Russell Armfield v. SONJA NICKLAUS , Warden

In the
United States Court of Appeals
For the Seventh Circuit
____________________
No. 18-3702
R USSELL A RMFIELD,
Petitioner-Appellant,
v.
S ONJA NICKLAUS , Warden,
Respondent-Appellee.
____________________
Appeal from the United States District Court for the
Northern District of Illinois, Eastern Division.
No. 1:17-cv-03331 — Thomas M. Durkin, Judge.
____________________
A RGUED O CTOBER 30, 2020 — DECIDED J ANUARY 11, 2021
____________________
Before M ANION , R OVNER , and S CUDDER , Circuit Judges.
M ANION , Circuit Judge. Russell Armfield, along with Kimo-
thy Randall and Tyrene Nelson, was charged with first-de-
gree murder in Illinois state court for the 2004 shooting death
of Al Copeland in southwest Chicago.
The jury convicted Armfield. He appealed the conviction
on the grounds that a transcript disclosed inadvertently to the
jury violated his constitutional rights under the Sixth

-- 1 of 23 --

2 No. 18-3702
Amendment’s Confrontation Clause. He lost. He then pur-
sued a collateral attack in state court alleging ineffective assis-
tance of counsel. He lost again. He then filed for federal ha-
beas relief via 28 U.S.C. § 2254. The district court denied relief
and Armfield appeals.
Although Armfield’s positions have been well briefed and
argued by appointed counsel, we affirm denial of habeas re-
lief on Armfield’s Confrontation Clause claim because the
state’s strong case against him renders any constitutional er-
ror harmless. We also reject Armfield’s ineffective assistance
claim; he cannot show trial counsel’s shortcomings resulted
in prejudice.
I. Background
Around 6:00 pm on August 17, 2004, Kimothy Randall
opened fire on Al Copeland’s vehicle while Copeland drove
by. Copeland’s car was struck by gunfire, as was a bystander’s
vehicle. No one was injured. Russell Armfield and Tyrene
Nelson were present.
Later that evening, between 8:00 and 9:00 pm, while riding
with Armfield and Nelson in a car driven by Randall’s girl-
friend, Randall spotted Copeland again. Randall told his girl-
friend to drive to his residence, where Armfield and Nelson
armed themselves. They tracked down Copeland as he drove
away from his own girlfriend’s home. As Copeland ap-
proached an intersection, Randall gave the signal: shoot
Copeland. Armfield and Nelson sprang from their car, ran to-
ward Copeland, and fired multiple shots into his vehicle, kill-
ing him.
The state charged Armfield, Randall, and Nelson with
first-degree murder. Armfield and his codefendants

-- 2 of 23 --

No. 18-3702 3
proceeded to trial before two juries—one jury for Armfield
and Randall, the other for Nelson. The two trials, though sep-
arate, occurred simultaneously before the same judge, with
the juries and defendants shuffling in and out depending on
the evidence presented.1
No doubt this arrangement contributed to the mishap at
the center of this habeas petition. During deliberations, the
Armfield/Randall jury requested a transcript of certain wit-
nesses’ testimony. The court, by mistake, tendered a trial tran-
script containing the prosecutor’s opening statements from
Nelson’s case. The Armfield/Randall jury had not heard this
version. Therein, the prosecutor referenced a videotaped
statement from Nelson that purported to implicate all three
defendants in the murder:
And, ladies and gentlemen, you’re also going to
see a statement given to a Cook County assis-
tant state’s attorney that was videotaped of
[Nelson] confessing to shooting Al Copeland
and laying out essentially the same facts that I
just told you. You will see him tell you how he
and his partners murdered Al Copeland in his
own words.
1 Trial courts sometimes employ this practice to increase efficiency.
Simultaneous trials can circumvent the need for duplicate presentation of
overlapping evidence. But their use does not come risk-free. Here, the trial
court’s confusion between two transcripts spawned years of postconvic-
tion litigation in state and federal court. Whatever resources the trial court
hoped to save were cancelled out long ago by the tax on judicial economy.
We take this opportunity to implore trial courts to exercise caution and
diligence when holding simultaneous trials. The mistake at the center of
this case was completely avoidable.

-- 3 of 23 --

4 No. 18-3702
Supp. App’x at 164.
Neither this snippet nor Nelson’s confession were pre-
sented as evidence of Armfield’s involvement. For that, the
state leaned primarily on eyewitness testimony rather than
physical evidence.
Two witnesses placed Armfield, Nelson, and Randall at
the 6:00 pm shooting scene. One of those witnesses actually
saw Randall pull the trigger and believed Armfield acted as a
lookout.
Grand jury testimony and a police statement from Ran-
dall’s sister revealed how the defendants obtained guns just
before they killed Copeland, though she recanted that story at
trial.
Three more witnesses detailed the defendants’ involve-
ment in the fatal 9:00 pm shooting. Copeland’s girlfriend and
a bystander watched Armfield and Nelson shoot Copeland.
The latter positively identified Armfield and Nelson as the
shooters; he knew them from the neighborhood. Randall’s
girlfriend (the driver) told police and the grand jury Randall
instructed Armfield and Nelson to shoot Copeland, and that
when Armfield returned to the car, he admitted to firing his
weapon. Like Randall’s sister, she recanted this account on
the stand.
Finally, the state introduced evidence regarding a subse-
quent shooting in March 2005 involving Nelson, following
which police confiscated one of the firearms used in
Copeland’s murder. Armfield played no part in this shooting.
Neither Armfield nor Randall put on a defense, and none
of the three defendants testified before the Armfield/Randall
jury.

-- 4 of 23 --

No. 18-3702 5
The jury convicted Armfield of first-degree murder.2 He
received a sentence of 33 years’ imprisonment. Armfield ap-
pealed on grounds that disclosing the reference to Nelson’s
confession deprived him of a fair trial, along the lines of Bru-
ton v. United States, 391 U.S. 123 (1968). The state appellate
court acknowledged the error in allowing Armfield’s jury ac-
cess to opening statements from a separate trial. It nonetheless
held this error non-reversible and further determined it to be
harmless beyond a reasonable doubt. The Illinois Supreme
Court denied review.
Armfield next launched a state collateral attack on the con-
viction. The basis: his trial counsel provided ineffective assis-
tance in multiple respects, including by failing to move to ex-
clude testimony about the March 2005 shooting that did not
involve Armfield. The state appellate court rejected his claim
for failure to satisfy prejudice under Strickland v. Washington,
466 U.S. 668 (1984). The Illinois Supreme Court denied re-
view.
Armfield filed for federal habeas relief. The district court
concluded the state appellate court did not unreasonably ap-
ply Supreme Court precedent to Armfield’s Confrontation
Clause claim or the related harmlessness analysis. Nor did the
state court’s prejudice determination unreasonably apply
Strickland. We granted Armfield’s request for a certificate of
appealability on these two issues.
2 The jury was also asked to determine whether Armfield personally
discharged a firearm during the commission of the offense. The jury found
he did not.

-- 5 of 23 --

6 No. 18-3702
II. Discussion
We review the district court’s denial of federal habeas re-
lief de novo, “but our inquiry is an otherwise narrow one.”
Schmidt v. Foster, 911 F.3d 469, 476 (7th Cir. 2018) (en banc).
A federal court may grant habeas relief following an adju-
dication on the merits in state court only if that decision (1)
“was contrary to, or involved an unreasonable application of,
clearly established Federal law, as determined by the Su-
preme Court of the United States,” or (2) “was based on an
unreasonable determination of the facts in light of the evi-
dence presented in the State court proceeding.” 28 U.S.C.
§§ 2254(d)(1)–(2); Schmidt, 911 F.3d at 476–77.
This standard is difficult to meet. “Unreasonable means
more than incorrect.” Winfield v. Dorethy, 956 F.3d 442, 451
(7th Cir. 2020). The inquiry is “whether the decision was un-
reasonably wrong under an objective standard.” Dassey v.
Dittmann, 877 F.3d 297, 302 (7th Cir. 2017) (en banc).
On both Armfield’s claims, the Illinois Appellate Court is-
sued the “last reasoned decision on the merits,” so we afford
its analysis deference so long as that analysis is reasonable.
Gage v. Richardson, 978 F.3d 522, 529 (7th Cir. 2020). Habeas
relief is warranted only if Armfield shows the state court’s de-
terminations were “so lacking in justification that there was
an error well understood and comprehended in existing law
beyond any possibility for fairminded disagreement.” Har-
rington v. Richter, 562 U.S. 86, 103 (2011). Given this demand-
ing standard of review, we cannot award Armfield the relief
he seeks.

-- 6 of 23 --

No. 18-3702 7
A. Confrontation Clause
The Sixth Amendment’s Confrontation Clause guarantees
a criminal defendant the right to be confronted with the wit-
nesses against him, so that he may cross-examine their testi-
mony and allow the jury to weigh their credibility. Douglas v.
Alabama, 380 U.S. 415, 418–19 (1965).
Armfield’s jury received information that Nelson made a
videotaped statement implicating Armfield in the murder. Yet
Nelson did not testify at Armfield’s trial. This information
reached the jury as would an ex parte affidavit or deposition,
thus depriving Armfield of the opportunity to rebut Nelson
through cross-examination. Armfield argues the disclosure
amounted to a violation of his constitutional rights and
should result in habeas relief.
The state appellate court acknowledged the “unquestion-
abl[e] … error by the trial court.” Short App’x at 35. But it held
the error did not create a constitutional violation contem-
plated by Bruton v. United States, 391 U.S. 123 (1968). Armfield
claims this ruling was contrary to and unreasonably applied
Supreme Court precedent, and that it rested on unreasonable
determinations of fact.
i. Constitutional violation?
In Bruton, at a joint trial for armed robbery, an investigator
testified to a codefendant’s confession that implicated peti-
tioner. 391 U.S. at 124. The codefendant did not testify. The
Court held the admission violated petitioner’s constitutional
right to cross-examine his codefendant. Id. at 126. This, de-
spite jury instructions prohibiting the confession’s considera-
tion toward determining petitioner’s guilt. The jury could not

-- 7 of 23 --

8 No. 18-3702
reasonably be expected to ignore the confession, which added
substantial strength to the prosecution’s case.
The Bruton Court also placed great emphasis on the con-
fession being admitted into evidence; it was not merely para-
phrased through attorney statements or argument. Rather,
the jury received its entire substance, as was true in Douglas,
where the prosecutor read a codefendant’s confession into the
record “under the guise of cross-examination to refresh [the
codefendant’s] recollection” after the codefendant refused to
answer questions about the crime. 380 U.S. at 416. That con-
fession inculpated petitioner. Though reading the codefend-
ant’s confession did not technically qualify as testimony, do-
ing so risked the jury equating it with evidence and created
an inference that the codefendant actually made the state-
ment. The inference could not be tested on cross-examination
because the prosecutor was not himself a witness; nor could
the codefendant be cross-examined on a statement “imputed
to but not admitted by him.” Id. at 419. This procedure denied
petitioner his right of confrontation.
The situation in Frazier v. Cupp, 394 U.S. 731 (1969), pre-
sented far less a threat to petitioner’s confrontation rights.3 In
that case, the prosecutor summarized anticipated testimony
from petitioner’s codefendant (who had already pled guilty
to the same offense) during opening statements. The sum-
mary itself “was not emphasized in any particular way,” but
it referenced a confession made by the codefendant. Id. at 733.
That testimony never materialized; the codefendant invoked
his right against self-incrimination when he took the stand.
3 The state appellate court did not consider Frazier, but the district
court did.

-- 8 of 23 --

No. 18-3702 9
Even if it had, the statement “was not a vitally important part
of the prosecution’s case,” and the jury was instructed that
opening statements must not be considered as evidence. Id. at
735.
These facts led the Frazier Court to conclude no constitu-
tional violation had occurred. In so holding, the Court re-
jected the same general argument Armfield makes: the refer-
ence to his codefendant’s confession in opening statements—
albeit in Nelson’s trial, not his own—“placed the substance of
[Nelson’s] statement before the jury in a way that ‘may well
have been the equivalent in the jury’s mind of testimony.’” Id.
at 734 (quoting Douglas, 380 U.S. at 419).
Resolving Armfield’s Confrontation Clause challenge
boils down to determining on which side of the Bruton/Frazier
line his case falls.
Armfield maintains his conviction flies in the face of Bru-
ton as well as analogous Supreme Court precedent addressing
the use of redacted codefendant confessions at joint trials. For
example, Armfield argues while the summary of Nelson’s
confession did not mention Armfield by name, that quasi re-
daction would still permit the jury to consider it against him.
The prosecutor’s summary stated “[Nelson] and his partners
murdered Al Copeland.” Having just sat through three days
of testimony corroborating the state’s theme that Armfield,
Randall, and Nelson acted as a team, a juror at Armfield’s trial
“need only lift his eyes to [Armfield], sitting at counsel table,”
to figure out the identity of Nelson’s “partners.” Gray v. Mar-
yland, 523 U.S. 185, 193 (1998) (holding confession redactions
that obviously refer to defendant fall within Bruton’s protec-
tive rule); but see Richardson v. Marsh, 481 U.S. 200, 211 (1987)
(“[T]he Confrontation Clause is not violated by the admission

-- 9 of 23 --

10 No. 18-3702
of a nontestifying codefendant’s confession with a proper lim-
iting instruction when, as here, the confession is redacted to
eliminate not only the defendant’s name, but any reference to
his or her existence.”).
Armfield further distinguishes his case from Frazier by
claiming his jury received the substance of Nelson’s confes-
sion, i.e., how Nelson and his partners murdered Copeland.
Because his jurors had already heard testimony about the
murder’s details, Armfield argues the disclosure of the fact
that Nelson confessed validated those facts in their minds.
On the flipside, Armfield’s jury was not exposed to Nel-
son’s confession itself. The wayward transcript contained
only opening statements from Nelson’s trial; it did not include
evidence that Armfield’s jurors did not observe.
Moreover, the allusion to Nelson’s confession was generic
and fleeting, occupying only seven lines of transcript text to-
ward the end of the prosecutor’s monologue. Nelson’s confes-
sion (and the fact that he gave one) played no part in the pros-
ecution’s case-in-chief against Armfield. In addition, the trial
judge instructed Armfield’s jury to consider only evidence in
the form of witness testimony, exhibits, and stipulations; the
judge then gave a clear follow-up instruction that opening
statements are not evidence.4 These instructions did not pre-
sent the same concerns outlined by the Court in Bruton. See
Frazier, 394 U.S. at 736 (“Even if it is unreasonable to assume
that a jury can disregard a coconspirator’s statement when
4 Query whether the jury would interpret this second instruction to
prohibit treating opening statements from Nelson’s trial as evidence, or
whether that instruction carried such force at all. We need not answer
these questions given our holding.

-- 10 of 23 --

No. 18-3702 11
introduced against one of two joint defendants, it does not
seem at all remarkable to assume that the jury will ordinarily
be able to limit its consideration to the evidence introduced
during the trial.”).
We need not answer whether the state appellate court un-
reasonably applied Supreme Court precedent or rested its de-
cision on an unreasonable interpretation of the facts when it
held the transcript mix-up caused no reversible constitutional
error. Even were Armfield’s Confrontation Clause rights vio-
lated, any such violation was harmless.
ii. Harmlessness
Federal habeas relief “is appropriate only if the prosecu-
tion cannot demonstrate harmlessness.” Davis v. Ayala, 576
U.S. 257, 267 (2015). And Armfield does not argue the jury’s
receipt of Nelson’s trial transcript constitutes “the rare type of
error” that overrides the harmlessness requirement. See id.
(citing Glebe v. Frost, 574 U.S. 21, 23 (2014)).
Procedural posture determines how we assess harmless
error. Courts reviewing cases on direct appeal may find a con-
stitutional violation harmless only if the error was “harmless
beyond a reasonable doubt.” Ayala, 576 U.S. at 267 (quoting
Chapman v. California, 386 U.S. 18, 24 (1967)). Collateral pro-
ceedings like this one require more from the habeas peti-
tioner. Armfield is “not entitled to habeas relief based on trial
error unless [he] can establish that it resulted in ‘actual preju-
dice.’” Brecht v. Abrahamson, 507 U.S. 619, 637 (1993). In other
words, relief is proper only if the federal court has “grave
doubt about whether a trial error of federal law had ‘substan-
tial and injurious effect or influence in determining the jury’s
verdict.’” O’Neal v. McAninch, 513 U.S. 432, 436 (1995).

-- 11 of 23 --

12 No. 18-3702
We employ Brecht’s “actual prejudice” test even if the state
appellate court reviewed the matter through Chapman’s harm-
less-beyond-a-reasonable-doubt lens. Jones v. Basinger, 635
F.3d 1030, 1052 (7th Cir. 2011); see also Ayala, 576 U.S. at 268–
70 (explaining the Brecht standard subsumes § 2254(d)’s re-
quirements when a federal habeas petitioner challenges the
state court’s Chapman finding). We employ a de novo review of
the entire record, asking “whether a properly instructed jury
would have arrived at the same verdict, absent the error.”
Czech v. Melvin, 904 F.3d 570, 577 (7th Cir. 2018).
Note: harmless-error review is distinct from assessing
whether there was enough evidence at trial to support a ver-
dict. Jensen v. Clements, 800 F.3d 892, 902 (7th Cir. 2015). The
question here is whether the error “had or reasonably may be
taken to have had” a substantial influence on the jury’s deci-
sion. Kotteakos v. United States, 328 U.S. 750, 764 (1946). When
sizing up the state’s case, we look at the case’s overall
strength, not just the evidence in the state’s favor. Jensen, 800
F.3d at 906. For cases involving Confrontation Clause errors,
we examine factors like the importance of the disclosed state-
ments to the prosecution’s case, whether the disclosure was
cumulative, the presence of corroborating evidence, and the
extent of cross-examination permitted. Id. at 904.
It’s obvious Armfield had no chance to cross-examine Nel-
son (or anyone) about Nelson’s statement to police. The short
summary of Nelson’s confession can also be considered
mostly cumulative; the disclosed opening statement indicates
the confession “lay[s] out essentially the same facts” as those
making up the state’s theory against Armfield. The summary
filled no gaps in the state’s evidence.

-- 12 of 23 --

No. 18-3702 13
The bigger question here is what role, if any, the inadvert-
ent disclosure played in the government’s case. Based on the
detailed, consistent testimony from several independent wit-
nesses confirming Armfield’s involvement, we cannot con-
clude the disclosure of Nelson’s confession had or reasonably
may have had a substantial influence on the jury’s decision.
Two separate witnesses placed all three defendants at the
6:00 pm shooting. Willie Williams spotted Nelson pushing
Randall in a wheelchair. Williams knew both men from the
neighborhood. He watched as Nelson handed Randall a pis-
tol. He saw Randall open fire on a gray Chevrolet. Williams
also noticed a third man, later identified as Armfield, standing
in a nearby alley. To Williams, it seemed Armfield was acting
as a lookout.
Yakirah Robinson was driving her own car close to the
gray Chevrolet. She saw Randall in a wheelchair with Nelson
standing behind him. She saw Armfield nearby, too. She
heard close-range gunshots and sped off without seeing who
fired them. When she reached safety, Robinson noticed her
car had been struck by at least one bullet. She made a police
report and watched as Al Copeland spoke with law enforce-
ment. His vehicle, a gray Chevrolet Cavalier, had also been
shot.
Three hours later (around 9:00 pm) and a few blocks from
the first shooting, Calshaun Vinson observed Copeland driv-
ing away from a restaurant. Vinson had known Copeland
since childhood. Vinson also saw all three defendants in a
black car. He could see Randall in the front passenger seat
with Nelson and Armfield in the rear. He could tell Randall’s
girlfriend, Ayeshia Floyd, was driving. Vinson knew the three
defendants.

-- 13 of 23 --

14 No. 18-3702
The vehicle in which Vinson was riding came to a stop.
From there, he could see Copeland’s car approach a nearby
intersection. Vinson watched Armfield and Nelson exit the
black car and run toward the intersection. He saw them shoot
at Copeland’s car.
Copeland had just dropped off Kawana Jenkins and her
three children at Jenkins’s home. As Copeland drove away,
Jenkins saw a young man flag him down. She saw Copeland
open his car door. She saw the young man open fire on
Copeland. She saw Copeland accelerate away, but as he
reached an intersection, another individual emerged and be-
gan shooting at Copeland. Copeland crashed. Jenkins ran to
his car and found him slumped over.
The assailants shot Copeland five times. He died before
reaching the hospital.
Physical evidence collected from the murder scene sup-
ported this two-shooter narrative. Investigators found bullet
fragments and spent cartridge cases of two different calibers,
9mm and .40 caliber, fired from two different guns. The spent
cases were found at different positions; the 9mm cases were
grouped in the middle of the street, and the .40-caliber cases
were grouped several addresses away, on the sidewalk.
In March 2005, Floyd gave a statement to investigators
about Copeland’s murder and testified before a grand jury.
With Floyd on the stand at trial, the state introduced her
grand jury testimony for impeachment purposes. Floyd told
the grand jury that, around 9:00 pm on August 17, 2004, she
was driving a car with Randall in the front passenger seat and
Nelson and Armfield in the rear.

-- 14 of 23 --

No. 18-3702 15
Floyd stated, after Randall spotted Copeland, he called his
sister and asked her to retrieve a hooded sweatshirt and some
urine bags5 from his house. Floyd drove the trio to Randall’s
home. There, Randall’s sister, Sinquis Prosper, brought the re-
quested items to the car. Floyd observed Randall’s sister cra-
dling the sweatshirt with two hands.
Floyd informed the grand jury she then drove the defend-
ants to the vicinity of Jenkins’ home, where Randall watched
Copeland drop off Jenkins and her children. Floyd witnessed
Randall instruct Armfield and Nelson to “take care of busi-
ness.” Floyd understood this to mean Armfield and Nelson
should shoot Copeland.
Armfield and Nelson exited the vehicle and headed to-
ward Copeland. Floyd heard several gunshots and then saw
Armfield and Nelson running back to her car. Armfield had a
gun in his hand and Nelson was holding his side as if carrying
a gun. They got back in Floyd’s car. Armfield, complaining
about Nelson’s hesitancy, exclaimed: “[He] didn’t want to
shoot until I started to shoot.”
Prosper testified at trial, too. She attested to the phone call
and visit from her brother occurring shortly before 9:00 pm
on August 17, 2004. She also confirmed Randall arrived in a
car driven by Floyd, with Armfield and Nelson sitting in the
rear. She gave Randall the requested sweatshirt and urine
bags. Prosper told police and the grand jury the sweatshirt
contained hard, heavy objects in its pockets. Though Prosper
did not look inside the pockets, she believed they contained
guns; she knew Randall kept guns in the house. The state
5 For Randall’s medical condition. He is paralyzed.

-- 15 of 23 --

16 No. 18-3702
introduced Prosper’s police statement and grand jury testi-
mony for impeachment purposes.
Despite these accounts, the prosecution’s case against
Armfield did not go unchallenged. No physical evidence tied
Armfield to either the 6:00 pm or the 9:00 pm shooting. No
suitable latent fingerprints were left on any of the cartridge
cases found at the murder scene. Police did confiscate a 9mm
pistol after responding to a March 2005 shooting involving
Armfield’s codefendant, Nelson (more on that, below). Alt-
hough forensic analysis matched that gun to the 9mm car-
tridge cases recovered from the murder scene, Armfield had
nothing to do with the events surrounding the pistol’s confis-
cation.
The state’s witnesses had their share of credibility issues.
Williams did not identify Armfield at the scene until he
picked Armfield from a police photo lineup in April 2005—a
full eight months after the shooting. Nor did he even recog-
nize Armfield in the courtroom at trial. Williams also admit-
ted to testifying while on heroin.
Vinson spoke to police the night of Copeland’s murder,
but he did not provide his full account until being charged
with a felony firearm offense in March 2005. He received no
promises for cooperating yet his gun charge was dismissed.
Prosper and Floyd each recanted at trial and told the jury
they had been threatened by police to lie. Prosper testified the
police and a state’s attorney instructed her to falsely claim she
felt heavy, hard objects in the sweatshirt, or else face jail. She
claimed the entirety of her grand jury testimony was fabri-
cated other than the fact that she delivered urine bags and a
sweatshirt to her brother on August 17, 2004.

-- 16 of 23 --

No. 18-3702 17
Floyd stated police gave her a bogus series of events to
memorize and regurgitate before the grand jury. She pur-
ported to comply only after investigators threatened her with
a first-degree murder charge and 12 years’ imprisonment on
three unrelated drug counts. Per Floyd’s revised story, she
and the three defendants spent the day Copeland died driv-
ing around smoking marijuana. Floyd confirmed the encoun-
ter with Prosper involving urine bags and a sweatshirt, but
she denied everything else she had told police regarding
Copeland’s murder. She claimed those additional events did
not happen. She also testified she could not remember what
she told the grand jury, adding it was all a lie anyway.
Floyd and Prosper’s claims of coercion conflicted with
their grand jury testimony, in which they stated no promises
or threats had been made in connection with their willingness
to talk with investigators. The government also put on wit-
nesses who denied Prosper and Floyd’s accusations of threats
from law enforcement. The jurors had the opportunity to
weigh those rebuttals against the allegations. They knew of
Prosper and Floyd’s close relationships with Randall and
could infer from them a motive to protect him. They also
heard a portion of Floyd’s grand jury testimony in which she
explained Randall abused her and she was scared he could
have her harmed or killed.
Armfield contends the flaws in the state’s case made the
question of his guilt a razor-thin call. The jury, after all, delib-
erated for nearly fourteen hours. At one point the jurors in-
formed the judge they reached an impasse and had to be in-
structed to keep deliberating. He also points to the jury’s con-
clusion that he did not fire a weapon during the murder as
proof that it rejected evidence to the contrary, such as Vinson

-- 17 of 23 --

18 No. 18-3702
and Floyd’s accounts. Since the jury discredited those key wit-
nesses, there was little left of the prosecution’s case and the
disclosed opening statements from Nelson’s trial must have
tipped the scale against him, he claims.
We disagree. First, that the jury determined Armfield did
not discharge a weapon does not mean the jury rejected
Vinson and Floyd’s testimony entirely. Nor does the special
verdict bear on Armfield’s guilt for first-degree murder be-
cause he still faced accomplice liability. The jury found him
guilty of that crime; he participated directly in bringing about
Copeland’s death. In other words, even if the jury discounted
testimony that Armfield himself fired at Copeland, that does
not elevate the disclosed summary of Nelson’s confession au-
tomatically (or at all); the special verdict does not tell us
whether the disclosure had a “substantial and injurious” in-
fluence on the jury’s ability to find Armfield guilty of murder.
More significant is the weight of evidence against Arm-
field. Multiple unconnected witnesses corroborated the
state’s theory: Armfield acted as part of a three-man crew that
tried to kill Copeland once, failed, tried again only hours later,
and succeeded. Two bystanders placed him, along with Nel-
son and Randall, at the scene of the 6:00 pm attempt on
Copeland’s life. Two more witnesses—one bystander and one
accomplice—confirmed Armfield and Nelson’s roles as trig-
ger men in the 9:00 pm shooting and additionally placed Ran-
dall at the scene. Physical evidence and an additional witness
supported the two-shooter theory. Another witness identified
all three men together, only minutes before the fatal shooting,
when delivering them a sweatshirt containing what she be-
lieved to be guns. For all their personal credibility baggage,
the witnesses’ narratives were consistent in substance and

-- 18 of 23 --

No. 18-3702 19
detail. Vinson’s testimony and Floyd’s grand jury statements
were especially damning.
The trial evidence alone spanned almost 450 pages of tran-
script. Not once did the state discuss or even hint at Nelson’s
confession. Only during deliberations was Armfield’s jury ex-
posed to a summary of Nelson’s confession—a summary last-
ing two sentences, made during opening statements (not evi-
dence) in another trial, that revealed no new details about the
murder. Review of the entire record leaves us with no “grave
doubt” about harmlessness. The disclosure was inconsequen-
tial next to the evidence, and a “properly instructed jury
would have arrived at the same verdict” absent the disclo-
sure. Czech, 904 F.3d at 577. Armfield’s Confrontation Clause
claim warrants no habeas relief.
B. Ineffective Assistance of Counsel
Armfield also maintains he received ineffective assistance
of counsel because his trial attorney did not move to exclude
evidence concerning a shooting that happened in March 2005.
Armfield raised this claim in a collateral postconviction pro-
ceeding. The state appellate court denied him relief. Armfield
now insists the state court unreasonably applied Strickland in
concluding trial counsel’s error did not prejudice his case.
At trial, Tykima Walker testified that on March 18, 2005,
she drove her two children to the Cook County jail to visit one
of the children’s fathers. Three men in another vehicle, a
Grand Prix, followed her. She identified one of them as Nel-
son. At some point along the way, they began shooting at her
car. When Walker arrived at the jail, Nelson and another of
the Grand Prix occupants exited their car and followed her
inside.

-- 19 of 23 --

20 No. 18-3702
Officer Frank Ramaglia responded to reports of a person
with a gun in a white Grand Prix. He located the Grand Prix,
parked. At this point, only one individual remained in the car:
a Calvin Armfield—not our petitioner, Russell Armfield.
Ramaglia testified he placed Calvin Armfield in custody and
retrieved three firearms from the car: a .380-caliber machine
pistol; a .40-caliber pistol; and a 9mm pistol. Forensic ballistics
analysis determined the 9mm pistol had been used in
Copeland’s murder. The jury viewed all three guns while the
prosecutor had Ramaglia verify chain of custody.
Strickland v. Washington provides the clearly established
federal law for Armfield’s ineffective assistance of counsel
claim. 466 U.S. 668 (1984). Pursuant to Strickland, Armfield
must show (1) counsel’s performance “fell below an objective
standard of reasonableness,” and (2) “there is a reasonable
probability that, but for counsel’s unprofessional errors, the
result of the proceeding would have been different.” Id. at 688,
694. Failure to prove either deficient performance or prejudice
defeats a petitioner’s claim. Winfield, 956 F.3d at 452.
For prejudice, a reasonable probability is one “sufficient to
undermine confidence in the outcome.” Strickland, 466 U.S. at
694. The “likelihood of a different result must be substantial,
not just conceivable.” Richter, 562 U.S. at 112.
The state appellate court did not address whether trial
counsel performed deficiently under Strickland’s first prong.
We need not address performance either if resolving the claim
on prejudice will do. McNary v. Lemke, 708 F.3d 905, 914 (7th
Cir. 2013). But the state court indeed assessed prejudice, and
our review is “doubly” deferential at this stage. Richter, 562
U.S. at 105.

-- 20 of 23 --

No. 18-3702 21
Here, even if counsel’s failure to object was sub-standard,
that failure did not prejudice Armfield’s case.
As with Armfield’s briefing of harmlessness, much of his
argument for prejudice rests on his view that the state brought
a weak case against him. See Strickland, 466 U.S. at 696 (“[A]
verdict … weakly supported by the record is more likely to
have been affected by errors than one with overwhelming rec-
ord support.”). Granted, the state’s evidence against Armfield
was not one-sided, but it was both robust and compelling. We
discussed the evidence’s strength (and shortcomings) in the
preceding section and need not repeat ourselves.
In addition, Armfield highlights the Illinois Appellate
Court’s adoption of its earlier assessment of Armfield’s case
on direct appeal. On direct appeal, the state court incorrectly
determined police recovered the 9mm pistol used in
Copeland’s murder from petitioner’s car, not that of Calvin
Armfield. Armfield argues this mistake constitutes an unrea-
sonable determination of fact that should result in habeas re-
lief.
But nothing indicates the jury made that same mistake.
Early on in closing arguments, while referring to Armfield
and Randall, the prosecutor said “they” are so bold as to shoot
up the area around the Cook County jail. That was technically
incorrect; only Nelson took part in the shooting near the jail.
Armfield and Randall had nothing to do with it. The misstate-
ment, however, was minor, and we can find no prejudice re-
sulting from it. Indeed, the prosecutor made sure to include
Calvin Armfield’s first name when later getting into the par-
ticulars of the March 2005 shooting, distinguishing that indi-
vidual from petitioner. At one point, the prosecutor began to
refer to Calvin Armfield as petitioner’s brother. Defense

-- 21 of 23 --

22 No. 18-3702
counsel objected and the judge sustained the objection, ex-
plaining, “There is no testimony about what relation the one
bore to the other.” Supp. App’x at 574. Given this context,
there was little room for the jury to confuse the two. Nor was
the jury likely to associate the .380- and .40-caliber guns re-
covered in March 2005 with Copeland’s murder. Nothing
linked those weapons to the fatal August 2004 shooting in any
way, and the jury’s exposure to them was momentary and
procedural at most.
Relatedly, Armfield implies trial counsel’s failure to object
allowed the state to unfairly lump him in with “superpreda-
tors” brazen enough to shoot up an area with heavy law en-
forcement presence near the jail and courthouse. The prose-
cution didn’t need to reference the March 2005 shooting for
the jury to draw that conclusion. Armfield fit the bill thanks
to his role in two shootings on the same day on the public
streets of the same neighborhood.
The state appellate court did not apply Strickland’s preju-
dice test unreasonably. Considering the strength of the pros-
ecution’s evidence against the secondary value added by the
March 2005 shooting, there exists no substantial likelihood of
a different result here.
III. Conclusion
The Illinois Appellate Court reviewed Armfield’s convic-
tion twice: once on direct appeal (his Confrontation Clause
challenge), and again through collateral proceedings (his in-
effective assistance of counsel challenge). In neither instance
did the state court resort to an unreasonable analysis that
would permit federal habeas relief. The state’s case against
Armfield was strong, with multiple, independent witnesses

-- 22 of 23 --

No. 18-3702 23
swearing to the same events and implicating him as a key
player in Al Copeland’s murder. Thus, Armfield cannot over-
come harmlessness or make a showing of prejudice, as re-
quired for his two claims. The district court’s judgment is
AFFIRMED.

-- 23 of 23 --