United States Court of Appeals
For the Eighth Circuit
___________________________
No. 10-2434
___________________________
Terrick Terrell Nooner
lllllllllllllllllllll Plaintiff - Appellant
v.
Ray Hobbs, Interim Director, Arkansas Department of Correction
lllllllllllllllllllll Defendant - Appellee
____________
Appeal from United States District Court
for the Eastern District of Arkansas - Pine Bluff
____________
Submitted: June 16, 2011
Filed: August 24, 2012
____________
Before BYE and MELLOY, Circuit Judges, and SMITH CAMP1, District Judge.
____________
MELLOY, Circuit Judge.
An Arkansas jury convicted Terrick Terrell Nooner of capital felony murder
committed in the course of a robbery. He received a sentence of death. In state court,
he unsuccessfully appealed his conviction. See Nooner v. Arkansas, 907 S.W.2d 677
1
The Honorable Laurie Smith Camp, United States District Judge for the
District of Nebraska, sitting by designation.
(Ark. 1995) (opinion on direct appeal). He then unsuccessfully sought state and
federal post-conviction relief. Nooner v. Arkansas, 4 S.W.3d 497 (Ark. 1999);
Nooner v. Norris, 402 F.3d 801 (8th Cir. 2005). Finally, he received permission from
our court to file a second petition for federal relief alleging new and previously
unavailable evidence demonstrated that he was actually innocent and that the state
had failed to disclose evidence favorable to the defense.
Nooner supported his second petition primarily with: (1) a recantation and
statement from an important witness for the state, Antonia Kennedy, (2) a confession
from a non-testifying co-defendant, Robert Rockett, and (3) expert testimony from
a video-image expert. Antonia Kennedy and Rockett both assert that Rockett, rather
than Nooner, committed the murder but that Nooner was present in a car outside the
building where the murder took place. The expert, applying techniques developed
after Nooner's trial for analyzing grainy and unclear video images (images that were
admitted into evidence at the trial), opined that Nooner could not have been the
purported murderer shown in the images because the height of the person in the
image more closely matched Rockett's height than Nooner's.
The district court2 granted Nooner two evidentiary hearings as to limited issues,
but ultimately denied relief. The court found that Nooner's evidence was neither new
nor previously unavailable. In the alternative, the court found that Nooner failed to
demonstrate the requisite diligence in presenting his claims. In the further alternative,
the court found that Nooner failed to make an adequate showing of actual innocence
because the recantations were unreliable, the expert's opinion was unconvincing, and
together with the trial evidence, Nooner's showing was insufficient to create the
requisite likelihood that a jury would have found him not guilty. Nooner appeals,
raising several issues. We affirm.
2
The Honorable J. Leon Holmes, United States District Judge for the Eastern
District of Arkansas.
-2-
I. Background
A. The Underlying Crime and the Trial
On March 16, 1993, at about 1:30 a.m., Scot Stobaugh entered a laundromat
in Little Rock. Later that morning, he was found dead on the floor of the laundromat.
He had been shot seven times.
The laundromat had three surveillance cameras. The three cameras recorded
to one videotape with the feeds from the cameras recording only one view at any
given moment. As a result, a video recording from the laundromat on the morning
of March 16 does not contain a continuous depiction of events surrounding the
murder. Rather, it shows bits and pieces of the events that unfolded after Stobaugh's
arrival.
The tape shows a second man entering the laundromat shortly after Stobaugh.
The second man appears to accost Stobaugh and walk Stobaugh towards the back of
the laundromat. The tape shows the two men walking between two rows of machines
and partially out of the camera's view. It then shows the second man possibly kicking
at someone on the floor and out of view with the second man's left arm visible in the
image and extended back but his right arm forward and out of the image. Stobaugh
is out of the image, but based on the sequence of frames, Stobaugh would have been
in front of the second man. The state argued that, given the second man's position,
this frame could depict the moment of a shooting. The state also argued that the
shooter was right handed and held the gun in the right hand not shown in the image.
All images are grainy and unclear, but certain aspects of the second man's
clothing are visible as is the rough relative size of the second man compared to
surrounding objects and compared to Stobaugh (subject to the limitation that both
men are wearing baggy clothes and hats). The camera recording the most useful
-3-
images was mounted above the men in the rear of the laundromat, and analysis of the
men's height and size is made difficult due to the camera's downward-facing
perspective and due to the additional fact that the images appear to show both men
while moving, standing with legs apart, or standing slightly stooped.
About ten days after the shooting, detectives focused their attention on Nooner
based in part on statements Antonia Kennedy and Rockett gave to investigators.
Officers had arrested Rockett for a robbery and murder at a convenience store in
another town (the "Stax murder"). Rockett told police that Nooner was responsible
for the Stax murder and for Stobaugh's murder. Although Rockett did not testify at
Nooner's trial, he told officers that he and Nooner had been together on the night of
Stobaugh's murder. According to Rockett, Rockett was driving a car Nooner had
stolen, and the two men were looking for someone to rob when Nooner saw Stobaugh
enter the laundromat. Rockett claimed Nooner told him to pull into the laundromat
and then exited the car, entered the laundromat, walked Stobaugh to the back of the
building, forced him to the floor, shot him seven times, kicked him, and took his
checkbook, social security card, and $100 cash. Rockett stated that Nooner used a
0.22-caliber Ruger pistol the two men had stolen. Eventually Rockett pleaded guilty
to first degree murder for his participation in Stobaugh's murder and was sentenced
to sixty-five years' imprisonment consecutive to other sentences, including a life
sentence for the Stax murder.4
Police also questioned a group of people associated with Nooner and Rockett.
These same people ultimately testified at Nooner's trial. These witnesses included
three sisters: Antonia Kennedy, Terri Kennedy, and Jazmaar Kennedy. Terri
Kennedy was Nooner's girlfriend at the time of Stobaugh's murder and the mother of
4
Although Rockett initially laid blame for the Stax murder on Nooner,
investigators came to believe this accusation was false and prosecutors dropped
charges against Nooner related to the Stax murder.
-4-
Nooner's child. Other witnesses included Johnny Martin and Isaac Warren, two men
who had been with Nooner shortly before Stobaugh's murder or who had been around
an apartment shared by the Kennedy sisters.
By any account, Antonia Kennedy was an important witness for the state at
Nooner's trial. She testified consistently with a statement she had given to police
during their investigation. She stated that Nooner told her the morning following the
murder that he had killed the man in the laundromat. Her testimony was detailed; she
recounted Nooner describing the victim's body as jerking after being shot and Nooner
claiming to have kicked the body. According to her testimony, Nooner bragged about
the killing. She quoted Nooner as stating that he "had to kill me a mother-f***er,
shoot me a mother-f***er," and as stating, "Sister-in-law, when I shot him, his body
was jerking, and I had to do it with him." Antonia Kennedy also stated that she had
seen Nooner with a Ruger 0.22-caliber pistol before the murder and after the murder.
She stated that, in the car Nooner and Rockett were using, she saw a checkbook
containing a check with Stobaugh's name on it. Finally, she described Nooner's
reaction to seeing the surveillance images in the news and his reaction to a news
report that the crime may have been a random, gang-initiation event. According to
Antonia Kennedy, Nooner asked if the images looked like him and acted "happy
because the police wasn't [sic] on to him."
On cross examination, Antonia Kennedy admitted that she had given an
inconsistent, video-recorded statement to a defense investigator after police initially
interviewed her. Antonia Kennedy had told the defense investigator that her initial
statement to police inculpating Nooner (the statement consistent with her
subsequently provided trial testimony) was a lie she gave to police while in custody
in Jonesboro, Arkansas, for a different matter. She claimed to the defense
investigator that she lied to police because police said they would take care of charges
against her in Jonesboro. At trial, however, she asserted that her trial testimony and
statement to police were truthful and that her statement to the defense investigator
-5-
was false. The jury also learned that, after Antonia Kennedy spoke to police, she did
not subsequently face any charges in Jonesboro. Finally, the jury learned that
Antonia Kennedy had known Nooner from when the two were minors and had been
together in a juvenile detention facility. Antonia admitted that the two had been
sexually involved in the past, Nooner was in a relationship with Antonia's sister Terri
Kennedy at the time of the murder, and Nooner had fathered a child with Terri
Kennedy. Finally, Antonia identified Nooner as the man in the video.
Johnny Martin, Antonia Kennedy's ex-boyfriend, described his familiarity with
Nooner and testified that he recognized Nooner in the surveillance images based on
his face, coat, and hat. Martin claimed that he had seen Nooner wearing the same
clothes the evening before the murder.
Isaac Warren testified that he saw Nooner in an apartment shared by the
Kennedy sisters and that he had talked to Nooner at the apartment about purchasing
a 0.22-caliber Ruger pistol. Warren stated that he saw Nooner with the gun before
the murder but not after. Warren also testified that, after Nooner saw images from the
crime scene, Nooner stated "they can't prove it" and "they had to shoot seven times."
Jazmaar Kennedy testified that she could identify Nooner in the surveillance
video based on the coat and hat he was wearing because they belonged to her sister,
Terri. Jazmaar had seen Nooner wearing those items the day before the murder.
An expert witness from the Arkansas Crime Lab testified that markings on five
of the seven bullets and all seven shell casings were consistent with the bullets having
been fired by a 0.22-caliber Ruger pistol with identifiable rifling patterns. The expert
stated that the remaining two bullets were too damaged for analysis.
-6-
In addition, a police investigator testified that surveillance images were
published in a local paper, thirteen callers contacted police to identify the assailant,
and all thirteen callers identified persons other than Nooner.
Finally, Nooner called his stepfather and Terri Kennedy as defense witnesses.
Both testified that Nooner was at his mother's and stepfather's house when Stobaugh
was murdered. Nooner's theory of the case at trial, then, was not that he was involved
as the driver and that Rockett shot Stobaugh. Rather, Nooner's theory as presented
to the jury was that Nooner was at his mother's home in bed when Stobaugh was shot.
Nooner's stepfather testified that Nooner was present at the home the night of the
murder without Terri Kennedy and without Nooner's child. Terri Kennedy, however,
testified that the child was at the stepfather's home with Nooner.
B. Successive Habeas Proceedings
In the present petition, Nooner raises his claim of actual innocence for two
purposes. First, he asserts actual innocence as a gateway issue to permit him to assert
a Brady5 violation alleging official concealment of evidence favorable to the defense.
Second, he asserts actual innocence as a stand-alone claim citing the theory that the
Supreme Court described, but did not rely upon, in Herrera v. Collins, 506 U.S. 390,
400 (1993).
In his direct appeal and initial state and federal collateral proceedings, Nooner
did not raise any of the issues he raises in the present habeas proceedings. Further,
he did not allege at any time that he was present at the laundromat and in the car.
Rather, in his initial appeal and prior collateral proceedings, he presented various
arguments and attacks upon the state's proof and the procedural integrity of the
underlying proceedings. At no time prior to filing the present petition did he abandon
5
Brady v. Maryland, 373 U.S. 83 (1963).
-7-
his version of the facts or his affirmative presentation of evidence at trial in which he
asserted that he was wholly uninvolved in events at the laundromat and at home in
bed at the time of Stobaugh's murder.
Nooner sought and received permission from our court to file the present
petition, alleging that Antonia Kennedy recanted her testimony and placed Rockett
and Nooner at the scene of the murder with Rockett as the shooter. He also alleged
Rockett recanted, confessed to being the shooter, and placed Nooner at the scene.3
Finally, Nooner alleged that a computer vision expert, Dr. Hartley, could accurately
measure the height of the second man in the surveillance video based upon analytical
techniques using the purportedly known height of the surveillance camera at the time
of the shooting and also using input gleaned from a single image.
The state identified a competing expert, Dr. Russ, to address Dr. Hartley's
methodology and conclusions. Ten days before a scheduled evidentiary hearing,
however, Nooner deposed Dr. Russ and discovered substantial inconsistencies in his
reported academic credentials. Based on these revelations, Nooner filed a motion
seeking permission to subpoena Dr. Russ's academic and employment records. The
district court denied the motion, stating that the evidence sought was merely
impeachment evidence and would require the authorization of discovery with
3
The Arkansas homicide statute defining felony murder provides "It is an
affirmative defense to any prosecution under subdivision (a)(1) of this section for an
offense in which the defendant was not the only participant that the defendant did not
commit the homicidal act or in any way solicit, command, induce, procure, counsel,
or aid in the homicidal act's commission." Ark. Code Ann. § 5-10-101(b). Although
the state makes brief reference to Nooner's guilt even if he were merely the driver,
neither the parties nor the district court fully developed these arguments below. We
mention this argument and statute merely to explain the posture of this case and to
explain why we do not address the obvious issue of accomplice liability as would
typically preclude a showing of actual innocence under traditional standards for
felony murder.
-8-
subpoenas to issue from three different district courts in other states with return dates
of less than one week. Nooner later filed a motion pursuant to Daubert4 seeking to
exclude Dr. Russ's testimony. The district court denied the motion.
During the first evidentiary hearing, Nooner presented testimony primarily
from Antonia Kennedy and Rockett. Nooner does not allege any procedural
irregularities regarding the first hearing. Rockett testified that he (Rockett) was the
second man in the video who shot Stobaugh and that Nooner was present, but
remained in the car. In addition, Rockett provided a handwritten declaration dated
August 8, 2007, as well as a typed declaration signed the same day. In the
declarations, Rockett claimed to have fabricated his original statements to police in
which he placed the blame on Nooner for Stobaugh's murder. According to Rockett,
he misdirected police as a means to put pressure on Nooner so that Nooner would not
inculpate Rockett in other crimes. Counsel for the state, however, exposed several
inconsistencies in Rockett's testimony and recantations when questioning Rockett at
the hearing. As accurately described by the district court:
In the typewritten declaration, Rockett said that he never saw a wallet
or a checkbook that belonged to Scot Stobaugh and that he had no idea
where the story came from that those items were taken from the robbery
and shooting at the . . . laundromat. He said that he never had them in
the car that he was driving where someone else could have seen them.
In the handwritten declaration, however, he said that he is the one who
gave Nooner the checkbook to show to Antonia Kennedy. At the
evidentiary hearing, he testified that the typewritten declaration was
wrong on that point. In his handwritten declaration, Rockett said that he
was lying in the floor of the Jacksonville [Arkansas] jail on March 30,
1993, when he heard Nooner talking to police and apparently placing the
blame on him with regard to a robbery in Jacksonville [the robbery
involving the Stax murder]. Rockett testified that it was at that time that
he decided he needed to place the blame for the murder of Scot
4
Daubert v. Merrell Dow Pharm., Inc., 509 U.S. 579 (1993).
-9-
Stobaugh on Nooner so as to keep Nooner from "snitching" on him.
Apart from the fact that it makes no sense for Rockett to "snitch" on
Nooner regarding the . . . laundromat [murder] in order to keep Nooner
from "snitching" on him regarding the Stax robbery [and murder],
Rockett's first statement regarding Nooner's murder of Scot Stobaugh
was given on March 27, 1993, three days before Rockett says he first
conceived of the idea of blaming Nooner for the Scot Stobaugh murder.
Rockett's declarations are inconsistent and unpersuasive.
In addition, Rockett denied having told Antonia Kennedy near the time of the
murder that he—rather than Nooner—killed Stobaugh. Further, Rockett admitted he
was left handed. Finally, even though Rockett now claims to have been the killer,
Rockett still maintains that Nooner bragged about killing Stobaugh. In this regard,
Rockett described Nooner as big talker who wanted to seem like a violent criminal,
but who did not actually shoot Stobaugh.
Antonia Kennedy testified that her original statement to police and her trial
testimony were false and that she had been telling the truth when speaking with the
defense investigator prior to Nooner's trial. She made the same assertion in a
declaration dated August 21, 2007, claiming to have testified falsely at the trial. In
addition, she claimed that Rockett had convinced her to lay blame for Stobaugh's
murder on Nooner. She, like Rockett, however, continued to maintain that Nooner
bragged about the killing. Contrary to Rockett's testimony, Antonia Kennedy
testified that Rockett admitted to her the morning after the murder that he (Rockett)
had killed Stobaugh. Also contrary to Rockett's testimony, Antonia Kennedy stated
that she had been sexually involved with Rockett.
Nooner's claim of prosecutorial misconduct alleged prosecutors knowingly
allowed Antonia Kennedy to testify falsely and failed to disclose the fact that police
improperly coerced her into making her initial statement inculpating Nooner.
Antonia Kennedy responded to questioning by the court at the evidentiary hearing,
however, and stated that no police officers asked her to make false statements or lie
-10-
regarding the identity of the killer. According to Antonia Kennedy, officers who
questioned her did not know whether Rockett or Nooner had been the shooter and did
not direct her to place the blame on either particular suspect.
During the second evidentiary hearing, the court received testimony from Drs.
Hartley and Russ. Over repeated objections from Nooner's counsel, the district court
extended the hearing until approximately 10:00 p.m. Both experts had submitted
reports, with Dr. Hartley's report setting forth his methodology and conclusions, and
Dr. Russ's report challenging Dr. Hartley. Nooner cross examined Dr. Russ,
challenging his credentials and exposing the inconsistencies that had led to Nooner's
motion for discovery and the motion to exclude Dr. Russ's testimony entirely.
In his report and in his testimony, Dr. Hartley described his general
methodology for image analysis and discussed specifically his assumptions and
corrections made in Nooner's case. In general terms, Dr. Hartley's method of image
analysis uses parallel lines within the image to project and calculate an off-image
convergence point. The effect relied upon is similar to the apparent convergence of
parallel rails of a railroad track at a distant horizon. Using distances determined
through the projection of this convergence point and using a purportedly known
height within an image (or using the purportedly known height of the camera), Dr.
Hartley draws conclusions based upon the relative heights of other objects in the
image.
We need not discuss in detail the geometric relationships, proofs, and
calculations involved, nor the theoretical viability of Dr. Hartley's techniques in
general. Dr. Russ and the state did not challenge the general theoretical
underpinnings of Dr. Hartley's method. In other words, given a hypothetical case
involving a clear image of a stationary object, the ability to clearly locate the top and
bottom of an object in an image, undisputed inputs, and undisputed assumptions, the
state and Nooner agree that Dr. Hartley's theoretical approach could be suitable for
-11-
some applications and some levels of proof in idealized situations. Accordingly, we
focus our factual recitation upon the areas of disagreement in order to assess the
reliability of Dr. Hartley's opinion in this case.
Ultimately, Dr. Hartley testified that he believed the height of the second man
in the image to be 66.1 inches, with a possible error of +/- 1.1 inches. Dr. Hartley
testified that he started with a poor quality non-digital tape of the laundromat. He
then digitized the six most potentially useful images and used a combination of
manual and automated steps to locate and approximate straight lines within the image.
He also used a manual method to locate and approximate a point on the floor directly
beneath the man's head representing a point between where the man's two feet
touched the floor. He also testified that, after calculating a height using his geometric
projection methods, he came up with a rough measurement of 67.6 inches +/- 0.5
inches (allowing for potential error in extracting measurements from the image and
due to the fact that a person stands taller when standing perfectly erect than when
walking mid-stride with legs angled away from vertical as the second man in the
image appeared to be positioned). Regarding this 67.6 inches +/- 0.5 inch figure,
which incorporated an estimated adjustment for stride (the range of 67.1–68.1 inches
before consideration of other adjustments), Dr. Hartley stated that he was 75–80%
confident in his result.
He then arrived at his overall adjusted measurement of height by making
adjustments based on the fact that the man in the image was wearing shoes and
wearing a hat. Regarding shoes, Dr. Hartley stated that he could not tell what type
of shoes the man in the image was wearing, but that he assumed shoes would add
between 0.90 and 1.25 inches depending on type, and applied an adjustment of 1.1
inches for shoes. Dr. Hartley's report stated that he arrived at this figure via simple
experiments. In his testimony, he admitted that his source for these estimates was a
simple examination of his own height when wearing two different pairs of his own
shoes. Regarding an adjustment for the hat, Dr. Hartley described no source for his
-12-
estimate. Regarding the adjustment for stride as embodied in the initial figure, Dr.
Hartley described walking along a wall "with a thing attached to [his] head" to
determine an average height reduction caused by a person's legs being in stride rather
than together and vertical. He also described references to this height-impacting
effect of walking (in contrast with standing erect with legs straight) in the literature.
Dr. Hartley then added these adjustments together, subtracted them from the height
determined through his geometric projection method, and concluded, with "99.1 %"
confidence, that Nooner could not be the shooter, stating that the actual height of the
man in the image was 66.1 inches +/- 1.1 inches (65–67.2 inches).
Finally, Dr. Hartley stated that, prior to his analysis, Nooner's counsel had not
provided him with the actual heights of the relevant people in the case. The record
shows that an arrest summary of the Little Rock Police Department lists Nooner's
height as 68 inches, and an admission to the Arkansas Department of Corrections lists
Nooner's height as 69 inches and his weight as 164 pounds. Rockett's arrest report
dated March 25, 1993 lists his height as 66 inches and his weight as 115 pounds.
Stobaugh's autopsy report stated that he was 70 inches tall and weighed 171 pounds.
Further, in the images, Stobaugh, like the second man, was wearing shoes and a hat.
The district court noted its impression that Stobaugh and the second man appeared
to be approximately the same size.
Dr. Russ criticized Dr. Hartley's methods on several grounds related to the
manual vs. automated treatment of the images and related to the several adjustments
described above. In addition, Dr. Russ criticized Dr. Hartley's overall failure to
identify another object in the image that could be physically measured (such as the
victim or one of the various pieces of laundry equipment) and failure to apply the
same analysis to such an object. According to Dr. Russ, such a comparison would
help to show how a height derived through Dr. Hartley's methodology as applied to
the actual images in this case related to "ground truth."
-13-
Finally, Dr. Russ indicated that it would have been possible at the time of the
murder (when it would be known that the camera had not been moved or altered) to
place a surveyor's rod at the location where the second man was standing, shoot an
image, overlay that image with the image of the man, and directly measure his height.
The parties agree that the laundromat was remodeled at some time after the trial
eliminating the possibility of locating the exact spot the second man was standing
based on landmarks in the image.
Dr. Hartley admitted on cross-examination that a direct-measurement,
surveyor's-rod technique as described by Dr. Russ could have been performed at the
time of the initial investigation. Dr. Hartley also admitted that such a method would
have been more accurate than his own method. Dr. Hartley made this admission,
however, late in the evening during the extended hearing. Dr. Hartley provided an
affidavit after the hearing in which he acknowledged that the technique using a
surveyor's rod could have been employed at the time of the initial investigation. Dr.
Hartley's affidavit, however, disowned the admission regarding the relative accuracy
of his method and the surveyor's rod method. According to the affidavit, Dr. Hartley
attributed his concession regarding the accuracy of the surveyor's-rod, direct-
measurement technique to the fact that he was tired and not focusing well during the
late-night hearing.
After the hearings, Nooner unsuccessfully sought an expanded evidentiary
hearing to address: (1) factual issues surrounding an argument by the state that he had
not met the diligence requirement in procuring the testimony from his three
witnesses; and (2) a plenary evidentiary hearing regarding innocence. In his request
for a plenary hearing, Nooner sought the opportunity to present evidence as to all
contested issues in the case, but did not identify with particularity any witnesses or
evidence that he sought to present. The district court denied the request for an
expanded hearing.
-14-
Next, the district court denied relief on several grounds. First, the court
concluded that the video-image analysis was not new evidence. The court noted that
the techniques Dr. Hartley applied were developed after the trial but held the actual
evidence was the height of the murderer, not the analytical techniques. Because the
surveyor-rod method could have been employed pre-trial, the court concluded,
"Before the remodeling and replacement of the objects in the laundromat, Dr.
Hartley's sophisticated methods of estimating the height of the murderer in the
surveillance video were not necessary. That they are necessary now does not make
the height of the murderer newly discovered evidence that could not have been
discovered previously through the exercise of due diligence."
The court then held that, even if it were appropriate to consider Dr. Hartley's
opinion, it was unreliable and unconvincing and Nooner had failed to show it is more
likely than not that no reasonable juror would have found him guilty beyond a
reasonable doubt. The court criticized Dr. Hartley's stride adjustment as unscientific.
The court also criticized the failure to perform a comparable measurement on another,
measurable object in the image as proof of the validity of Dr. Hartley's technique as
applied to the images in this case. The district court also noted that it did not believe
Dr. Hartley adequately addressed the fact that the second man in the image appeared
to be slightly bent at the waist and leaning forward. Further, the court noted simply
that Rockett was 66 inches tall and 115 pounds, Nooner was 68–69 inches tall and
165 pounds, and Stobaugh was 70 inches tall and 171 pounds. The court also noted
that Stobaugh and the murderer appeared to be approximately the same size. Finally,
the court emphasized that other witnesses identified Nooner by his clothing not
merely by his height or appearance, and that the height evidence did not undermine
these other identifications.
Regarding Antonia Kennedy and Rockett, the court noted inconsistencies
between their respective statements and internal inconsistencies in Rockett's
declarations and testimony, as quoted above. In addition, the court noted apparent
-15-
inconsistencies between the statements and events observed in the video images and
Rockett's statement. For example, the court noted that Rockett admitted he was left
handed whereas the shooter appeared to be right handed. Also, Rockett claimed to
be the driver and stated he never let Nooner drive, whereas the images appear to show
headlights moving outside the laundromat while the murderer is in the laundromat,
as though the car is being repositioned to leave. Further, the court emphasized that
both Antonia Kennedy and Rockett still maintain that Nooner confessed to being the
shooter, although both claim now that they believe Nooner's confessions to them were
false.
Finally, the court addressed the general skepticism with which the Supreme
Court and Eighth Circuit have treated recantation testimony, the general lack of risk
to Rockett in recanting (he has already been sentenced to life imprisonment for his
participation in Stobaugh's murder), and the fact that the evidence Nooner relies upon
"to show his innocence contradicts rather than supplements the evidence upon which
he relied at trial" because all purportedly new evidence would place Nooner in the
car, whereas at trial he claimed to be home in bed.
Even though the court rejected the attempted showing of actual innocence, the
court held in the alternative that most of the purported Brady evidence Nooner relied
upon for his substantive claim was neither new nor previously unavailable. One
possible exception was the purported coercing of a false statement from Antonia
Kennedy by police investigators. As to this point, the court stated that Antonia
Kennedy had answered direct questions from the court at the evidentiary hearing
disavowing any claim to have been instructed or asked by police to testify falsely.
Nooner then filed a motion to reconsider. The court rejected the motion,
issuing a detailed order expanding on its analysis. In the motion, Nooner asserted it
was not known whether the camera Dr. Hartley used for a reference height was the
same or different than the camera in place at the time of the murder. The court stated
-16-
that this assertion undercut Nooner's position, because Dr. Hartley relied upon the
measured height of the camera as his only reference point. The court concluded:
At one point Nooner criticizes the Court's observations regarding the
relative size of the gunman as compared to Scot Stobaugh as
"unscientific and unreliable," but the Court was simply describing how
reasonable jurors would likely see the evidence. . . . Nooner would
have the Court say that reasonable jurors would be more likely to
believe Dr. Hartley's calculations than what they can see with their own
eyes, but jurors are not obliged to, and are not likely to, disregard what
they can see in deference to an esoteric mathematical and technological
analysis the accuracy of which cannot be verified. Although Nooner
asserts that the Court is "flatly incorrect" in saying that it is impossible
to say the extent to which Dr. Hartley's calculations correspond to
ground truth, the Court does not agree. Because he did not calculate the
height of any object in the videotape the height of which could be
independently confirmed, it is impossible to know the extent to which
Dr. Hartley's calculations correspond to ground truth.
II. Discussion
A. Successive Habeas Standards
Nooner asserts a stand-alone claim of actual innocence in addition to asserting
actual innocence as a "gateway" issue to permit his presentation of a Brady claim in
this successive petition.5 For gateway purposes, to enable presentation of a claim
5
The Supreme Court has suggested that a stand-alone claim of actual innocence
may be cognizable. See House v. Bell, 547 U.S. 518, 554 (2006); Herrera, 506 U.S.
at 417. The Court has not articulated a standard of proof that might apply to such a
claim, but some justices have stated that any such standard would be more stringent
than that required to prove actual innocence for gateway purposes. See, e.g., Herrera,
506 U.S. at 426 (O'Connor, J., concurring) (explaining that an "extraordinarily high"
standard of review would apply to any such claim because the conviction under attack
-17-
alleging a constitutional violation, a state prisoner seeking relief in a successive
federal habeas proceeding must show that:
(i) the factual predicate for the claim could not have been discovered
previously through the exercise of due diligence; and
(ii) the facts underlying the claim, if proven and viewed in light of the
evidence as a whole, would be sufficient to establish by clear and
convincing evidence that, but for constitutional error, no reasonable
factfinder would have found the applicant guilty of the underlying
offense.
28 U.S.C. § 2244(b)(2)(B). The phrase "evidence as a whole," as used in
§ 2244(b)(2)(B)(ii), refers to the entirety of the trial evidence as well as new evidence
offered in the collateral proceedings. Regarding evidence offered in the collateral
proceedings, the petitioner must "come forward not only with new reliable evidence
which was not presented at trial, but . . . come forward with new reliable evidence
which was not available at trial through the exercise of due diligence." Kidd v.
Norman, 651 F.3d 947, 953 (8th Cir. 2011).
When assessing the likely impact of this overall body of evidence upon
reasonable jurors, the court itself does not view the evidence through a personal and
presumably would have resulted from an underlying state trial free of any prejudicial
constitutional error, unlike gateway claims of actual innocence where the prisoner
asserts the underlying conviction is tainted by error of constitutional magnitude);
Dansby v. Norris, 682 F.3d 711, 716 (8th Cir. 2012) (stating that the standard
applicable to a stand-alone claim of actual innocence, if cognizable at all, would be
extraordinarily high and higher than the standard for a gateway assertion of
innocence). In any event, there is no authority suggesting the standard would be more
lenient than that applied to gateway claims of actual innocence. Accordingly, where
it is determined that a prisoner fails to make an adequate showing of actual innocence
for gateway purposes, it is unnecessary to address separately the possibility of a
stand-alone claim of actual innocence.
-18-
subjective lens. Rather, the court must conduct an objective analysis of how
hypothetical, rational jurors likely would view the evidence. See Schlup v. Delo, 513
U.S. 298, 329 (1995) ("It is not the district court's independent judgment as to
whether reasonable doubt exists that the standard addresses; rather the standard
requires the district court to make a probabilistic determination about what
reasonable, properly instructed jurors would do."). Here, the district court determined
that the purportedly new evidence was unreliable and that no reasonable juror would
fail to find Nooner guilty beyond a reasonable doubt based upon the totality of the
evidence.
The parties disagree as to whether our entire review is de novo or whether the
district court's underlying reliability assessment is a factual determination that we
must review with deference. In similar contexts, we have held that determination of
the reliability of purportedly new evidence is "a credibility determination which is
entitled to great deference." Amrine v. Bowersox, 238 F.3d 1023, 1029 (8th Cir.
2001) (deferring to a district court's credibility assessment regarding the reliability
of a recanting witness's new testimony); see also Storey v. Roper, 603 F.3d 507, 524
(8th Cir. 2010) (applying clear error standard to review district court's assessment of
new evidence); Wooten v. Norris, 578 F.3d 767, 782 (8th Cir. 2009) (reviewing for
clear error a district court's factual determination regarding the credibility of evidence
of actual innocence). Accordingly, we review the district court's credibility
determinations for clear error, and we review de novo its overall assessment of how
hypothetical, rational jurors would view the entire record as to the issue of actual
innocence. See Kidd, 651 F.3d at 952 n.5 (stating in the context of an actual
innocence analysis that a district court's "credibility determination is virtually
unreviewable on appeal"); Raymond v. Weber, 552 F.3d 680, 683 (8th Cir. 2009)
("We review a district court's denial of habeas relief for clear error in respect to
findings of fact, and de novo for questions of law or for mixed questions of law and
fact.").
-19-
B. Actual Innocence—New Evidence and Adequacy of Nooner's Showing
Nooner presents several interrelated arguments relevant to the threshold
showing of actual innocence and to the district court's various procedural rulings. For
the reasons stated below, we find clear error neither in the district court's
determination that the expert testimony is not "new" evidence nor in its reliability and
credibility assessments. We also find no error in the district court's predictive
analysis of how actual jurors would view the totality of the evidence regarding actual
innocence. In addition, even assuming error occurred as to one or more of the
procedural issues that Nooner raises, we find no prejudice that might undermine the
district court's credibility assessments or actual innocence determination.
i. New Evidence
Dr. Hartley's opinion of the second man's height cannot be deemed new
evidence. See Kidd, 651 F.3d at 953 (holding that "new evidence" is evidence that
could not have been offered at trial through the exercise of due diligence). Although
Dr. Hartley contests the accuracy that could have been attained with the surveyor's-
rod, direct-measurement method described above, he admits such a technique could
have been employed prior to trial. In our view, minor potential differences in the
relative accuracies of Dr. Hartley's current method and the simple method using a
surveyor's rod are not material to the issue of whether the present height evidence is
"new evidence" as described in Kidd. See McDonald v. Bowersox, 125 F.3d 1183,
1186 (8th Cir. 1997) (per curiam) (rejecting a petitioner's argument that the improved
ability to diagnose and treat mental afflictions in the ten years since a trial made a
psychiatric opinion proffered post-trial "new evidence").
Neither before Nooner's 1993 trial, before his initial petitions for collateral
relief, nor prior to 2007, did Nooner proffer evidence attempting to distinguish his
own height from the height of the second man in the video images. At all times,
-20-
however, there existed the capability through simple methods to ascertain the height
of the second man in the image.6 For whatever reason, defense counsel elected not
do so. Accordingly, the present availability of a different measuring technique does
not make an opinion based on the new technique "new evidence." See id.; Noel v.
Norris, 322 F.3d 500, 503 (8th Cir. 2003) (rejecting an argument that brain-scan
evidence based on a technique developed after sentencing could serve as "newly
discovered evidence" in support of a claim for initial habeas relief).
We note also that there is nothing to suggest Nooner's failure to present some
form of height evidence at trial is attributable to anything other than trial strategy or
a lack of diligence. Even if we were to view Rockett's present confession as true,
then Nooner would have to have known at the time of trial that Rockett was the
shooter—Rockett placed Nooner outside the laundromat in the car. Nooner, however,
chose neither to stand mute nor rely upon the purported knowledge that Rockett was
the shooter (and use height evidence to prove Rockett rather than Nooner was the
second man in the video). Instead, Nooner elected to present testimony placing
himself at home in bed at the time of the murder. Previously available evidence is not
new simply because a defendant elects to change strategies long after the conclusion
of his state trial. See Wainwright v. Sykes, 433 U.S. 72, 90 (1977) ("[T]he state trial
on the merits [should be] the 'main event,' so to speak, rather than a 'tryout on the
road' for what will later be the determinative federal habeas hearing.").
6
We do not purport to address a situation where the newly proffered scientific
evidence truly represents exonerating evidence that could not have been obtained at
trial. For example, we do not suggest DNA analyses cannot be deemed new evidence
if the trial occurred before there were any scientific capabilities to accurately identify
the source of a biological sample using DNA. We hold merely that where the
capability of obtaining data from evidence existed at trial, and the defendant elected
not to utilize that capability, data derived from a subsequently developed technique
generally should not be deemed "new."
-21-
ii. Credibility Assessment
We also hold that even if Antonia Kennedy's recantation, Rockett's confession,
and Dr. Hartley's analysis could be deemed new evidence, the district court did not
clearly err in its credibility assessments. See Amrine, 238 F.3d at 1028 (8th Cir.
2001) (affirming a denial of habeas relief where "the only new evidence proffered . . .
was unreliable, [and] the court concluded that no further Schlup analysis was
necessary."). The district court identified internal inconsistencies in Rockett's
handwritten and typed declarations and in Rockett's testimony from the evidentiary
hearing, as well as inconsistencies between Antonia Kennedy's recantation and
Rockett's confession. For example, Rockett denied in one written statement that he
had seen Stobaugh's checkbook or had it in a car, but stated in his other written
statement that he possessed the checkbook and gave it to Antonia Kennedy. Rockett
claimed with great specificity to have formulated his plan to pin Stobaugh's murder
on Nooner while in police custody on a particular date and in response to overhearing
Nooner talking to police. Rockett, however, first inculpated Nooner prior to the day
that Nooner spoke to police and could have been overheard by Rockett. Further,
Antonia Kennedy stated that Rockett admitted to killing Stobaugh the morning after
the murder and that Rockett told her to place the blame on Nooner. Rockett,
however, denied telling Antonia Kennedy the day after the murder that he, rather than
Nooner, had shot Stobaugh. As a result, it is not clear if Nooner now asserts Antonia
Kennedy and Rockett created the purported plot to falsely blame Nooner the morning
after the killing or sometime after Rockett was in police custody.
Additional inconsistencies bear on the witnesses' credibility and motivations
for recanting or confessing. Rockett denies having been sexually involved with
Antonia Kennedy, but she claims to have been sexually involved with Rockett prior
to Stobaugh's murder. Further, Antonia Kennedy's history of making contradictory
statements in this case includes not only her versions of the facts as delivered to
police, the defense investigator, and the jury, but also her allegation of police
-22-
misconduct. She signed a declaration asserting that police told her to lie and blame
Nooner, but she told the district court during the evidentiary hearing that police did
not tell her to lie or name Nooner as the shooter. The district court emphasized that
Rockett had nothing to lose by confessing because he was already serving a life
sentence for his participation in Stobaugh's murder. Finally, both Antonia Kennedy
and Rockett maintain that Nooner claimed to be the shooter the morning after the
murder, even though they assert that Rocket rather than Nooner actually was the
shooter.
Nooner challenges the district court's credibility assessment in two ways. First,
he argues we owe no deference to the district court's credibility determinations
because they are not separate factual determinations, but rather, are merely part of the
overall actual innocence analysis that we should review de novo. Second, he attempts
to minimize the impact of the inconsistencies and argues he was improperly denied
discovery and an expanded hearing as necessary to fully analyze the evidence of
actual innocence.
As stated above, circuit precedent requires that we treat the district court's
credibility/reliability assessment as a factual finding, and—as in most similar
circumstances—defer to the superior fact-finding ability of the district court who
possessed more than a cold record and who took live testimony from the witnesses
and observed their demeanor at the evidentiary hearing. Kidd, 651 F.3d at 952 n.5.
Regarding Nooner's factual arguments, he fails to reconcile the listed
inconsistencies. Rockett and Antonia Kennedy described Nooner as a "big talker"
who wanted to be viewed as a killer and who falsely claimed Rockett's exploits as his
own. Nooner argues that Antonia Kennedy's selective recantation and her continued
adherence to the claim that Nooner made a morning-after confession is consistent
with the view that Nooner was merely a "big talker." This argument is tenuous at best
and simply fails to convince that the district court committed clear error, especially
-23-
in light of the skeptical manner in which courts are to treat long-after-the-fact
recantation testimony. See United States v. Holmes, 421 F.3d 683, 688 (8th Cir.
2005) ("[W]e view [recanted testimony] and its potential evidentiary value at a new
trial with suspicion."); United States v. Rouse, 410 F.3d 1005, 1009 (8th Cir. 2005)
("The stability and finality of verdicts would be greatly disturbed if courts were too
ready to entertain testimony from witnesses who have changed their minds, or who
claim to have lied at the trial." (citation and internal quotation marks omitted)).
Nooner also fails to demonstrate clear error in the district court's assessment
of Dr. Hartley's reliability. Dr. Hartley's methods were, in the abstract and applied to
hypothetical, clear images, unchallenged. Here, however, Dr. Hartley took several
steps in reaching his conclusions that are easily understood by lay persons and lack
scientific rigor. As noted, he manually identified a location on the image he believed
to represent a point on the floor beneath the second man's head to use as a bottom
point for measurements. He empirically assessed the impact of shoes on height by
using a sample size of two pairs of his own shoes and estimating how much height
shoes (generically defined) might add to the man in the image. He assessed a height
reduction due to stride by placing an unidentified "thing" on his own head and tracing
his height on a wall was he walked. And, he simply created what he believed to be
a conservative estimate of a height that might be added due to a hat.
Adding to these potential sources of unreliability are the facts that (1) even
before making the shoe and hat adjustment Dr. Hartley was only 75–80% confident
in his result; (2) the height evidence in no manner explained why the two men in the
image appeared to be roughly the same size even though Rocket was fifty-five pounds
lighter than Stobaugh and forty-nine pounds lighter than Nooner, whereas Nooner
and Stobaugh were roughly the same height and weight; (3) Dr. Hartley performed
no control measurement upon Stobaugh or any object in the image to illustrate how
his method actually performed when applied to the particular, poor-quality image
obtained from the videotape in this case; and (4) Dr. Hartley's conclusion was
-24-
inconsistent with the other identification testimony which relied upon clothing rather
than height.
Finally, the district court stated that Dr. Hartley had not sufficiently taken into
account the fact that the second man in the image appeared to be bending at the waist.
Nooner argues Dr. Hartley adequately addressed this issue through a discussion of
"foreshortening," a visual effect caused when an image is taken from above.
According to Dr. Hartley, the foreshortening effect causes the top and bottom of an
object in an image to appear with slightly different scales such that the object appears
to be leaning towards the camera. The district court did not clearly err in this regard
because, whereas foreshortening may explain why an object appears to be leaning,
it does not explain why a person in an image would appear to be bending at the waist.
This point is particularly strong in the present case where the critical image shows the
second man from a partial side view in which he appears to be bending slightly at the
waist in a direction perpendicular to the line of sight of the camera. This visual image
is simply inconsistent with Nooner's explanation of foreshortening.
iii. Actual Innocence
Having determined that the Nooner's evidence is not reliable and Dr. Hartley's
opinion is not new, it is unnecessary to conduct a further Schlup analysis. Amrine,
238 F.3d at 1028. In the alternative, however, even considering the proffered
evidence in the context of the trial evidence, the evidence was "not particularly
convincing and easily marginalized." Storey, 603 F.3d at 524 (internal quotation
marks omitted) (reviewing a district court's assessment of purportedly "new evidence"
and holding the petitioner failed to demonstrate actual innocence). Here, the jury that
convicted Nooner actually knew Antonia Kennedy had given two inconsistent
statements prior to trial. Hypothetical, rational jurors, then, would have to reconcile
the fact that she previously was confronted with her conflicting statements and
declared the statement inculpating Nooner to be the truth whereas her long-after-the-
-25-
fact reversal and partial recantation disclaimed the truth of that statement. Such a jury
would also be required to reconcile Nooner's evidence as presented at trial—in which
he claimed to be home in bed at the time of the shooting—with his morning-after
confession to Antonia Kennedy, and with Rockett's only-partially-consistent
confession (all of which placed him at the scene). Further, even the evidence Nooner
chose to present at trial was internally inconsistent in that his step-father testified
Terri Kennedy and Nooner's son were not at home with Nooner the night of the
murder, whereas Terri Kennedy testified that the child was with Nooner on that night.
Finally, nothing in the evidence Nooner asserts as new undermines the testimony of
Jazmaar Kennedy or Johnny Martin who testified that they had seen Nooner wearing
the same clothes as the second man in the image prior to the murder. Similarly, the
purportedly new evidence does not undermine the testimony of Isaac Warren who
stated that he saw Nooner with a 0.22-caliber Ruger pistol prior to the murder, but not
after the murder.
Nooner does not address these factual issues head-on, arguing instead that he
was denied the opportunity to present additional evidence needed to prove his
innocence, that the district court erred in denying discovery and in admitting and
relying upon the opinion of Dr. Russ, and that the district court misconstrued the
actual innocence standard. For the reasons explained below, these arguments fail to
show that "no reasonable factfinder would have found [Nooner] guilty of the
underlying offense." 28 U.S.C. § 2244(b)(2)(B)(ii).
Regarding the opportunity to present additional evidence, Nooner argues he
should have been granted a plenary evidentiary hearing. We review the district
court's decision to grant, deny, or limit the scope of an evidentiary hearing during
habeas proceedings only for abuse of discretion. See Schriro v. Landrigan, 550 U.S.
465, 480–81 (2007) (holding that a district court did not abuse its discretion in
refusing an evidentiary hearing on a habeas claim); Schlup, 513 U.S. at 332
(remanding and noting "the District Court's ability to take testimony from the few key
-26-
witnesses if it deems that course advisable" (emphasis added)); Wadlington v. United
States, 428 F.3d 779, 784 n.2 (8th Cir. 2005) ("[T]he court's failure to hold an
evidentiary hearing to evaluate Hood's new testimony was not an abuse of discretion.
An evidentiary hearing is not necessary where, as in this case, the district judge
observed the demeanor and credibility of the witness at trial or is otherwise
thoroughly familiar with the record of the case." (internal citation omitted)). On
appeal Nooner argues specifically that he should have been permitted to call Jazmaar
Kennedy and Johnny Martin to challenge their trial testimony in light of the other
evidence he presented at the hearing.
In contrast to his specific argument to our court, Nooner asked the district court
in general terms to present evidence or arguments beyond the expert opinion of Dr.
Hartley and the recantations of Antonia Kennedy and Rockett. Nooner argued
generally that all earlier identification evidence based on the video was unreliable in
light of Dr. Hartley's new opinion, the recantation, and the confession. In attacking
the reliability of earlier video-identification testimony, however, Nooner named
several witnesses other than Jazmaar Kennedy and Johnny Martin. Nooner did not
seek to present testimony specifically from Jazmaar Kennedy or Johnny Martin, nor
did he allege that either witness had recanted.
Even if he had specifically sought to introduce testimony from Jazmaar
Kennedy or Johnny Martin, though, these witnesses identified Nooner in the
surveillance video based on the clothing he was wearing, and it is difficult to
appreciate how Dr. Hartley's discussion of height would impact identifications related
to clothing. In short, Nooner made neither a specific request nor a specific proffer to
show what he wanted to place into evidence.7
7
Further, at oral argument to our court, counsel stated that no preparations had
been made to present such witnesses. Accordingly, it is not the case that Jazmaar
Kennedy and Johnny Martin were present to testify or had been subpoenaed.
-27-
Given the vague nature of Nooner's request, the district court did not abuse its
discretion in holding a hearing on only the three items the district court properly
understood to be most pertinent to Nooner's claims: the expert witness testimony and
the veracity of the two recantations. Schlup does not hold habeas prisoners have
unlimited rights to plenary hearings. Instead, the Court stated lower courts may hear
evidence from "key witnesses." See Schlup, 513 U.S. at 332 (noting the "fact-
intensive nature of the inquiry" and the "District Court's ability to take testimony from
the few key witnesses").
To the extent Nooner argues procedural error surrounding Dr. Russ infected
the factual analysis, his arguments are without merit. As to both the motion for
discovery and the motion to exclude, any error that may have occurred caused no
prejudice to Nooner and provides no basis to disturb the district court's judgment.
See Barrett v. Rhodia, Inc., 606 F.3d 975, 980 (8th Cir. 2010) ("We will not reverse
a district court's ruling on the admissibility of expert testimony absent a clear and
prejudicial abuse of discretion.") (internal quotation marks omitted); Sheets v. Butera,
389 F.3d 772, 780 (8th Cir. 2004) ("'Our review of questions concerning discovery
matters is very deferential. . . . We will not reverse such a determination absent a
gross abuse of discretion resulting in fundamental unfairness in the trial of the case.'")
(quoting SDI Operating P'ship v. Neuwirth, 973 F.2d 652, 655 (8th Cir. 1992)).
Prejudice stemming from the admission of Dr. Russ as an expert necessarily would
have to present itself through the court's reliance upon Dr. Russ's technical expertise,
i.e. reliance upon his opinions as an expert. Here, however, to the extent the district
court relied upon Dr. Russ, it did so only as to matters well within the understanding
of lay persons.
We have held that "[w]here the subject matter is within the knowledge or
experience of lay people, expert testimony is superfluous." United States v.
Coutentos, 651 F.3d 809, 821 (8th Cir. 2011) (internal quotation marks omitted). We
have also held that, as to issues of medical causation or deliberate indifference to
-28-
medical needs, expert testimony is unnecessary where the medical condition or issue
of causation "would have been obvious to a layman." Roberson v. Bradshaw, 198
F.3d 645, 648 (8th Cir. 1999). We believe it follows naturally from these cases that
no prejudice can be shown based on the improper admission of an expert or the denial
of discovery regarding an expert where a court's only reliance upon the expert goes
to matters on which the expert testimony was, in effect, superfluous.
The district court identified the failure of Dr. Hartley to conduct a comparative
measurement of some other person or object in the image to prove how his method
as applied to the image corresponded to "ground truth." Dr. Russ leveled this same
criticism at Dr. Hartley. We do not believe criticism of a measurement technique
based on the failure to perform a control measurement is the type of criticism that lies
beyond the understanding of lay persons. Similarly we do not believe criticism of Dr.
Hartley's ad hoc methods of determining adjustments for various methods lie outside
the realm of a lay person's understanding.
Nooner also argues the district court misconstrued the standard of Schlup v.
Delo and 28 U.S.C. § 2244(b)(2)(B) by failing to view the evidence through the eyes
of hypothetical, rational jurors. Nooner argues the district court did not pose the
question, "what would reasonable jurors do," but rather, assumed jurors to be flawed
human beings skeptical of science and technology who would never trust Dr.
Hartley's testimony over what they could see with their own eyes. Nooner's argument
appears to be wholly dependent upon the following passage from the district court's
written order denying a motion to alter or amend the initial judgment:
Nooner would have the court say that reasonable jurors would be more
likely to believe Dr. Hartley's calculations than what they can see with
their own eyes, but jurors are not obliged to, and are not likely to,
disregard what they can see in deference to an esoteric mathematical and
technological analysis the accuracy of which cannot be verified.
Although Nooner asserts that the Court is "flatly incorrect" in saying
-29-
that it is impossible to say the extent to which Dr. Hartley's calculations
correspond to ground truth, the Court does not agree. Because he did
not calculate the height of any object in the videotape the height of
which could be independently confirmed, it is impossible to know the
extent to which Dr. Hartley's calculations correspond to ground truth.
Nooner seizes upon the phrase "esoteric mathematical and technological analysis."
His emphasis, however, is misplaced. It seems clear the court was merely describing
how rational jurors would view the evidence the court had already determined to be
unreliable. In fact, the district court incorporated into the same passage its criticism
of Dr. Hartley based on his failure to conduct a control measurement of some other
person or object in the image.
Nooner's argument that the district court applied the wrong standard, then, is
really an argument that the district court should have been compelled to find Dr.
Hartley's opinions and conclusions (and the selectively recanted testimony) sufficient
to impart reasonable doubt in the minds of reasonable jurors. In this regard, however,
Nooner misconstrues the posture of the case. The evidence was neither proffered in
an effort to obtain an evidentiary hearing nor was it presented as a preliminary
manner that might require a court to view it in a light most favorable to Nooner.
Instead, Dr. Hartley's analysis was introduced at the evidentiary hearing to be
assessed by the court for reliability and for its likely impact upon potential jurors.
That Dr. Hartley's opinion may have been admissible does not mean the finder of fact
was required to rely upon it. The district court did not depart from the proper
predictive analysis merely because it determined Dr. Hartley's opinion was
unconvincing in light of other evidence.
Finally, Nooner argues that the extension of the second evidentiary hearing into
the night was structural error that infected the entire proceedings or, at a minimum,
was prejudicial in that Dr. Hartley conceded the prior availability and accuracy of a
simpler height-measurement technique. In arguing that the error was structural and
-30-
requires no showing of prejudice, Nooner characterizes potential performance
deficiencies that might arise due to exhaustion or confusion of counsel as the
complete denial of counsel. He points to actual conflict of interest cases and the
complete denial of counsel as structural error that obviates any need to show
prejudice. See, e.g., Holloway v. Arkansas, 435 U.S. 475, 490–91 (1978) (deeming
an actual conflict of interest by counsel as structural error). We find no authority to
support his attempted analogy. The list of errors deemed structural is not a list we
expand lightly, and we decline the invitation to label any error by the district court
in exercising its discretion regarding management of the evidentiary hearing in this
case as structural. See United States v. Jones, 662 F.3d 1018, 1028 (8th Cir. 2011)
("These extreme errors have been recognized in a very limited set of circumstances.").
Regarding Dr. Hartley's concession, we find no prejudice. Even in his post-
hearing affidavit, Dr. Hartley did not disown his statement that the surveyor-rod
method for measuring the height of the second man in the image could have been
performed at the time of the initial investigation. Dr. Hartley retracted his statement
that the simpler method would have been more accurate. The minor differences in
the relative accuracy of the two methods, however, is not outcome determinative in
this case. Accordingly, we find no prejudice.
Because the district court did not err in its credibility assessments, and because
Dr. Hartley's opinion is not "new" evidence, Nooner cannot meet the threshold
requirements of § 2244(b)(2)(B). In the alternative, we agree with the district court's
analysis of the evidence as a whole and the determination that Nooner failed to prove
it more likely than not that no reasonable juror would have found him guilty beyond
a reasonable doubt. Nooner's arguments that these determinations were infected with
structural error or prejudice are without merit.
-31-
We affirm the judgment of the district court.8
______________________________
8
The pending motion to expand the record regarding extension of the second
evidentiary hearing is denied as moot given our resolution of this issue. The pending
motion for recusal of all circuit judges also is denied.
-32-