course, Stokes’ refusal to allow the AIDS evidence to be used made this

strategy more challenging. But even without the AIDS evidence, Aiken emphasized that

Stokes could be managed in a maximum-security environment for the rest of his life. Trial

counsel used this testimony to argue that Stokes’ life should be spared.

The jury was not convinced. They deliberated for 3 hours and 15 minutes before

returning a death sentence finding four of the six aggravating factors alleged by the State

to make Stokes eligible for the death sentence and recommended the death penalty.

44
 II.

Now, over two decades later, the majority grades trial counsel’s strategic decisions

about mitigation evidence as ineffective. In my view, that conclusion ignores the reality of

trial work and conflicts with Supreme Court precedent.

Importantly, Stokes did not press the mitigation strategy in his state PCR efforts.

Since he did not, he must satisfy Martinez v. Ryan, 566 U.S. 1 (2012), which “provides a

narrow exception to the general rule . . . that errors committed by state habeas counsel do

not provide cause to excuse a procedural default.” Gray v. Zook, 806 F.3d 783, 788 (4th

Cir. 2015). Martinez permits a petitioner to excuse certain procedurally defaulted

ineffective-assistance-of-trial counsel claims. Id. at 789. But the petitioner must establish

cause to excuse the procedural bar before the federal court will consider the merits of that

defaulted claim. This standard reflects the “well-established principle that [f]ederal habeas

courts reviewing the constitutionality of a state prisoner’s conviction and sentence are

guided by rules designed to ensure that state-court judgments are accorded the finality and

respect necessary to preserve the integrity of legal proceedings within our system of

federalism.” Owens v. Stirling, 967 F.3d 396, 422 (4th Cir. 2020) (alteration in original)

(internal quotation marks omitted).

Thus, Stokes must show that the underlying ineffective assistance of trial counsel

claim is substantial, and that PCR counsel was deficient in not pursing that claim. Id. at

423. And if Stokes crosses those two hurdles, he must then show trial counsel was deficient

and that the deficiency prejudiced his defense. While ordinarily I would address these

issues in that order, here I will address them chronologically because this really is

45
somewhat of a chicken or the egg dilemma. PCR counsel could not have been ineffective

unless trial counsel was as well. In my view, neither was defective.

A.

I begin with a discussion of Stokes’ trial counsel because I do not believe that Stokes

has shown a substantial underlying ineffective of assistance of counsel claim. Where a state

prisoner claims ineffective assistance of counsel as the basis of habeas relief, the Court

must also review the claim through the highly deferential lens of Strickland v. Washington,

466 U.S. 668 (1984). First, a petitioner must show counsel’s performance was deficient

and fell below an objective standard of reasonableness. Id. at 687–88. Second, the

petitioner must show the deficient performance prejudiced the defense, meaning “there is

a reasonable probability that, but for counsel’s unprofessional errors, the result of the

proceeding would have been different.” Id. at 694. But in considering these two factors,

the bar is higher for strategic decisions. Much higher. Counsel is “strongly presumed to

have rendered adequate assistance and made all significant decisions in the exercise of

reasonable professional judgment.” Id. at 690. “[S]trategic choices made after thorough

investigation of law and facts relevant to plausible options are virtually unchallengeable.”

Id.

Stokes’ complaints involve classic strategic decisions that the Supreme Court has

determined to be “virtually unchallengeable.” Sims and Johnson—seasoned trial lawyers

with experience trying capital cases in Orangeburg County—felt the personal mitigation

evidence cut both ways. Despite the potential benefits of that evidence to humanize Stokes,

there were risks. Their strategy throughout the trial was to admit guilt, portray the other

46
participants in the murder as more culpable than Stokes and ask for mercy. And they felt

that jurors might view mitigating evidence as to Stokes’ background as a poor excuse for

him committing a gruesome murder. They questioned whether those jurors, from a poor

county in South Carolina, many of whom may have had tough upbringings and life

experiences, might take offense at the mitigating evidence. In other words, “the question

at that point [was] whether or not putting out the background of the individual and the kind

of life that they had had as a child would be effective in light of the facts of the case and

the people that you had on the jury.” J.A. 3472.

In weighing the pros and cons, Sims and Johnson evaluated how the trial was going.

They had been with the jurors throughout the trial. They saw how the jury reacted to the

opening statements and closing arguments and to the evidence presented. They also drew

upon their knowledge of the community where the jurors lived. Recall that Sims and

Johnson lived in that community too.

All of these things went into making the decision, that at that time, in that venue,

with that jury, against the evidence that had been presented so far in that trial, they should

not introduce certain mitigating evidence. Sims and Johnson made the choice—the

excruciating choice—that the benefits of the mitigation evidence, in this situation, were

not worth the risks.

To be sure, one could have a different view. Like the majority, one could conclude

that in the face of the inevitably gruesome evidence about what Stokes did, the best chance

to save his life was to attempt to humanize that conduct through the personal mitigation

evidence. Maj. Op. at 28.

47
 But trial counsel considered that approach. Their best judgment, sitting in counsel’s

chair with an appreciation of the dynamics at the moment, was that the mitigating evidence

would do more harm than good. We are in no position to label that decision unreasonable.

In fact, the Supreme Court has provided guidance in the context of similar arguments.

Substituting Stokes for the petitioner in Wong v. Belmontes, 558 U.S. 15, 25 (2009), Stokes

would argue for a “more-evidence-is-better” approach; “after all, what is there to lose?”

“But here there was a lot to lose. A heavyhanded case to portray [Stokes] in a positive light,

with or without experts, would have invited the strongest possible evidence in rebuttal—

the evidence that [Stokes] was responsible for not one but two murders.” Id. The Supreme

Court tells us that we should not second guess those decisions.

Despite that, the majority engages in just that sort of second guessing, chastising

Sims and Johnson about how they should have tried the case. According to the majority,

the concern counsel identified “reflects a misunderstanding of the duty to mitigate. Trial

counsel were not obliged to ask the jury to excuse Stokes’s actions. Instead, their duty was

to mitigate Stokes’s ‘moral culpability.’” Maj. Op. at 27. Remarkably, the majority goes

on: “[i]f trial counsel believed the jury would not have responded well to a presentation

that minimized Stokes’s conduct, their duty was to find a way to convey this highly

significant evidence without doing so.” Maj. Op. at 29.

I agree with the majority that the purpose of mitigating evidence is not to excuse

conduct, but to help place the moral judgment about the conduct in a context more

favorable to the defendant. And in sociology classes, law review articles and even appellate

court chambers, that distinction may seem clear. But in the fast-paced and intense context

48
of a trial, particularly one in which a defendant’s life is in the hands of twelve jurors from

the community, the line is blurry. Even if presented in the best way by the most capable of

lawyers, it seems far from unreasonable for Sims and Johnson to be concerned that the jury

would not accept that distinction. In my view, this is the exact type of strategic judgment

to which we must defer. Indeed, Strickland tells us that. It counsels us to make every effort

to “eliminate the distorting effects of hindsight” and to “indulge a strong presumption that

counsel’s conduct falls within the wide range of reasonable professional assistance . . . .”

Strickland, 466 U.S. at 689.

Likely recognizing the difficult hurdle of claiming ineffective assistance based on

strategic decisions, Stokes tries to masquerade his criticism of those strategic decisions as

criticism about preparation. He is right that without a reasonable investigation, counsel

does not get the benefit of the strong deference we afford to strategic decisions. Seizing on

that law, Stokes argues that trial counsel did not sufficiently investigate mitigation evidence

and did not do so soon enough. Although the majority largely agrees with Stokes, in my

view this argument also falls short. “The question is whether an attorney’s representation

amounted to incompetence under prevailing professional norms, not whether it deviated

from best practices or most common custom.” Harrington v. Richter, 562 U.S. 86, 105

(2011) (internal quotation marks omitted). Capital sentencing counsel has an “obligation

to conduct a thorough investigation of the defendant’s background . . . to discover all

reasonably available mitigating evidence,” but the investigation need only be “reasonably

thorough.” Owens, 967 F.3d at 413 (internal quotation marks and citations omitted). We

are to employ a highly deferential view of trial counsel’s performance, recognizing the

49
dangers of second-guessing counsel’s assistance after an adverse sentence and

acknowledging there are “countless ways to provide effective assistance in any given case.”

Strickland, 466 U.S. at 689.

Here, as noted above, the record demonstrates the significant work that trial counsel

and their team did in interviewing witnesses and working with experts to develop a trial

strategy. As described above, trial counsel’s investigative efforts involved interviewing a

significant number of Stokes’ family and friends to learn about his upbringing. Through

those efforts, trial counsel learned extensive and specific information about the traumatic

and abusive upbringing Stokes endured. While I recognize not every detail of mitigation

evidence was discovered, that, of course, is not required. What is clear is that trial counsel’s

investigation went well beyond broad outlines of Stokes’ childhood troubles.

This simply is not a situation where trial counsel overlooked “red flags” about

Stokes in investigating their case. Rompilla v. Beard, 545 U.S. 374, 392 (2005). It is also

not a case where trial counsel “did not even take the first step of interviewing witnesses or

requesting records.” Porter v. McCollum, 558 U.S. 30, 39 (2009). Indeed, trial counsel’s

investigative efforts here were more than sufficient, particularly when compared to the

attorneys’ work in Bobby v. Van Hook, 558 U.S. 4, 9–10 (2009) where the record there

showed that counsel “looked into enlisting a mitigation specialist when the trial was still

five weeks away” and were in touch with expert witnesses “more than a month before

trial.” The Supreme Court there found that trial counsel’s performance was not

constitutionally deficient in terms of the timing and scope of the investigation, and that

50
“even if . . . counsel performed deficiently by failing to dig deeper, [the defendant] suffered

no prejudice as a result.” Id. at 12.

Interestingly enough, the majority’s analysis confirms that trial counsel’s

investigation went “beyond some investigations that have been deemed unreasonable.”

Maj. Op. at 26 n.9. It also rightly notes that counsel responded to that investigation by

identifying witnesses who could have presented the personal mitigating evidence available.

“[M]embers of Stokes’s family and a social worker were prepared to testify. A neurologist

and psychiatrist were prepared to testify about Stokes’s HIV status, but presumably could

have offered other personal testimony about Stokes instead.” Maj. Op. at 27. The absence

of mitigating evidence about Stokes’ upbringing and childhood was not a matter of

preparation, or lack thereof. It was the result of strategic decision-making by trial counsel

in the thick of an intense trial.

In hindsight, one could argue that counsel could have done more. Hindsight, after

all, is always twenty-twenty. But that is simply not the standard we apply here. “Strickland

does not require counsel to investigate every conceivable line of mitigating evidence no

matter how unlikely the effort would be to assist the defendant at sentencing.” Wiggins v.

Smith, 539 U.S. 510, 533 (2003).

Trial counsel conducted the “reasonably thorough” investigation required of capital

sentencing counsel. Owens, 967 F.3d at 413, 417 (finding no deficiency in counsel’s

mitigation team’s efforts and further rejecting the complaint that counsel failed to present

mitigating evidence within their possession, noting that counsel “judged with reasonable

competence in avoiding such ‘double-edged’ evidence” (quoting Gray, 529 F.3d at 239));

51
see also Strickland, 466 U.S. at 691 (1984). With a thorough investigation of law and facts

completed, we must credit counsel’s exercise of reasonable professional judgment.

Strickland, 466 U.S. at 690. One thing professional judgment teaches is that the evidence

can cut both ways. Here, trial counsel thought the burdens outweighed the possible

benefits. Stokes has not presented sufficient evidence to overcome the presumption of

adequate assistance for trial counsel’s strategic decisions.

B.

I turn now to Stokes’ PCR counsel. Stokes claims that not only was his trial counsel

ineffective, his PCR counsel was also. Like trial counsel, PCR counsel investigated

mitigation evidence. They hired a mitigation investigator to conduct more interviews of

family, friends, teachers and Stokes’ ex-wife. That investigation also unearthed even more

aggravating evidence against Stokes. They had the benefit of trial counsel’s consultation

with a social worker who was a part of Stokes’ trial defense and had met with Stokes as

well. But, Stokes’ PCR counsel, like trial counsel, decided not to pursue the mitigation

claim, focusing instead on an intellectual disability claim and an actual conflict of trial

counsel claim, which they felt were the stronger claims. *

*
Stokes claimed that trial counsel labored under an actual conflict of interest
because the State’s key witness at trial was Stokes’ ex-wife. Trial counsel Sims prosecuted
Stokes in his earlier assault case. Stokes claimed this conflict prejudiced him because Sims
failed to explore several lines of inquiry in cross-examining Stokes’ ex-wife during the
sentencing phase allegedly due to this conflict. I am not convinced that trial counsel Sims
labored under any actual conflict of interest. This issue was adjudicated below and as the
district court recognized, the PCR court credited Sims’ testimony that he knew Stokes’ ex-
wife would testify and his declaration that there was nothing in his earlier prosecution that
would inhibit his defense. Sims testified that he had a theory in mitigation as to how to
(Continued)
52
 Stokes claims in doing so, PCR counsel’s assistance was ineffective. He claims that

the mitigation issue should have been pressed. And as he and the majority note, PCR

counsel now agree. In this collateral proceeding, PCR counsel conceded that they should

have pursued the claim.

The testimony of PCR counsel certainly supports Stokes’ claim. But their testimony

as a whole must be considered from counsel’s perspective at that time and without the

“distorting effects of hindsight.” Strickland, 466 U.S. at 689. And importantly, PCR

counsel admitted that if they thought the ineffective assistance of counsel claim was strong,

they would have presented that claim. In their own words, they “made some sort of

judgment, explicit or implicit” in deciding not to pursue the referenced mitigation claim.

J.A. 3259. That was the judgment at the time.

That testimony suggests that PCR counsel considered presenting mitigating

evidence but decided against it. In other words, like trial counsel, they made a strategic

decision not to include that and other claims that, at the time, they considered weaker.

Instead, they felt the best approach was to focus only on the strongest claims.

Consistent with that conclusion, the record reveals that Stokes’ pro se application

for habeas relief includes the mitigating evidence issue. Then PCR counsel filed an

address the incident involving Stokes’ ex-wife and noted that one of the issues he was
trying to show was Stokes’ remorse. As for his representation as a whole, when asked if he
labored under a conflict Sims stated that “if I thought I couldn’t have represented Mr.
Stokes to the best of my ability I would not have been in the case.” J.A. 1618.

53
amended application removing the mitigation issue. Removing an existing ground provides

additional evidence of a conscious decision.

And let’s not forget that PCR counsel are experienced death penalty lawyers. One

of Stokes’ PCR attorneys is currently a law professor and director of death penalty

litigation at a law school who transitioned to that role after working almost exclusively on

post-conviction and federal habeas cases while in private practice. He was trained in the

development and presentation of mitigating evidence in death penalty cases and had done

this work before. While even the best lawyers can make mistakes, PCR counsel’s

experience is even more evidence that counsel made a strategic decision not to pursue the

mitigating evidence claim.

Finally, I find it significant that PCR counsel acknowledged “falling on [the] sword”

for Stokes. J.A. 3044. In fairness, counsel admitted that in using that term, he meant “I

didn’t do as good a job as I should have.” J.A. 3044. But to me, when considered in the

context of his testimony as a whole, PCR’s admission amounts to acceptance of

responsibility in hindsight for a failed effort; not necessarily an effort that was ineffective

at the time. In other words, PCR counsel, after his efforts proved unsuccessful, stated that

he would have done things differently as he thought about that case several years later.

Accepting that as true, it does not change the fact that PCR counsel, at the time, made a

strategic decision to pursue the claims that they felt were the strongest. That view of the

decision in hindsight is precisely what the Supreme Court has prohibited.

While one might reasonably say PCR counsel took the wrong approach in dropping

the claim, it was hardly unreasonable. In presenting arguments, lawyers often debate

54
whether to pursue all potential arguments hoping one will stick—or to laser in on the best

arguments because of concerns that the weaker ones may dilute the stronger ones.

Reasonable minds can differ on that question. But the Supreme Court tells us that this

approach is not unreasonable. “Even if some of the arguments would unquestionably have

supported the defense, it does not follow that counsel was incompetent for failing to include

them.” Yarborough v. Gentry, 540 U.S. 1, 7 (2003). As is the case with almost any trial

decision, “[f]ocusing on a small number of key points may be more persuasive….” Id.

Thus, “[w]hen counsel focuses on some issues to the exclusion of others, there is a strong

presumption that he did so for tactical reasons rather than through sheer neglect.” Id. at 8;

see also Workman v. Superintendent Albion SCI, 915 F.3d 928, 942 (3d Cir. 2019)

(discussing this presumption in an attorney’s decision to pursue some claims and decline

to pursue others as a tactical choice in relation to Martinez). Guided by Strickland, we must

indulge the strong presumption that counsel’s conduct falls within the wide range of

reasonable professional assistance. Strickland, 466 U.S. at 689. Here, PCR counsel made

a strategic decision to focus on other claims and that was reasonable under Strickland.

C.

But even if Stokes could establish his trial counsel and his PCR counsel were

deficient, he must show prejudice under Strickland. To show prejudice, a petitioner must

demonstrate a reasonable probability that “at least one juror would have struck a difference

balance” and voted against the death penalty after having heard additional available

mitigation evidence. Wiggins, 539 U.S. at 537. “To assess that probability, we consider

‘the totality of the available mitigation evidence—both that adduced at trial, and the

55
evidence adduced in the habeas proceeding’—and ‘reweig[h] it against the evidence in

aggravation.’” Porter, 558 U.S. at 41 (quoting Williams v. Taylor, 529 U.S. 362, 397–98

(2000)).

Considering the totality of the evidence, both aggravating and mitigating, I do not

see where Stokes has “affirmatively prove[n] prejudice.” Strickland, 466 U.S. at 693. Even

crediting mitigating evidence derived from both trial counsel and PCR counsel’s efforts

and interviews, as well as potential testimony about Stokes’ troubled childhood, neglectful

parents, abuse, and trauma, and even if federal habeas counsel’s child development expert

testified, the aggravating evidence is simply overwhelming. It is hard to conjure a more

horrific set of facts. The jury, of course, heard it all. They learned of how Stokes plotted

several months in advance to murder Connie Snipes, a complete stranger, as he sat in a jail

cell for having already committed yet another violent act. They heard testimony about the

gruesome rape and murder of Connie. They learned that the murderers scalped her head,

stabbed and mutilated her nipples and body, and cut her vagina out of her body. They heard

the testimony of a forensic pathologist and then saw the graphic pictures of Connie’s

mutilated and dismembered body. They heard that Stokes committed another horrific

murder mere days later. They also heard about his criminal history and that he assaulted

his ex-wife and served prison time for that conduct. And on top of it all, Stokes wrote a

letter admitting to much of the detail about his role in the murders of Connie and Doug

Ferguson. That letter is laced with profanity, graphic and detailed in nature, but shows little

remorse—“[t]elling their family sorry would only make them hate me more.” J.A.1225.

56
 In light of this overwhelming evidence, I do not see how Stokes satisfies his burden

or how any additional expert or fact testimony about his upbringing and difficult childhood

would outweigh the gruesome and horrific nature of Connie Snipes’ murder. Stokes and

the majority rightly note that the relevant question is whether “there is a reasonable

probability that at least one juror would have struck a different balance.” See Wiggins, 539

U.S. at 537. But that does not mean we compromise our objective analysis or decline to

view the evidence “taken as a whole.” Id. at 538. If we do, we water down the prejudice

requirement to something akin to anything is possible. See Cullen v. Pinholster, 563 U.S.

170, 189 (2011) (noting that a reasonable probability sufficient to undermine confidence

in the outcome requires a “‘substantial’ not just ‘conceivable,’ likelihood of a different

result”). That, however, is exactly what Stokes asks us to do. His argument boils down to

conjecture, speculating that it just takes one juror; so, the mitigating evidence could have

made a difference. Of course, it could have. Anything could have made a difference. That

is not, however, the approach the Supreme Court requires. Stokes must show that there is

a “reasonable probability that at least one juror would have struck a different balance.”

Wiggins, 539 U.S. at 537 (emphasis added). A reasonable probability is one “sufficient to

undermine confidence in the outcome” but in the capital sentencing context, this means

whether the sentencer would have concluded that the “balance of aggravating and

mitigating circumstances did not warrant death.” Owens, 967 F.3d at 412 (internal

quotation marks omitted). Objectively considering the facts here, there is no basis to

conclude that presenting the mitigating evidence would have had any effect on the outcome

of Stokes’ sentence.

57
 III.

In my view, the record does not support the conclusion that the choice Sims and

Johnson made was unreasonable. It does not support the conclusion that PCR counsel was

unreasonable. But the record does support the district court’s conclusion that even if we

are to conclude that their representation was deficient, Stokes faced no prejudice because

of the overwhelming and horrific aggravating evidence before the jury. Therefore, I

respectfully dissent.

58