filed a concurring opinion.
Back in 2007, I voted to allow the applicant in this cause to proceed on the merits of a subsequent writ application because I believed that her application contained previously unavailable specific facts sufficient to make out a prima facie case for a constitutional claim of actual innocence.1 Now that the applicant has had a chance to develop a record in support of her claim of actual innocence, I agree with both Judge Keasler (and, implicitly, Judge Cochran) that her evidence falls short of satisfying the “Herculean” burden imposed on applicants making a bare claim of actual innocence under the standard we laid out in Ex parte Elizondo,2 And yet, the convicting court has recommended that we grant the applicant a new trial, and the State has declared itself content to go along with that ultimate recommendation. I write separately to explain why, particularly in light of my concurring opinion last year in Ex parte Robbins,3 I, too, am content to grant the applicant a new trial in this case — but on the basis of the inadvertent use of false evidence rather than actual innocence.
*835A bare claim of actual innocence and a claim that false evidence was inadvertently used to obtain a conviction both fall along a continuum of due process violations. At one end of the continuum is a claim that the State has knowingly used false or perjured testimony. Here, due process is primarily concerned with the fairness of the trial.4 Because of the State’s complicity in undermining the integrity of the process, the standard for materiality is comparatively low: a reasonable possibility that the false or perjured testimony contributed to the conviction.5 At the other end of the continuum is a bare claim of actual innocence. An actual innocence claim does not depend upon a showing of misconduct of any kind on the part of the State. The due process concern is with the accuracy of the result.6 For that reason, an actual innocence claim will result in habeas corpus relief only upon a showing of extreme materiality: the applicant must be able to show by clear and convincing evidence that, given the newly available evidence of innocence in addition to the inculpatory evidence presented at trial, no reasonable juror would have convicted him.7 And, as Presiding Judge Keller has advocated, and I agree, “[t]he unknowing use of perjured or false testimony falls [or at least should fall] in between these endpoints, with a mid-level standard (or standards) of materiality.”8 Thus, as the particular due process claim moves from the fairness end of the continuum toward the accuracy end, the standard for materiality should rise concomitantly, culminating in the Herculean burden associated with a bare claim of actual innocence.
Though I do not think that the applicant has proven actual innocence in this case, I do believe that she has established that her conviction violated her right to due process. She has proven to my satisfaction that her conviction was based in critical part upon an opinion from the medical examiner that he has now disowned because it has been shown by subsequent scientific developments to be highly questionable. In Robbins, a non-death-penalty capital murder case, the applicant raised claims of actual innocence and the inadvertent use of false evidence. I joined the majority opinion denying relief on both claims, but wrote separately to express my view that, “when it comes to claims of the inadvertent use of false evidence, we must not be overly liberal in how we characterize ‘false’ evidence.”9 One immediately obvious difference between Robbins and the instant case is that Robbins did not involve intervening scientific developments — the medical examiner in Robbins simply changed her mind between the time of her trial testimony and the post-conviction writ proceedings.10 Here, the medical *836examiner currently believes that intervening research in the science of biomechanics has undercut his trial testimony, and the convicting court has accepted that as a sufficiently accurate statement of the condition of current scientific/medical knowledge to justify the conclusion that the applicant’s trial was rendered unfair — and quite possibly inaccurate — by the medical examiner’s now-disowned trial testimony. Under these circumstances, I am far less hesitant to characterize the trial testimony as “false.”
Nor does it trouble me that, when the applicant filed her subsequent writ application in 2007, she did not include a false-evidence claim, per se. Having filed her subsequent writ application two-and-a-half years before we issued our opinion in Ex parte Chabot,11 she could not have known that we would come to entertain claims of the inadvertent use of false evidence.12 Even so, embedded within her claim of actual innocence are facts sufficient to allege such a claim,13 and in pursuing her actual innocence claim, she has ultimately developed facts sufficient to prove it.
My concurring opinion in Robbins was fueled by a concern that the materiality standard we have applied to claims of the inadvertent use of false or perjured evidence since Chabot may not be sufficiently stout. To my way of thinking, such claims fall closer to the accuracy-of-the-result actual-innocence end of the continuum than the fairness-of-the-process State-complicity end. Accordingly, while a claim that the State inadvertently used false or perjured testimony should not be subject to quite the same Herculean standard of materiality as an actual innocence claim, still, if we are to recognize it as a valid due process claim at all, we should subject it to a materiality standard that is relatively high. I did not think that Robbins could meet such a high standard of materiality and believed that to grant him relief under the circumstances of that case would effectively undermine our actual innocence jurisprudence. I therefore voted to deny relief in Robbins.
By contrast, on the facts of this case, as detailed in Judge Cochran’s concurring opinion, it is evident to me that the applicant can meet practically any standard of materiality less than that applicable to a bare actual innocence claim. In Ex paHe Chavez, I recently took the position that a subsequent post-conviction habeas applicant should not be permitted to proceed to the merits of his inadvertent-use-of-false-evidenee claim because, regardless of whatever standard of materiality we might apply, the subsequent applicant could not possibly meet it.14 In this case, my view is just the opposite. While she cannot satisfy the Herculean materiality standard to satisfy a claim of actual innocence, the present applicant has demonstrated materiality under whatever marginally lesser standard we may care to carve out to *837govern a claim that due process has been violated by the inadvertent use of false evidence. That is enough to convince me that the Court appropriately adopts the convicting court’s ultimate recommendation — acceded to by the State — to grant the applicant a new capital murder trial.
COCHRAN, J., filed a concurring opinion in which WOMACK, JOHNSON, and ALCALA, JJ., joined.
This case raises the same novel and difficult issue for the criminal-justice system that this Court faced, and, I maintain, fumbled in Ex parte Robbins:1 Changing science has cast doubt on the accuracy of the original jury verdict.2 Dr. Roberto Bayardo, who performed the autopsy of the victim in this case, testified unequivocally at trial that three-and-a-half-month-old Brandon Baugh “came to his death as a result of a severe closed head injury ... characteristic of abuse, homicide.” He concluded, without a scintilla of doubt, that the cause of Brandon’s death was a severe closed-head injury and the manner of death was homicide. Dr. Bayardo was the State’s star witness at trial on the cause and manner of Brandon’s death. But, based on advances in the science of pediatric head trauma, he has since changed his mind: “Based on the physical evidence in the case, I cannot determine with a reasonable degree of medical certainty whether Brandon Baugh’s injuries resulted from an intentional act or an accidental fall.” This scientific uncertainty about Brandon’s manner of death raises an extremely serious concern about the accuracy of the original jury verdict.3 I write separately to provide some factual context for the habe-as judge’s recommendation to grant a new trial, the State’s decision to agree with that recommendation, and this Court’s adoption of the habeas judge’s findings of fact and ultimate recommendation.
I.
On the morning of January 21, 1994, Eryn and Melissa Baugh left their infant son, Brandon, with applicant, their regular babysitter.4 That day, both applicant and Brandon disappeared. A kidnapping investigation began the next day. On February 1, applicant was arrested by the FBI in Kansas City. At first, applicant denied any knowledge of Brandon’s location or well-being. Later, she stated that Brandon’s grandmother, driving a car with Oklahoma license plates, picked him up during the afternoon of January 21. Eventually, applicant admitted that Brandon was dead, but claimed that his death *838was an accident. She also said that she had buried his body in a wooded area near Waco, that she had used a spade to dig the grave, and that she could take officers to his grave.
The single contested issue in the 1995 capital-murder trial was whether applicant intended to kill Brandon or whether she recklessly, negligently, or accidentally caused his death. In her statement, applicant contended that Brandon’s death was an accident-he accidentally fell from her arms onto a linoleum-covered concrete floor.5 The State’s primary evidence to prove that applicant intended to kill Brandon consisted of the circumstantial evidence produced by the autopsy.6 At trial, Dr. Roberto Bayardo, the long-time and highly experienced medical examiner for Travis County, testified that it was “impossible” for Brandon’s extensive brain injuries to have occurred in the way that applicant stated. He said that her story was false and “incredible.” In his opinion (and that of Dr. Sparks Veasey, the Deputy Chief Medical Examiner of Lubbock County), Brandon’s injuries must have resulted from an intentional blow.7 He concluded, “I would say the baby was caught up with the hands by the arms along the body and then swung and slammed very hard against a surface.” In Dr. Bayardo’s opinion, Brandon’s death was the result of child abuse: “this is the worst case of head injury [at the hands of a person] I [have] ever seen.” The jury agreed and convicted applicant of capital murder in May 1995.
This Court affirmed applicant’s conviction and death sentence on direct appeal8 and denied relief on her initial writ application in 2002.9 The federal district judge denied her federal habeas petition in 2004,10 and the Fifth Circuit affirmed that denial in 2006.11
The Honorable Jon Wisser presided over applicant’s 1995 trial and is currently presiding over her subsequent writ application. Judge Wisser was sufficiently troubled by the preliminary scientific evidence initially presented to him that, on April 4, 2007, he recalled applicant’s original death warrant and rescheduled her *839execution to give her sufficient time to gather additional material for this subsequent writ application.
On May 23, 2007, three weeks before her re-scheduled execution date of June 13, 2007, applicant filed this subsequent application for habeas relief based on recent scientific advances in the area of biomechanics and physics — advances that led Dr. Bayardo to recant his conclusive opinion that Brandon’s head injuries could not have been caused by an accidental, short-distance fall. Dr. Bayardo’s 2007 affidavit stated,
Since 1995, when I testified at Cathy Henderson’s trial, the medical profession has gained a greater understanding of pediatric head trauma and the extent of injuries that can occur in infants as a result of relatively short distance falls, based in part on the application of principles of physics and biomechanics. Specifically, and as shown in the reports that I have read, even a fall of a relatively short distance onto a hard surface can cause the degree of injury that Brandon Baugh experienced. If this new scientific information had been available to me in 1995, I would have taken it into account before attempting to formulate an opinion about the circumstances leading to the injury.
I have reviewed the affidavit of John Plunkett dated May 18, 2007,12 and I agree with his opinion. Based on the physical evidence in the case, I cannot determine with a reasonable degree of medical certainty whether Brandon Baugh’s injuries resulted from an intentional act or an accidental fall. In fact, had the new scientific information been available ,to me in 1995,1 would not have been able to testify the way I did about the degree of force needed to cause Brandon Baugh’s head injury.
Faced with this recantation, we held that applicant’s first two claims — (1) she is innocent of capital murder, and (2) but for constitutional errors she would not have been found guilty of capital murder — satisfied the requirements of article 11.071, § 5(a). We then remanded her application to the trial court for consideration of the merits. Judge Wisser held hearings between November 17, 2008, and March 5, 2009, and listened to the testimony of twelve witnesses. Seven of these witnesses were medical doctors and four were scientists with Ph.Ds. The twelfth witness was Linda Icenhauer-Ramirez, one of applicant’s trial attorneys.
At the habeas hearing, Dr. Bayardo testified that, because of recent scientific knowledge about how head injuries occur, he would no longer use words like “impossible” or “incredible” to describe applicant’s version of the events. Also, he would no longer assert that Brandon’s injuries, if they were caused by an accidental fall, would have to be the result of a fall from a height of over two stories. He stood by the cause of death: a heavy blow to the head. But he would change the manner of death from “homicide” to “undetermined.” He stood by his trial testimony that the comminuted depressed fracture of the back of the skull, which caused *840radiating fractures, was the result of a single blow. Dr. Bayardo said that, because Brandon sustained just one injury and that was to the back of his head, he doubted his prior finding of “homicide.” In other child homicides, the infants had multiple injuries to the side or front, rather than the back, of the head.13 Dr. Bay-ardo flatly contradicted his trial testimony when he concluded, “I don’t believe it’s a case of child abuse.”
Applicant also called Dr. Monson, an assistant professor of biomechanics at the University of Utah, who studies traumatic brain injury in children; Dr. Plunkett, a forensic pathologist who studies pediatric head trauma; and Dr. Van Ee, who has a Ph.D. in biomedical engineering and studies impact and orthopedic biomechanics. They testified that the application of biom-echanics to the study of pediatric head trauma and the medical community’s recognition of the role of biomechanics in determining causes of injury are recent and still developing. The medical community did not recognize a role for biome-chanics in cases such as this one in 1998, when applicant filed her first habeas application.
These experts testified about “drop” experiments conducted with crash-test dummies and infant cadavers that measured the impact and injuries involved in short-distance falls, and how those experiments show the potential for head injury and death when babies fall short distances. Dr. Monson testified that he calculated the g-force involved in the fall (as described by applicant) to be 120 to 163g. Dr. Monson could not rule out the possibility that Brandon’s death resulted from a short-distance fall: “So recognizing that the calculated values are well above when simple skull fracture may occur and also recognizing that there aren’t data defining exactly when a fracture of this severity may occur, I have to conclude that you simply cannot rule out that possibility.”
Dr. Plunkett testified about his playground-equipment study that included cases in which small children who fell short distances suffered complex skull fractures, brain injury, and death. He also could not rule out the possibility of an accident.
Q. As a pathologist who has done intensive work in this area and now is an author and speaker and expert in the field, do you think it’s scientifically plausible to offer an opinion on the cause of an infant’s death in a case like this without any review or application of biomechanics?
A. Not today, that’s not acceptable.
Q. Why not?
A. Unless your experience is in the area of bioengineering, very few physicians have the necessary knowledge to evaluate, to rigorously evaluate, Brandon’s injury. It’s got to go beyond medicine.
Q. Based on your studies and your work, Dr. Van Ee’s work, the CRABI14 dummy tests that you’ve seen, all of your review of the medical reports, and a portion of Brandon Baugh’s skull this morning, in your opinion, can any person conclude rationally and with certainty that the death of *841Brandon Baugh resulted from an intentional murderous act?
No. P>
Could Brandon, in your opinion, have suffered a complex comminuted skull fracture from a drop of four feet— four and a half feet landing on the back of his head on a hard floor? <o
Yes. ¡>
Dr. Van Ee performed an accident reconstruction using the CRABI-6 infant dummy. He testified that Brandon was probably traveling 11 miles per hour when he hit the concrete floor. “Severe injury is certainly a possibility and may even be likely for this type of impact.” He testified that Brandon’s skull fracture could have resulted from a fall of about four feet onto a carpeted or linoleum-covered concrete surface. He said there was now no “correct scientific support” for Dr. Bayar-do’s trial testimony.
Applicant’s experts generally concluded that it was possible that a short-distance fall, like the one that applicant reported, could have caused Brandon’s injuries. The State’s experts then testified about the limitations and shortcomings of the studies and experiments described by applicant’s experts and the rarity of head injuries with diffuse brain injury and complex skull fractures such as those Brandon suffered.
The State called Dr. Rangarajan, who has a Ph.D. in biomedical engineering and engineering mechanics. He designs crash-test dummies for government and industry use. Dr. Rangarajan designed a newborn-infant dummy for car-seat tests in Japan, and he testified about the limitations of using a dummy’s response to predict injury. He noted that “dummies are calibrated to perform well in automotive seated posture.” And, the dummies are only biof-idehc (able to produce true human-like responses) within calibration limits. He said that there are not enough biomechanics data to assess the probability of injury, but he did not know whether the tests conducted by the defense experts were an accepted way of predicting severe head injury: “I don’t know how to answer the question because I cannot say. I’m not like — I’m not the general secretary of the scientific community.”
Dr. Case, a medical examiner, neuropa-thologist, and forensic pathologist who has published articles on short-distance falls and pediatric head injuries, also testified for the State. She has written on how to distinguish between accidental head injuries and inflicted or abusive head injuries. A marker, “the presence of the diffuse distribution of subdural blood over the cerebral convexities,” signals brain injury produced by a blow to the head. Short falls, on the other hand, cause focal injuries, not diffuse ones.15 Only “one to two to three percent of all short falls will result in a simple linear skull fracture.” Brandon’s injury, “a depressed comminuted skull fracture,” signaled that he had suffered a blunt impact blow. “In my opinion, this is not going to result from a short fall [of] less than six feet. This fracture is not a fracture that I have ever seen in a short fall or that I have ever seen described in a short fall.” Brandon had a “diffuse subdural hemorrhage,” but Dr. Case could not tell whether it was the result of a single impact or multiple impacts.
Dr. Case disagreed with Dr. Bayardo about the significance of the fact that there were no other injuries on Brandon’s body. In her experience, 25 to 50 percent of children who died from abusive head inju*842ry had no other marks on their body. Dr. Case specifically criticized Dr. Plunkett’s work and testified that it was “on the fringe” and not widely accepted in the professional-medical community. Dr. Case thought the defense’s focus on the fracture was misguided because “You have to focus on the entirety of the head injury.”
Dr. Pustilnik, the Chief Medical Examiner for Galveston County and an assistant professor in pathology at UTMB, testified for the State and said that biomechanical engineering “looks at the body in what they call finite-point analysis.... In certain instances it falls apart when you have to look at differential ... mechanisms when you have both rigid structures in the body interacting with liquid or gelatinous structures in the body.” He said that there was “absolutely” a danger in taking readings from accelerometers in the crash-test dummies and making assumptions about how an infant of those dimensions would behave:
[IJt’s not any good to use the CRABI for head injury in these young infants because no one has studies on head injuries on infants the way they have studies on CRABI models.... They don’t have — you know, three-month-olds don’t have accelerometers behind their ears, so you can never know.
Dr. Pustilnik also criticized the cadaver studies because there were too many unknowns: Were the children stored cold, room temperature, or frozen? Were they dried out? Were they dropped with the scalp intact? “You just don’t know. It’s not in the papers.” It was Dr. Pustilnik’s opinion that Brandon suffered “multiple impacts to the back and left or perhaps even right left-side back area” of his head.
Dr. Jenny, a professor of pediatrics and the director of the child protection program at a hospital in Providence, Rhode Island, has studied child abuse and car safety and has specialized in treating babies with head injuries. She also questioned the usefulness of the defense studies and testified that, based on the severity of Brandon’s fractures, it was “highly likely” that he had been a victim of multiple impacts or crush injuries to his head. The fracture pattern was not consistent with a short fall from a care-giver’s arms. Dr. Jenny characterized Dr. Plunkett’s testimony about playground falls as disingenuous, and Dr. Van Ee’s accident reconstruction “drop” tests with the CRABI-6 as lacking a proper protocol. She said it was “impossible” for Brandon to have suffered his injuries in a short fall. She disagreed with the other State’s experts that such injuries were possible, though improbable, from a short fall. Those falls create “linear parietal” fractures, rather than complex fractures.
Dr. Gill-King, the director of the Laboratory of Forensic Anthropology and Human Identification, and a professor of forensic anthropology and pathology, who focuses on the material properties of bone, testified for the State. He likened the radiating fractures found on Brandon’s skull to those created from shooting a BB-gun multiple times at a window: “A fracture stops another fracture.” He disagreed with Dr. Bayardo that there was one blow to Brandon’s head because the injuries were distinct outside-to-inside injuries: “I concluded there were at least three separate applications of force. There may have been more, but there were at least three.” He also stated that the CRABI-6 dummy’s head does not in any way accurately portray the properties of a human infant head.
In rebuttal, applicant called Dr. Opho-ven, a pediatric forensic pathologist and medical examiner, and Dr. Stephens, a forensic pathologist and former medical examiner.
Dr. Ophoven described the “pendulum swing” in the medical community with re*843spect to pediatric injuries: before the 1980’s, a doctor would generally accept a family member’s report that a child’s head injury occurred accidentally. In the late 1980’s, doctors began to routinely disbelieve family members’ reports of children’s accidental head injuries and to assume abuse. Now, with studies applying biome-chanics to the field of pediatric head injuries, doctors are more cautious about “ruling out” the possibility that a child’s head injury occurred accidently. After reviewing the materials in this case, including the State’s experts’ testimony, Dr. Ophoven stated that short falls like that described here “rarely cause fatal injuries, but have the potential to kill.” She stated that biomedical analysis is important in cases like this one, where there is no evidence of pre-existing or fresh abuse or assault and where the history comes in as a possible accidental fall. She said that Dr. Van Ee’s and Dr. Monson’s reports support their opinions that fatal injuries can occur from a fall of 46 inches onto the concrete surface described here. She also said that Brandon’s injuries were consistent with a single blow; she did not see evidence of multiple impacts. She based this on “confluence bleeding from a single area of impact,” and the “clear continuity between the fractures in the back with the fractures on the side[.]” She said that she would not rely on a forensic anthropologist, like Dr. Gill-King, to help her with the cause of death or nature of injuries or number of impacts because “theirs is the area of bones and their training is not in making a determination and rendering it.”
Recent studies have clearly said, very clearly, you can not tell the difference between an accidental fracture and an inflicted fracture by how bad it is, how complicated it is, whether or not it crosses the suture line, whether or not it comes apart on the sides, that you cannot look at the fracture and say, I see abuse.
Dr. Ophoven summarized her testimony as follows:
So this issue here is, is it my opinion that there is science to study the force that could be generated in a fall? Yes. Do we have reasonable expectations of what’s the amount of force it takes for those falls to cause a fatal injury or death? Yes. Could those forces have been generated in this case? Yes. Do I have an opinion about what happened to Brandon? I can’t answer the question, but what I can say is he sure could have fallen and died.
The defense also called Dr. Stephens, who said that a medical examiner’s change of the manner of death from “homicide” to “undetermined” represents a paradigm shift. He did not regard Dr. Plunkett’s work as marginal but instead opined that some professionals in the medical community would not let go of old beliefs when faced with studies that challenged them. He personally had seen cases in which short-distance falls had killed children. He did not agree with much of what Dr. Pustilnik had to say, and discounted the significance of fracture lines hitting one another. He also agreed with Dr. Bayar-do that the injuries Brandon sustained were the product of a single blow or impact.
Applicant also called her trial attorney to testify that she filed a pre-trial motion for funds to employ a biomechanical expert, but her motion was denied.
In sum, all but one of these ten medical and scientific experts agreed that Dr. Bay-ardo’s trial testimony was now known to be scientifically inaccurate: Brandon’s autopsy results did not establish that his death was the product of an intentional homicide. Indeed, all but one of these experts basically admitted that science cannot answer the question of whether Brandon’s death was the result of an inten*844tional homicide. It could have been an intentional homicide; it could have been an accident. Based upon the totality of the evidence, Judge Wisser recommended that this Court
vacate the judgment of conviction in this cause, and ... order that Applicant be returned forthwith from her present place of confinement to the custody of the Sheriff of Travis County Texas, where she may thereafter be held to answer any indictment or other charges made against her arising out of the death of Brandon Baugh.
II.
Dr. Bayardo’s change in opinion on the manner of death from “homicide” to “undetermined” does not mean applicant is actually innocent of homicide. Nor does it mean that his trial testimony was “false” at the time it was given, based upon the state of scientific knowledge that he relied upon at that time. Due process was not violated at the time of trial, but nevertheless, the scientific testimony that supported a finding of “homicide” in the original trial has been retracted. Dr. Bayardo’s current scientific uncertainty, as well as the uncertainty of all but one of the experts at the habeas evidentiary hearing, casts a pall upon the basis for the jury’s verdict and upon its accuracy. At worst, the result of a change in the manner of Brandon’s death to “undetermined” is only an admission that science cannot resolve the issue of whether Brandon’s death was the result of a homicidal act. The jurors would have to decide that crucial question based upon the rest of the evidence.
The problem is that we do not know whether the jury would have found that applicant intentionally (as opposed to recklessly, negligently, or accidentally) caused Brandon’s death absent Dr. Bayardo’s expert scientific opinion.16 I recognize that this case does not fit neatly into our habe-as statute or our actual-innocence jurisprudence.17 But until the Supreme Court (or *845this Court) holds that a conviction later found to be based upon unreliable scientific evidence violates the Due Process Clause,18 I will stick by what I said in my Robbins dissent.
Who should decide whether the newly discovered unreliability of the expert scientific testimony was so crucial to the original jury’s verdict that the accuracy of that verdict can no longer be relied upon?
I fall back upon the wisdom and experience of the habeas judge — the “Johnny — on—the—Spot” factfinder to whom we will defer whenever the record supports his essential factual findings.19
Judge Wisser held a series of live hearings so that he could hear, first-hand, from all the experts. And he has concluded that the accuracy of the verdict can no longer be relied upon. The following findings— all supported by the record — are especially important to Judge Wisser’s recommendation of a new trial:
7. Dr. Roberto Bayardo served as the Chief Medical Examiner of Travis County for twenty-eight years, and personally conducted the autopsy of Brandon Baugh in February 1994.
8. [Applicant] claims that the infant died when he accidently fell from her arms to the concrete floor of her home, a distance of approximately four-and-one-half feet. At the trial, however, Dr. Bayardo strenuously disagreed that the infant’s death could have been accidental. Dr. Bayardo told the jury in no uncertain terms that “it would have been impossible” for an accidental fall to have produced the injuries sustained by the infant; that the claim of an accidental fall therefore was “incredible” and that to have sustained these injuries from a fall, the infant “would have to fall from the height higher than a two story building.” For these reasons, Dr. Bayardo opined in his autopsy report that the manner of death was “homicide,” meaning that the only way the infant could have sustained his fatal injuries was by means of a deliberate and murderous blow struck by Applicant.
9. Dr. Sparks Veasey, then the Deputy Chief Medical Examiner of Lubbock County, also testified for the State at the trial. He reviewed Dr. Bayardo’s autopsy report and photographs, and testified that the death resulted from blunt force trauma, probably resulting from “slamming into a wall or floor.” As Dr. Plunkett has pointed out, however, Dr. Veasey’s essentially repetitive testimony was uninformed by modern scientific learning and suffered the same vices as did the testimony of Dr. Bayardo. The State did not call Dr. Veasey as a witness at the eviden-tiary hearing. The Court does find that, if the jurors had heard Dr. Bay-ardo’s re-evaluation, they would not have credited the then-conflicting testimony of Dr. Veasey.
10. Based on the trial court record and the Court’s personal recollection of the trial, the Court finds that the trial testimony of Dr. Bayardo was the critical evidence upon which the con*846viction of Applicant rested, and was the evidence upon which the essential element of culpable mental state hinged. The “impossible,” “incredible,” “two story building” and “slamming” testimony of the State’s chief expert medical witness ruled out accidental cause. On that basis, the State persuaded the jurors that Applicant was guilty of murder beyond a reasonable doubt.
[[Image here]]
18.Dr. Bayardo and Dr. Plunkett both testified at the evidentiary hearing. Based upon the content of their written submissions, the testimony of each witness during direct and cross-examination, and the demeanor of each witness, the Court finds the Affidavit and testimony of Dr. Bayardo, and the Affidavit, report and testimony of Dr. Plunkett, are true, and that each of these witnesses is credible.
[[Image here]]
17. Dr. Bayardo testified at the evidentia-ry hearing that he continues to hold the opinions that he expressed in his Affidavit, and that [he] continues to share the views expressed by Dr. Plunkett in the latter’s evaluation. Addressing the words he used in his testimony to the jury in 1995, Dr. Bayardo testified at the evidentiary hearing that he would not use “impossible” or “incredible” were he to testify to the jury today, “because of the new knowledge about how these types of injuries occur.”
18. Dr. Bayardo also explained that he used the “two-story fall” analogy in his 1995 testimony because “that’s what I was taught during my residency and during my training that that’s the way that these types of injuries happen,” and “at that time, we didn’t have any information about the biom-echanical way of explaining these injuries.” The two-story analogy “would have been my usual answer in cases like this” in 1995, but he would not say so today, because the new scientific developments in biomechanical investigations “put a doubt in my mind.”
19. In his 1995 Medical Examiner’s Report Dr. Bayardo states that, in his professional opinion, the manner of the infant’s death was “homicide.” At the evidentiary hearing, however, Dr. Bayardo explained that he wrote “homicide” in 1994 because “at that time we didn’t have any information about the biomechanical way of explaining these injuries[.]” He then testified that if he were preparing his report today, he would not have opined that the manner of death was homicide, but instead, “I would leave it undetermined,”....
20. Applicant’s witness, Dr. Peter J. [Stephens], is also a former Medical Examiner. He testified that this reevaluation of the manner of death by Dr. Bayardo — from “homicide” to “undetermined” — was a significant “paradigm shift” representing a very fundamental re-evaluation.
21. In addition, Dr. Bayardo explained that there were other circumstances that now cause him to doubt that Brandon Baugh’s death was a homicide. He testified that
All the previous cases I’ve seen that were the result of a homicide injury had the multiple, recent, and other injuries and also had multiple fractures of ribs and extremities, and they also had multiple bruises or scrapes of similar ages. And this baby did not have any of those injuries, and that the location [of the injury] was also different from the other cases I’ve seen....
*84722. Because Dr. Bayardo’s testimony at trial was the critical evidence upon which the conviction of Applicant rested, and was the testimony upon which the essential element of culpable mental state hinged, the Court finds that if Dr. Bayardo’s re-evaluation had been presented to the jury in 1995, no rational juror could have or would have convicted Applicant of capital murder beyond a reasonable doubt in light of this new evidence.
In his Finding of Fact Number 23, Judge Wisser explicitly addressed the “guilty conscience” evidence detailed by Judge Keasler:
23. In making its Finding No. 22, the Court has not overlooked the State’s evidence at trial concerning Applicant’s flight after the death of the infant. While flight might be some evidence of a guilty conscience, it is equivocal evidence at best, and on the basis of personal recollection, the Court finds that the evidence of flight did not have the capacity to prove the mens rea of capital murder beyond a reasonable doubt.
24. In making its Findings, the Court has also considered the Applicant’s evidence of the new scientific analysis, unavailable at the trial, upon which Dr. Bayardo based his re-evaluation. The Court has done so in order to assure itself that Dr. Bayardo’s reevaluation is based upon a solid scientific foundation, and therefore supports the Court of Criminal Appeals’ statement that the re-evaluation is a “material exculpatory fact.”
25. In this regard, the Court has read, heard, considered, and evaluated the reports and materials of Drs. Plunk-ett, Stephens, and Monson upon which Dr. Bayardo relied in his Affidavit, . the trial testimony of these three experts, and the corroborating reports and evidentiary hearing testimony of Drs. Janice Ophoven and Chris Van Ee.
26. The Court finds that Dr. Plunkett’s Affidavit, his own report and materials, as well as the report of Dr. Peter Stephens, and the report of Dr. Kenneth Monson and the observed testimony of each of these witnesses at the evidentiary hearing, are truthful and credible.
27. Dr. Plunkett’s report in the present case, and those of the other experts, show that the 1995 testimony of the State’s chief experts was, at bottom, scientifically flawed and grounded upon the belief that “short-distance falls” can never case fatal infant head trauma. This belief no longer enjoys acceptance in today’s scientific community.
28. Dr. Monson made and explained a number of biomechanical calculations that he performed as an expert witness for Applicant, and concluded that the infant’s assumed fall from Applicant’s arms “had the potential to produce a serious injury,” and that “the possibility [of accidental death] cannot be ruled out given the current state of knowledge.” The Court finds that Dr. Monson’s Report and testimony are truthful and credible.
29. Dr. Monson testified that, from an assumed fall height of 46 inches from Ms. Henderson’s arms, the head of the infant would have been traveling at a speed of almost 11 miles an hour when it struck the floor, and that because the floor was a concrete, unyielding surface, the deceleration from 11 miles per hour to zero was virtually “instantaneous.”
30. Applicant’s witness Dr. Kenneth Monson, and State’s witness Dr. Na-garanjan Rangarajan, largely agree *848that scientific evidence available today cannot rule out the possibility of accidental death.
31. At trial, Dr. Bayardo testified that the fatal trauma sustained by Brandon Baugh was the product of a single blow to the infant’s head. This, of course, is consistent with the infant having fallen from Applicant’s arms, and reaching a speed of eleven miles per hour when the back of his head struck the concrete floor, and thereby absorbed all the energy (G force) of the fall “virtually instantaneously.” No evidence was introduced by the State that would have permitted the jury to find, much less beyond a reasonable doubt, that the infant sustained fatal trauma by reason of multiple blows to the head.
32. At the evidentiary hearing, Dr. Bay-ardo provided a detailed explanation of his conclusion that the infant died as a result of a single blow to the head. He used his autopsy photos for this purpose, as well as explaining the “soft tissue” and other analyses that he observed or performed during the autopsy. Based both upon its assessment of Dr. Bayardo’s testimony at the evidentiary hearing, and upon its personal recollection of Dr. Bayardo’s testimony at the trial of this case in 1995, and at many other criminal trials at which this Court served as trial judge, this Court finds that Dr. Bayardo’s “single blow” analysis is truthful and credible.
33. Drs. John Plunkett, Janice Ophoven, and Peter Stephens each testified at the evidentiary hearing that they agreed with Dr. Bayardo’s single blow analysis. The Court finds that all this testimony is truthful and credible.
34. The reports of Drs. Plunkett, Stephens, Ophoven, and Monson, and these witnesses’ corresponding testimony at the evidentiary hearing, support a finding that the biomechanical analysis that has been presented and used in this case was not available to Applicant in November 1998. Although biomechanical studies and analyses had been used theretofore for accident prevention research and application — helmets, airbags, seat-belts, child restraint seats, etc. — it was not until the early 2000’s that research began to focus for the first time on short distance falls and other traumatic events involving head injuries to infants and young children.
35. The Court also heard the testimony of the State’s experts, Drs. Mary Case, M.D., Carole Jenny, M.D., N. Rangarajan, Ph.D., Stephen Pustilnik, M.D., and Harrell Gill-King, Ph.D. The Court finds that their testimony is truthful and credible.
36. However, despite finding that both Applicant’s and State’s expert witnesses were truthful and credible, the Court finds that Dr. Bayardo’s reevaluation is based upon credible, new scientific evidence, and that the re-evaluation is, as the Court of Criminal Appeals stated, a “material exculpatory fact.”
37. Accordingly, the Court finds that Applicant has met her burden of proof under Sections, 5(a)(1) and 5(e) of Article 11.071, namely, that the factual basis of her claim was unavailable to her at the time she filed her prior application on November 17, 1998, and that the basis of her claim was not ascertainable through the exercise of reasonable diligence on or before that date.
Judge Wisser’s single Conclusion of Law is: “Applicant Cathy Lynn Henderson has proved by clear and convincing evidence *849that no reasonable juror would have convicted her of the capital murder of Brandon Baugh in light of the new evidence presented in her Application.” Judge Wis-ser signed his findings and conclusions on May 14, 2012. On June 13, 2012, the State filed its response:
The State has reviewed the habeas court’s findings and conclusions on the first subsequent writ application. Given that Dr. Roberto Bayardo, one of the State’s major witnesses, has changed his opinion concerning the cause and manner of death of the infant Brandon Baugh, and in light of the court’s findings that this new medical testimony would have impacted the jury’s decision, the State does not oppose the habeas court’s recommendation that the applicant’s cause be remanded for a new trial.
The State’s position is not based on the belief that the applicant is not guilty. Nor do we agree with the theory of biomechanics as presented by defense experts and relied upon by Dr. Bayardo. But we do believe that the community must have confidence in a fair process and accurate outcome. To this end, we believe Dr. Bayardo’s reevaluation and the habeas court’s recommendation are important enough to merit a reconsideration of all the evidence, including the new scientific theories, by a jury.
Our decision not to file objections to the habeas court’s findings, nor to oppose the court’s recommendation for a new trial, is done so that this matter can be fully and fairly litigated.20
Under the standard for determining a bare claim of actual innocence announced in Ex parte Elizondo, an applicant must show “by clear and convincing evidence that no reasonable juror would have convicted him in light of the new evidence.”21 This is what applicant claims; this is what the trial court concluded — that no reasonable juror would have found applicant guilty of the capital murder of Brandon Baugh — at least not to a level of confidence beyond a reasonable doubt.
Judge Wisser did not have to find that Brandon’s death was an accident to conclude that applicant was entitled to a new trial based on Dr. Bayardo’s changed testimony and the new expert testimony concerning scientific advances in biomechanics and forensic pathology. Judge Wisser’s factual finding that “if Dr. Bayardo’s reevaluation had been presented to the jury in 1995, no rational juror could have or would have convicted Applicant of capital murder beyond a reasonable doubt in light of this new evidence” is entitled to deference because it is supported by the record.22 That does not mean that applicant is actually innocent of capital murder. It simply means that the crucial evidence that had supported both the cause of Brandon’s death and applicant’s intent to cause his death has been retracted. The present guilty verdict is based on scientifically unreliable evidence, but, after another capital murder trial, a guilty verdict could be based on scientifically reliable evidence or evidence that forthrightly admits that science cannot resolve the question of either causation or intent.
Given Judge Wisser’s profound concerns about the impact of Dr. Bayardo’s expert testimony at trial on the critical, disputed issue of applicant’s intent, I agree that *850applicant did not receive a fundamentally fair trial based upon reliable scientific evidence.23 Despite every participant’s hon*851esty and good faith, this-as the District Attorney of Travis County forthrightly recognizes — is a case that should be retried to ensure the accuracy of our verdicts and the integrity of our system. With these comments, I join the Court’s order.
ALCALA, J., filed a concurring opinion.
Like the majority of the Court, I conclude that the death sentence imposed against Cathy Lynn Henderson, applicant, must be vacated and that she must receive a new trial for the charge of capital murder of Brandon Baugh. In reaching this conclusion, I follow the recommendation of the trial court and the State’s attorney and join the Court’s majority opinion granting relief. I also join the concurring opinion by the Honorable Judge Cochran with two exceptions. I disagree with her conclusions that (1) this case is the same as Ex parte Robbins, 360 S.W.3d 446 (Tex.Crim. App.2011), and (2) this case permits us to decide whether there is a due-process violation outside the context of a death-penalty case. I write separately to explain why I conclude that this case presents more compelling reasons for granting relief than those presented in Robbins. See id.
In Robbins, this Court denied the applicant relief. Id. at 448. If this case was factually identical to Robbins, the same precedent that was used to deny relief in Robbins would compel denying relief in this case. See id. Instead, the Court grants relief in this case. I conclude that, although they share many factual similarities, Robbins and this case differ as to the findings of fact rendered by the respective trial courts: This trial court finds that new scientific evidence is the basis for ordering a new trial, whereas the Robbins trial court found that use of false evidence was the basis for ordering a new trial. Id. at 457.
The Robbins trial court’s findings stated that medical examiner “Dr. Moore’s trial opinions were not true. They were based on false pretenses of competence, objectivity, and underlying pathological reasoning, and were not given in good faith.” Id. at 477 (Alcala, J., dissenting). The trial court characterized Dr. Moore’s testimony as “expert fiction calculated to attain a criminal conviction.” Id. Furthermore, the trial court found that Dr. Moore was “biased toward the State” at the time she testified. Id. at 474 (Cochran, J., dissenting). In my dissenting opinion in Robbins, I concluded that the record supported the trial court’s characterization concerning the falseness of the testimony and that the use of that testimony violated the Due Process Clause of the Fourteenth Amendment. Id. at 476-77 (Alcala, J., dissenting); see also Ex parte Chabot, 300 S.W.3d 768, 770-71 (Tex. Crim.App.2009); Ex parte Napper, 322 S.W.3d 202, 242 (Tex.Crim.App.2010).
Here, the trial court has not made any factual findings to suggest that, at the time that it was introduced, the medical evidence underlying applicant’s conviction was known to have been false. More specifically, nothing in the trial court’s findings suggests that Dr. Bayardo based his testimony on false pretenses of competence, a lack of objectivity, prosecutorial bias, or expert fiction calculated to attain a criminal conviction. The absence of these types of findings distinguishes this case from Robbins and renders it a new-science case rather than a false-testimony case. Compare id. at 457. As Judge Cochran accurately observes in her concurring opinion today, Dr. Bayardo’s testimony *852was “based upon the state of the scientific knowledge” and was not known to have been false at the time it was given. For this reason, I join Judge Cochran’s opinion today, although I did not in Robbins. See id. at 476 n. 1 (Alcala, J., dissenting).
In Robbins, I explained that I did not join Judge Cochran’s dissenting opinion “because the change in Dr. Moore’s testimony is not due to new scientific principles but is instead, according to her, due to her having more experience as a medical examiner, and according to the trial court’s findings, due to her trial testimony being the result of prosecutorial bias.” Id. Today, I join Judge Cochran’s concurring opinion because this case falls squarely within her assertion that executing a defendant whose conviction is premised on now-discredited scientific theories violates due process, even though those scientific theories were once considered valid and true at the time they were applied.
Furthermore, although I disagree with the Honorable Judge Price’s analysis of Robbins, I agree with his conclusion that this case presents a stronger reason to grant relief than that presented in Robbins: Without relief, applicant will be executed for a conviction that we now know was premised largely on faulty science.
The Supreme Court has succinctly observed that “the penalty of death is qualitatively different from any other sentence.” Lockett v. Ohio, 438 U.S. 586, 604, 98 S.Ct. 2954, 57 L.Ed.2d 973 (1978) (internal quotations omitted). Among these differences is that a death sentence “is unique in its total irrevocability.” Furman v. Georgia, 408 U.S. 238, 306, 92 S.Ct. 2726, 33 L.Ed.2d 346 (1972) (Stewart, J., concurring). The Court has held that the “qualitative difference between death and other penalties calls for a greater degree of reliability when the death sentence is imposed.” Lockett, 438 U.S. at 604, 98 S.Ct. 2954. This heightened need for reliability requires a mechanism that enables judicial enforcement of that sentence to evolve with the science that serves as the basis for imposition of that sentence.
Whether we ultimately apply the faulty-science theory to due-process complaints beyond the death-penalty context is a question for another day. The holding of this case is quite narrow: Due process prohibits the execution of a person when faulty science was essential to the State’s establishment of an element necessary for conviction — here, that the cause of death of the complainant was intentional — and the habeas record shows that today’s scientific community reaches a different consensus — here, that the cause of death is undetermined.
In accordance with the trial court’s recommendation, I join in the Court’s judgment granting relief and remanding for a new trial.
KEASLER, J., filed a dissenting opinion, in which KELLER, P.J., and HERVEY, J., joined.
Like this Court’s order remanding the matter to the trial court for findings of fact and conclusions of law, the Court’s opinion today grants Cathy Henderson relief without one word of analysis why she is entitled to it. In fact, the Court does not even identify the legal basis for granting Henderson relief. Instead it issues a legally hollow opinion with a staggering result. Readers of both our remand order in this matter and today’s opinion will undoubtedly and justifiably be both baffled and appalled by the Court’s opinion. I count myself among them. The unmistakable message of the Court’s per curiam opinion is this: despite applicable legal precedent to the contrary and overwhelming inculpatory facts, we grant Henderson relief solely because we want to. And to future applicants, this message’s implica*853tion is clear: with luck, your writ application may also be viewed with such grace.
In her subsequent application for a writ of habeas corpus, Henderson asserts she is entitled to relief for three reasons: (1) she is actually innocent of capital murder because no reasonable juror would convict her of capital murder in light of new scientific evidence (Herrera1-type claim); (2) but for constitutional errors — namely a violation of Ake v. Oklahoma2 and a Fifth Amendment claim previously raised and rejected on direct appeal3 — no rational juror could have found her guilty beyond a reasonable doubt in light of the new evidence (Schlup4-type claim); and (3) she is no longer death eligible. The Court came to the breathtaking conclusion that Henderson satisfied Texas Code of Criminal Procedure article 11.07, § 5(a) with her allegation that then Travis County Medical Examiner Dr. Roberto Bayardo’s reevaluation of his opinion at trial was newly discovered evidence that established her innocence. The Court remanded the matter to the trial court for further proceedings on Henderson’s first two claims and dismissed her third.5
After several evidentiary hearings, this matter now returns to us with the trial judge’s findings of fact and conclusions of law recommending that we grant relief on Henderson’s first actual-innocence claim. Dr. Bayardo’s reevaluation of whether the injuries suffered by Brandon Baugh, the three-and-half-month-old victim of this capital murder, were intentionally inflicted is the crux of Henderson’s actual-innocence claim and the foundation of the trial judge’s recommendation to grant relief. Despite our instructions in the remand order, the trial judge did not enter findings of fact or conclusions of law on Henderson’s Sehlup claim, which if the Court were to expressly reject Henderson’s actual-innocence claim, as it should, would require remanding to the trial judge to address this issue.
By explicitly stating that the trial judge’s findings of fact are supported by the record, the Court, by implication, reaches the opposite conclusion as to the trial judge’s conclusions of law — that they do not share the same record support. But it nonetheless “accepts” the trial judge’s recommendation to grant relief and give Henderson a new trial on an unknown basis. Surely, it cannot be actual innocence otherwise the Court would have found the trial judge’s conclusions are supported by the record or would have expressly found Henderson proved her actual innocence. In its zeal to grant Henderson relief, the Court is forced to look elsewhere to accomplish its goal, and left without a clear legal path, the Court takes the indefensible position to grant relief without justification or explanation. The facts adduced at trial and in the subsequent evidentiary hearings in light of our actual-innocence case law make it clear why the majority could not adopt the trial judge’s conclusions and grant Henderson relief on actual-innocence grounds; the burden is too high, the inculpatory facts are too great, and the “new evidence” is too weak.
To put the issue of Dr. Bayardo’s reevaluation in perspective, it is appropriate to *854start with his testimony at Henderson’s 1995 capital murder trial. At the time, Dr. Bayardo had been Travis County’s Chief Medical Examiner for eighteen years and throughout his career he had performed approximately 15,000 autopsies. He personally performed Brandon Baugh’s autopsy and testified to the extent of Brandon’s injuries at trial. Dr. Bayardo concluded, “It is my opinion, based upon the autopsy findings, that the decedent, Brandon Baugh, came to his death as a result of a severe closed head injury. There was comminuted fracturing of the back of the skull and subdural and subarachnoid hemorrhages implying that a severe force had been given to the head and characteristic of abuse, homicide.”6 Dr. Bayardo also concluded that it would have been “impossible” and “incredible” for a fall from four to four-and-a-half feet to have caused Brandon’s injuries. He also testified that Brandon’s injuries were not accidental because the fractures crossed the suture lines found in an infant’s not-fully formed skull which would require a severe degree of force. Further, the injury’s location — the back of the head — was a characteristic of an abused child because most accidents occur on the sides of the head. Dr. Bayardo explained that “[tjhis is an injury that you see when a baby’s head is slammed or thrown very forceable against a flat surface” and in order for Brandon’s injury to result from a fall, “he would have to fall from a height higher than a two-story building.”7 On cross-examination, Dr. Bayardo conceded that his testimony was limited to the cause of death and he was unable to tell the jury the exact nature of the severe force or how that force was inflicted.
At the habeas hearing and in his affidavit, Dr. Bayardo testified that after reviewing reports from Drs. John Plunkett, Peter Stephens, and Kenneth Monsoon— Henderson’s proffered experts who discussed how biomechanics could explain that Brandon’s injuries were accidental— he would no longer testify that the manner of Brandon’s death was homicide; instead he would now conclude that it was “undetermined,” as opposed to accidental. According to his affidavit, Dr. Bayardo also claims that he would not be able to testify about the degree of force needed to cause Brandon’s head injury and would not conclude that an accidental fall was “impossible” or “incredible” in explaining the cause of Brandon’s injuries.
Even after the benefit of multiple evi-dentiary hearings where the trial judge took testimony from expert after expert, Henderson is no closer to establishing her innocence than she was in her claims she asserted in her subsequent application.8 Henderson presents a bare innocence claim. We have labeled the burden of establishing a bare claim of actual innocence a Herculean task.9 In satisfying this heavy burden, Henderson’s newly discovered evidence must constitute affirmative evidence of her innocence.10 Henderson must show by clear and convincing evidence that no reasonable juror would have convicted her in light of the new evidence.11 Whether an applicant satisfies this burden requires the evaluation of “the *855probable impact of the newly available evidence upon the persuasiveness of the State’s case as a whole, so we must necessarily weigh such exculpatory evidence against the evidence of guilt adduced at trial.”12
The trial judge concluded that Henderson “has proved by clear and convincing evidence that no reasonable juror would have convicted her of the capital murder of Brandon Baugh in light of the new evidence presented in her Application.” 13 The Court’s opinion acknowledges that in most circumstances we appropriately defer to and accept the trial judge’s findings of fact and conclusions of law when they are supported by the record. However, we may make contrary findings and conclusion when our independent review of the record reveals the findings and conclusions are not supported by the record.14 Here, the trial judge’s conclusion is not supported by the record and is demonstrably wrong because it implicitly miseharacterizes Dr. Bayardo’s new opinion, improperly focuses on Dr. Bayar-do’s reevaluation in isolation, and fails to weigh such exculpatory evidence against all of the evidence of guilt adduced at trial.
Dr. Bayardo’s testimony was merely one piece of evidence that established Henderson’s guilt at trial and shed light on her intent that fateful day. Henderson was the last one to have seen Brandon the day he went missing after he was dropped off at Henderson’s home where she eared for Brandon and Megan Baugh. There were no calls to 911 that day from Henderson’s home or from her neighborhood. There were no other pleas for help. She had the Baughs’ emergency contact numbers. But they were not called. After Brandon died, she wrapped his body in a blanket and put his body in a Bartles & Jaymes wine cooler box and secured it with tape later matched up to tape found in her home. With Brandon’s body in the trunk of her car, Henderson put her daughter Jennifer and Megan, Brandon’s older sister, in the car with a spade and her neatly packed suitcase and drove to Round Rock to get her car’s oil changed.
On her way out of town and with the two girls in tow, she drove to the bank to withdraw $2200 through a cash advance on her credit card. After repeated attempts proved unsuccessful, she was finally able to withdraw $1000. Henderson told the bank employee that she needed the money because her father just died and she needed to be with her family. At some point in the afternoon, she stopped at McDonald’s to get the girls something to eat. She then drove to Holland, Texas where her husband’s relatives lived. She arrived unexpectedly. They had not seen Henderson in five or six months. Once there, she told her husband’s relatives that she needed to go to the store and would be back soon. She asked her eleven-year-old niece to babysit Jennifer and Megan. She never returned. Instead, she continued north where outside Waco, just off a country road near a stand of trees, she buried Brandon in the box, using the spade she packed. Investigators would later find Brandon’s diaper bag in a ditch in the vicinity where they found Brandon’s buried body.
Later that night she checked into a motel in Blackwell, Oklahoma under the name Tracy Simms and listed a Missouri ad*856dress. The next day, Henderson arrived in Trenton, Missouri and dropped in on a longtime friend Linda Brewer. It too was an unexpected visit. The two discussed Henderson’s troubling custody issues and how she was ordered to have only supervised visitation with her other daughter. Henderson told Brewer that the children she watched were picked up by their grandparents who would take care of them for a week so she decided to visit her in Missouri. Only three days after Brandon was killed and while having a few margaritas with a friend, Henderson admitted that she had “killed somebody or murdered somebody.”15
Henderson and Brewer then drove to Independence, Missouri to see other friends. After her arrival in Independence, Henderson told Brewer that she had a new identity, was getting new license plates for her car, and wondered how she would look with red hair. Brewer began to realize Henderson’s trip was no longer just an opportunity to come back and visit. When asked why she needed to change her identity, she responded, “I can’t do life. I don’t want to talk about it anymore.” With the use of a Social Security card another friend found, she assumed the identity of Patricia Keith. Reluctant to drive her vehicle because she was afraid the police would be looking for it, she got some old Missouri license plates and put them on her car. She dyed her hair red. With her new identity, she rented an apartment under her assumed name. She also began a sexual relationship with the male friend who gave her the Social Security card and new license plates and wanted him to move in with her. She attempted to find a job in Independence. She did all of this within three or four days after Brandon’s death.
When a police officer knocked at the door of her apartment looking for information on Henderson and Brandon, she claimed that she had never seen the person (Henderson) in the pictures. At first, the officer did not recognize her from the pictures. But the officer returned, and Henderson gave a false name. After Henderson consented to a search of the apartment, the officer discovered the Social Security card Patricia Keith’s name along with the receipt for the oil change and apartment rental receipt. She was subsequently arrested and claimed that she did not know what this was about, threatened the officers with a false imprisonment suit, and complained that she was going to miss a hair appointment.
She was then interviewed by an FBI agent to whom she gave conflicting stories. Her first story was that she did not have any information about Brandon. In her second story, she claimed that the Baughs’ grandmother came to pick Brandon up. She then claimed she packed up the girls and brought them to the bank, McDonald’s, and the relatives’ home in Holland and drove to Blackwell, Oklahoma, checking into the motel under Tracy Simms and finally arriving in Trenton and Independence, Missouri. But when the agent suggested that Brandon was dead, and perhaps it was an accident, Henderson said “yes.” When asked “Did you bury him?” she responded “Of course, I did. He’s just a baby.” Henderson’s final version was that around 10:30 in the morning, Brandon fell from her arms and hit the tile floor. She stated she attempted CPR for about an hour, but she knew he was dead. She admitted to burying Brandon near Waco with a spade she brought from home.
Dr. Bayardo was not the only witness who gave expert testimony concerning the manner of Brandon’s death. The jury heard from Lubbock County Deputy Chief *857Medical Examiner Dr. Sparks Veasey III. Like Dr. Bayardo, Dr. Veasey concluded that Brandon’s injuries were “consistent with a baby’s head [being] slammed into a blunt object, a baby being held by the legs and slammed into a wall or a floor. They are consistent with a baby being forcefully — extremely forcefully thrown into a blunt — into a blunt object.”16 He further concluded that Brandon’s injuries were inconsistent with an accidental drop from a distance of four to five feet and was certain that Brandon’s injuries were not the result of an accident. In addition to Dr. Veasey, Henderson’s own expert corroborated Dr. Bayardo’s and Dr. Veasey’s conclusions. After reviewing the autopsy report, photographs, and videotape, Dr. Kris Sperry, the Fulton County Deputy Chief Medical Examiner in Atlanta, Georgia, opined that Brandon’s crushing skull fractures were not accidental.
The trial judge’s conclusion that Henderson proved by clear and convincing evidence that no reasonable juror would have convicted her of capital murder simply failed to weigh Dr. Bayardo’s reevaluation against the evidence of guilt adduced at trial. The trial judge came to the remarkable and unsupported mixed finding and conclusion that
Because Dr. Bayardo’s testimony at trial was the critical evidence upon which the conviction of Applicant rested, and was the testimony upon which the essential element of culpable mental state hinged, the Court finds that if Dr. Bayardo’s reevaluation had been presented to the jury in 1995, no rational juror could have or would have convicted Applicant of capital murder beyond a reasonable doubt in light of this new evidence.17
Contrary to Judge Cochran’s belief, the trial judge’s opinion that “no rational juror could have or would have convicted Applicant of capital murder beyond a reasonable doubt in light of this evidence” is a conclusion of law, and therefore not entitled to deference.
’ The trial judge claimed to be mindful of Henderson’s flight, only to “find[ ] that the evidence of flight did not have the capacity to prove the mens rea of capital murder beyond a reasonable doubt.” But “[w]e have repeatedly held that flight is evidence of a circumstance from which an inference of guilt may be drawn.”18 Notably, the trial judge did not hear from Dr. Veasey or Dr. Sperry in connection with Henderson’s application and Henderson does not directly challenge their conclusions. We do not know if their conclusions would change like Dr. Bayardo’s when presented with Henderson’s experts’ affidavits and reports. On this record, we cannot assume they would. However, the trial judge cavalierly found that “if the jurors had heard Dr. Bayardo’s re-evaluation, they would not have, credited the then-conflicting testimony of Dr. Veasey.”19 Such idle speculation is certainly not a finding that this Court should adopt. Even if true, what about Dr. Sperry’s testimony offered by Henderson herself? The findings and conclusions are silent about the potential impact of Dr. Sperry’s corroborative testimony of both Dr. Veas-ey’s and Dr. Bayardo’s opinions at trial.
*858More importantly, it is the trial judge’s failure to consider all of Henderson’s actions after Brandon’s death that is the most troubling. After Brandon’s death, there was no attempt to call for help. She attempted to hide the evidence of Brandon’s death by burying him near Waco with the spade she brought from home. While having drinks with friends she admitted to killing or murdering someone. Further, Henderson did not merely engage in a very deliberate plan to flee. She actively attempted to evade law enforcement through assuming a new identity, changing her appearance, changing her li-cence plates, and starting a new life in Missouri while leaving her family behind. And when the jig was up, she first claimed not to know what happened to Brandon before finally admitting to burying him and changing her version of events to describe an accidental death. These are not the acts of an innocent person. In the face of common sense and our case law, it would be preposterous to conclude that Dr. Bayardo’s new opinion that the manner of death should be undetermined, as opposed to accidental, would undermine all of the incriminating evidence establishing Henderson’s guilt.
By leaving its rationale unstated, the Court avoids confronting our holding in Ex parte Robbins,20 our recent opinion addressing a factually similar actual-innocence claim and precedent determinative of Henderson’s actual-innocence claim. In Ex parte Robbins, Dr. Patricia Moore, an assistant medical examiner, testified at Robbins’s capital murder trial that the child victim’s death was caused by “asphyxia due to compression of the chest and abdomen and that the manner of death was homicide.”21 Many years after the jury found Robbins guilty of capital murder, Dr. Moore’s report and conclusions were reexamined by several other medical examiners who disagreed with her conclusions. Dr. Moore herself also reevaluated her report and came to the conclusion that her opinion had changed and the cause and the manner of the victim’s death should be listed as undetermined.22 Like Henderson, Robbins claimed this reevaluation was newly discovered evidence that demonstrated his actual innocence.23 In denying Robbin’s actual-innocence claim, we held that
[Robbins] failed to prove that the new evidence unquestionably establishes his innocence. Moore can no longer stand by her trial testimony, but rather than completely retracting her trial opinion, she is of the current opinion that the cause and manner of death of [the victim’s] death are ‘undetermined.’ Moore cannot rule out her trial opinion as a possibility of how [the victim] died. Hence, Moore’s reevaluation falls short of the requisite showing for actual innocence because it does not affirmatively disprove that [Robbins] intentionally asphyxiated [the victim].24
Like Dr. Moore’s reevaluation in Ex parte Robbins, Dr. Bayardo’s reevaluation did not render void his trial testimony.25 The jury could have considered Dr. Bayar-do’s testimony that the manner of death was “undetermined” and still have found Henderson guilty based on all of the evidence presented at her trial, including Dr. Véaseos and Dr. Sperry’s expert conclusions.26 Henderson’s reliance on Dr. Bay-*859ardo’s reevaluation merely serves to retroactively impugn the State’s case at trial and does not affirmatively demonstrate her innocence.27 On this record and particular claim of actual innocence, we are compelled to follow Ex 'parte Robbins and conclude that Henderson failed to satisfy her Herculean burden — to show by clear and convincing evidence that no reasonable juror would have convicted her in light of Dr. Bayardo’s reevaluation.
In its response, the State does not contest the trial judge’s findings and conclusions, and does not oppose granting Henderson relief. The State takes this position while simultaneously making it clear that it does not believe that Henderson is not guilty and disagrees with the biomechanical theory presented by Henderson’s experts and relied upon by Dr. Bayardo in his reevaluation. The State’s concern for the community’s confidence in the criminal justice system expressed in its response is laudable. But it is neither legally controlling, nor a particularly persuasive argument for granting Henderson relief when the State still contests Henderson’s underlying factual contentions. The State’s acquiescence cannot not bridge the gulf between Henderson’s asserted claims and the burden she must satisfy to be legally entitled to relief.
Today, the Court’s decision casts aside its established legal principles and grants relief to an applicant not entitled to it. The Court accomplishes this feat by abandoning all standards necessary for an applicant to obtain relief, only to replace them with an unexplained, ad hoc determination. “Because we want to” is not a substitute for legal reasoning. And any suggestion to the contrary is untenable. Further, we have denied past applicants similarly requested relief on similar evidence. We owe a duty to all applicants that we will measure the merits of their claims equally.
Henderson’s Herrera-type claim of actual innocence should be denied and her pending Schlup-type actual-innocence claim should be remanded to the trial court for findings of fact and conclusions of law. It is a travesty to grant this child killer relief on some unknown legal principle while her tiny, defenseless victim lies dead and reburied. Therefore I dissent with all the vigor at my command.
HERVEY, J., filed a dissenting opinion in which KELLER, P.J., and KEASLER, J., joined.
Something is missing here. While the Court states that it accepts the trial court’s recommendation granting relief, it does so without providing any legal basis for that ruling, and I cannot find a ground upon which relief should be granted. And to justify its decision, the Court makes a quantum leap from “advances in science” to granting relief, which presents a whole new dilemma for the criminal justice system and this case in particular.
The real issue in this case is whether the admission of potentially unreliable evidence requires this Court to grant relief regardless of the state of the remaining record. Eleven expert witnesses testified at Applicant’s writ hearing, some for both sides. The trial court found all of the experts credible but focused on a change in the testimony of expert Dr. Bayardo to conclude that Applicant had proven by clear and convincing evidence that no reasonable juror would have convicted her in light of the “new evidence.” This Court now defers to that conclusion. However, nowhere in Dr. Bayardo’s altered testimony does he state or indicate that his original opinion was false, nor does he refute *860the medical science relied upon by the other State experts. Instead, Dr. Bayardo changed his opinion on the manner of death from “homicide” to “undetermined” based upon changes in the science upon which he relied. While a change in Dr. Bayardo’s testimony could render his opinion unreliable, “unreliable” testimony does not equate with “false testimony” or “innocence,” nor does it automatically require a new trial.
The admission of expert testimony is governed by Texas Rule of Evidence 702, and to be admissible under this rule, the party offering the scientific expert testimony must demonstrate by clear and convincing evidence that such testimony is both relevant and reliable. Kelly v. State, 824 S.W.2d 568, 572 (Tex.Crim.App.1992). The focus of the reliability analysis is to determine whether the evidence has its basis in sound scientific methodology such that testimony about “junk science” is weeded out. Jordan v. State, 928 S.W.2d 550, 555 (Tex.Crim.App.1996).
Whether the science at issue is a “hard” science1 or a “soft” science,2 “reliability should be evaluated by reference to the standards applicable to the particular professional field in question.” Coble v. State, 330 S.W.3d 253, 274 (Tex.Crim.App.2010); see Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993) (holding that when the subject of the expert’s testimony is scientific knowledge, the basis of his or her testimony must be grounded in the accepted methods and procedures of science). Therefore, a change in the science upon which an expert relied in providing his trial testimony might indeed undermine the reliability of his testimony. But it does not necessarily follow that the testimony was “false” or that the existence of new science necessarily implicates innocence. For example, in this Court’s unanimous opinion in Ex parte Miles, 359 S.W.3d 647 (Tex.Crim.App.2012), we adopted the trial court’s conclusion that the “gunshot-residue standards, as testified to at trial, are no longer reliable.” Id. at 663. However, we did not adopt the trial court’s conclusion that “Applicant should also be granted relief, independently, on the ground of flawed forensic testimony.” Id. Instead, we looked to all of the evidence presented and only then determined that actual-innocence relief was warranted. Other decisions from this Court recognize both the advancement in science and the legislative directive to apply that science to our caselaw through Chapter 64 of the Texas Code of Criminal Procedure. While further testing may prove to be inconclusive or even exculpatory, relief may nonetheless be denied based on the volume of other evidence. See Gutierrez v. State, 337 S.W.3d 883 (Tex.Crim.App.2011); Prible v. State, 245 S.W.3d 466 (Tex.Crim.App.2008).
Furthermore, even if a change in the underlying science means that the expert testimony was unreliable, it does not automatically result in a due process violation (and thus a new trial). An additional analytical step is required. Only when the admission of unreliable testimony was harmful is due process implicated and a new trial appropriate. See Coble, 330 S.W.3d at 280. Accordingly, relief should only be granted if the applicant demonstrates the error affected his or her substantial right to a fair trial. See Tex. R.App. Proc. 44.2(b); Coble, 330 S.W.3d at 280 (explaining that harm occurs when the error had a substantial and injurious effect or influence in determining the jury’s verdict). A criminal conviction should not be overturned by the erroneous admission of *861evidence “if the appellate court, after examining the record as a whole, has fair assurance that the error did not influence the jury, or had but a slight effect.” Johnson v. State, 967 S.W.2d 410, 417 (Tex.Crim.App.1998).
This reliability analysis is consistent with our policy interest in the finality of convictions. The Supreme Court has emphasized its enduring respect for “the State’s interest in the finality of convictions that have survived direct review •within the state court system.” Calderon v. Thompson, 528 U.S. 538, 555, 118 S.Ct. 1489, 140 L.Ed.2d 728 (1998). “Without finality, the criminal law is deprived of much of its deterrent effect.” Teague v. Lane, 489 U.S. 288, 309, 109 S.Ct. 1060, 103 L.Ed.2d 334 (1989) (plurality op.). To abide by Judge Cochran’s suggestion that any intervening scientific development should result in a new trial would seriously undermine the stability of our criminal justice system. Most convictions involve some type of scientific evidence, whether hard (e.g., DNA or urinalysis) or soft (e.g., eyewitness identification or forensic psychiatrists testifying about future dangerousness). Rarely is a case wholly dependent on science alone. Thus, if we were to grant a new trial with every scientific advancement, without proof that the original science was indeed faulty, the finality of convictions would be illusory. While the evolution of science is important to the improvement of our system, each case must be decided individually, taking into account all of the evidence.
In summary, intervening scientific developments might result in unreliable expert testimony, and the admission of this unreliable evidence might rise to the level of a due process violation. But this case does not present us with such a scenario. Dr. Bayardo changed his opinion on the manner of death from “homicide” to “undetermined” based upon changes in the science upon which he relied. Perhaps this makes Dr. Bayardo’s trial testimony unreliable, but as Judge Cochran acknowledges in her concurring opinion, this “does not mean applicant is actually innocent of homicide” or that “[Dr. Bayardo’s] trial testimony was ‘false’ at the time it was given, based upon the state of scientific knowledge that he relied upon at that time.” The “new evidence,” even if based on “new science,” must still affirmatively establish that the applicant is entitled to relief. See Ex parte Spencer, 337 S.W.3d 869, 879 (Tex.Crim.App.2011).
Additionally, when the entire record is considered, it is clear that Appellant’s substantial rights to a fair trial were not affected by the admission of Dr. Bayardo’s testimony. See Coble, 330 S.W.3d at 280 (“In making a harm analysis, we examine the entire trial record and calculate, as much as possible, the probably impact of the error upon the rest of the evidence.”). As Judge Keasler competently outlines in his dissenting opinion, there is ample evidence to support that Appellant intentionally caused Brandon’s death. See Tex. Penal Code § 19.03(a)(8). Certainly, in this context, Dr. Bayardo’s unreliable testimony could not have a substantial and injurious effect or influence in determining the jury’s verdict. See Coble, 330 S.W.3d at 280. The record provides fair assurance that the error would not influence a jury, or would have but a slight effect. See Johnson, 967 S.W.2d at 417.
For these reasons, I respectfully dissent.
. Ex parte Henderson, 246 S.W.3d 690, 693 (Tex.Crim.App.2007) (Price, J., concurring).
. See Ex parte Brown, 205 S.W.3d 538, 545 (Tex.Crim.App.2006) ("Establishing a bare claim of actual innocence [under the standard set out in Ex parte Elizondo, 947 S.W.2d 202 (Tex.Crim.App.1996)] is a Herculean task.”).
.360 S.W.3d 446, 463-68 (Tex.Crim.App.2011) (Price, J., concurring).
. Id. at 464 (Price, J., concurring).
. United States v. Bagley, 473 U.S. 667, 678-80 & n. 9, 105 S.Ct. 3375, 87 L.Ed.2d 481 (1985).
. Robbins, at 463 (Price, J., concurring).
. Elizondo, supra, at 209.
. Ex parte Chavez, 371 S.W.3d 200, 216 (Tex. Crim.App.2012) (Keller, P.J., dissenting). The majority in Chavez seems to have applied the same standard of materiality for the unknowing use of false or perjured evidence as for the knowing use. See Chavez, supra, at 208-09 (applying "reasonable likelihood” standard of materiality). Needless to say, I am not in favor of that. Robbins, supra, passim (Price, J., concurring).
. Robbins, supra, at 465 (Price, J., concurring).
. Id. at 463 (“[Ujnlike Henderson, [the medical examiner’s] re-evaluation was not attributed to new scientific information but was instead based on the same autopsy findings and other information that she had at the time of trial.”).
. 300 S.W.3d 768, 771 (Tex.Crim.App.2009).
. Cf. Chavez, supra, at 205 ("Chabot was the first case in which we explicitly recognized an unknowing-use due-process claim; therefore, that legal basis was unavailable at the time applicant filed his previous application. See Tex.Code Crim. Proc. art. 11.07 § 4(a)(1).”).
. See Ex parte Napper, 322 S.W.3d 202, 243 (Tex.Crim.App.2010) ("[W]e observe that our 'actual innocence’ jurisprudence could encompass a claim on habeas that involved newly discovered evidence that a witness's testimony was false.”); Chavez, supra, at 216 n. 30 (Keller, P.J., dissenting) ("Although Eli-zondo was decided as an actual innocence case, it would also have satisfied the predicate, under today's standards, for an unknowing-use-of-perjured-testimony claim because the recantation of a testifying witness was at issue.”).
.Chavez, supra, at 220-21 (Price, J., dissenting).
. 360 S.W.3d 446 (Tex.Crim.App.2011).
. For a thorough discussion of the jurisprudential issues raised by the problem of "shifting science" and the reliability of criminal verdicts over time, see Caitlin Plummer & Imran Syed, “Shifted Science" and Post-Conviction Relief, 8 Stan. J.C.R. & C.L. 259 (2012).
. See “Shifted Science, supra note 2 at 263-64 (noting the problem of "what happens when scientific testimony that led to a conviction is later proved to have been completely invalid — a phenomenon we call 'shifted science.’ Science — its very character defined as the search for what lies beyond the present horizon of human understanding — advances, changes, and evolves constantly. Unfortunately, courtroom procedures, practices, and the institutional knowledge of legal actors are just the opposite: the law takes pride in its consistency and glacial pace of adaptation. If science is a leading indicator of where society is headed, the law is a lagging indicator of truths society recognized years before, and may since have moved above and beyond. When the two come together in the case of scientific evidence being used in courtrooms, it should be no surprise that the results are often less than palatable.”).
.These background facts are taken, largely verbatim, from our opinion on direct appeal. Henderson v. State, 962 S.W.2d 544, 548-49 (Tex.Crim.App.1997).
. Agent Napier testified that
[I]n telling the story, one time she would say that [Brandon] fell out of her arms, that he dropped out of her arms, and the final version is that while she's holding him with one hand ... as she’s trying to turn the recorder off with the other one so she can answer the phone, that he pushes his feet against the wall, and that motion then causes him to flip out of her hands and on to the floor.
. The trial judge noted in his factual findings that applicant’s flight and conduct after Brandon’s death was evidence of her guilty conscience that supported a finding of some culpable mental state, but that "equivocal” evidence was not sufficient to establish, beyond a reasonable doubt, an intent to cause Brandon's death. See Finding of Fact Number 23, quoted on page 19.
. Dr. Veasey did not dispute Dr. Bayardo’s opinion that "this was the result of single massive blow,” but he had a "suspicion ... that this may have been more than one impact causing these fractures[.]” The defense expert, Dr. Kris Sperry, testified that, although Brandon’s death was not an accident, it could have been the result of an impulsive, reckless act rather than a volitional, intentional or knowing one. The jury was charged on capital murder, intentional injury to a child, and reckless injury to a child.
. Henderson v. State, 962 S.W.2d 544 (Tex. Crim.App.1997).
. Ex parte Henderson, No. WR-49,984-01 (Tex.Crim.App. March 6, 2002) (not designated for publication).
. Henderson v. Dretke, No. A-02-CA-758-SS, 2004 WL 5295477 (W.D.Tex. March 31, 2004) (not designated for publication).
. Henderson v. Quarterman, 460 F.3d 654 (5th Cir.2006).
. Dr. Plunkett’s affidavit includes the following:
Neither I ..., nor anyone else, can prove Brandon’s injury and death was an accident. However, because of the new scientific information and analysis now available to scientifically evaluate Brandon’s injury and death, neither may anyone prove that Ms. Henderson intentionally caused it. It is impossible for any qualified scientist or physician to conclude, whether to a reasonable degree of medical certainty, or beyond a reasonable doubt, that any intentional and deliberate act by Ms. Henderson caused Brandon Baugh’s death, or that [his] injuries are such as to rule out an accidental cause.
. At trial Dr. Bayardo had said just the opposite.
Q. Okay. Can you describe for the jury what is the significance of the location of the injury in this case?
A. Yes. Also the location is quite characteristic of an abused child because most of the accidental injuries are going to occur on the sides of the head in the parietal bone. So whenever we see a fracture that includes the occipital bone, we immediately think of abuse.
. CRABI stands for "Child Restraint Air Bag Interaction.”
. Dr. Bayardo testified similarly at trial: "By just throwing the baby or just the baby falling from somebody's arms, the injury would have linear, a single short fracture on the side of the head or maybe in the front, but not on the back or the occipital bone.”
. I agree with Judge Price that
while the applicant’s flight undoubtedly evinces a guilty conscience, it provides little rational basis to conclude she felt guilty of an intentional or knowing murder, as opposed to a reckless or negligent homicide or even an excusable accident.
Ex parte Henderson, 246 S.W.3d 690, 694-95 (Tex.Crim.App.2007) (Price, J., concurring).
. As one law professor — addressing the tension between the governing framework for collateral relief and the issues presented by the "Shaken Baby Syndrome” cases — put it:
Until scientific consensus has been achieved, the criminal justice system must find its own solutions to the problem of a diagnosis already morphed and still in transition.
To date, our system has failed. In place of adaptation, we have seen massive institutional inertia. Once the SBS prosecution paradigm became entrenched, the crime became reified. Deferential review standards and a quest for finality perpetuated die system’s course. How expeditiously, and how deliberately, this course is righted will inform the meaning of justice.
Complicating the endeavor, SBS prosecutions raise discomfiting possibilities that diverge from those presented by the innocence archetype. Here, no other perpetrator can be held accountable; indeed, no crime at all may have occurred. The problem is not individual, but systemic, and its source is error, not corruption. Responsibility is diffuse: prosecutors and scientists may each legitimately point fingers. Most fundamentally, scientific developments have cast new doubt without yet creating certainty in its place. The story of SBS thus challenges current notions of wrongful convictions. Underlying conceptual frameworks must evolve accordingly.
For now, we find ourselves situated in an extraordinary moment; one which tests our commitment to innocence that is not proven, but presumed.
Deborah Tuerkheimer, The Next Innocence Project: Shaken Baby Syndrome and the Criminal Courts, 87 Wash. U.L.Rev. 1, 58 (2009).
. In Han Tak Lee v. Glunt, the Third Circuit intimated that advances in forensic science may support a due-process claim. In sending an arson case back to district court for an evidentiary hearing, the Han Tak Lee court stated, "If Lee's expert’s independent analysis of the fire scene evidence — applying principles from new developments in fire science— shows that the fire expert testimony at Lee’s trial was fundamentally unreliable, then Lee will be entitled to federal habeas relief on his due process claim.” Han Tak Lee v. Glunt, 667 F.3d 397, 407-08 (3d Cir.2012).
. Ex parte Robbins, 360 S.W.3d 446, 473-74 (Tex.Crim.App.2011) (Cochran, J., dissenting).
.The Travis County District Attorney’s Office should be commended for its fidelity to the admonition that "[i]t shall be the primary duty of all prosecuting attorneys ... not to convict, but to see that justice is done.” Tex. Code Crim. Proc. art. 2.01.
. 947 S.W.2d 202, 209 (Tex.Crim.App.1996).
. FF & CL 22.
. See, e.g., State v. Edmunds, 308 Wis.2d 374, 746 N.W.2d 590, 592 (Wis.Ct.App.2008) (defendant, infant’s babysitter, was originally convicted of first degree reckless homicide based on evidence of "shaken baby syndrome,” but was entitled to new trial based upon her claim of newly discovered evidence in the form of advances in medical science that cast serious doubt upon the cause of infant’s death).
In Edmunds the court noted that the State had, in the original trial of the case, called numerous expert witnesses who testified, to a reasonable degree of medical certainty, that the cause of the infant’s death was violent shaking or violent shaking combined with an impact that caused a fatal head injury. Id. The defense’s own expert agreed with that cause of death, but suggested that the fatal injury could have occurred before the infant was brought to the defendant-babysitter’s home. Id. Ten years after the original trial, the defendant filed a subsequent motion for new trial alleging that there had been significant developments in the medical community concerning whether the infant’s symptoms showed either "shaken baby syndrome” or shaking combined with head trauma. Id. at 593. "The experts explained that there was not a significant debate about this issue in the mid-1990s and that the opinions offered in Edmunds’s first postconviction motion would have been considered minority or fringe medical opinions.” Id. The State presented four experts who disagreed with the defense experts and maintained that the evidence at trial established that the infant had been violently injured while in the defendant’s care. Id. The trial judge agreed that the defense had presented newly discovered evidence, but denied relief because he concluded that Edmunds had not established that there was a reasonable probability of a different result with the new expert evidence. Id. at 594. The court of appeals reversed and remanded the case for a new trial based upon the new scientific advances.
The court of appeals explained,
The newly discovered evidence in this case shows that there has been a shift in mainstream medical opinion since the time of Edmunds’s trial as to the causes of the type of trauma [the infant] exhibited. We recognize, as did the circuit court, that there are now competing medical opinions as to how [the infant's] injuries arose and that the new evidence does not completely dispel the old evidence. Indeed, the debate between the defense and State experts reveals a fierce disagreement between forensic pathologists, who now question whether the symptoms [the infant] displayed indicate intentional head trauma, and pediatricians, who largely adhere to the science as presented at Edmunds’s trial. However, it is the emergence of a legitimate and significant dispute within the medical community as to the cause of those injuries that constitutes newly discovered evidence.... Now, a jury would be faced with competing credible medical opinions in determining whether there is a reasonable doubt as to Ed-munds’s guilt. Thus, we conclude that the record establishes that there is a reasonable probability that a jury, looking at both the new medical testimony and the old medical testimony, would have a reasonable doubt as to Edmunds's guilt.
Id. at 598-99. Thus, the court of appeals remanded the case for a new trial, not because the defendant had established that she was actually innocent of the offense, but because she had established a reasonable probability that a new jury, hearing both the new scientific evidence and the old medical testimony, would have a reasonable doubt as to her guilt. Id. at 599. This is the same standard as that for establishing prejudice in an ineffective assistance of counsel claim or materiality in a Brady claim. The defendant’s position is not that he has proven actual innocence, but that he has raised sufficient doubts as to the accuracy and reliability of the original verdict, that he is entitled to a new trial. See also Burrv. Brariker, No. 1:01CV393, 2009 WL 1298116, at *21 (M.D.N.C. May 6, 2009) (federal magistrate recommended granting habeas relief to state-court death-sentenced capital murder defendant for death of child based on counsel's failure to call biomechanical experts, including Dr. Plunkett, and specialized doctors who would have disputed cause of death as an intentionally inflicted blow to the head rather than an earlier short-distance fall; "This court concludes that there is a reasonable probability that the jury, or at least one juror, in Petitioner's case would have formed a reasonable doubt as to his guilt on the charge of first-degree murder had they heard the expert medical testimony that should have, and could have, been presented on Petitioner’s behalf.”), report adopted, 2012 *851WL 1950444 (M.D.N.C. May 30, 2012) (not designated for publication). See generally “Shifted Science," supra note 2 at 267-71 (discussing Edmunds, "shaken baby syndrome,” traumatic head injuries and the emergence of new scientific theories that undermine confidence in the original trial verdict).
. Herrera v. Collins, 506 U.S. 390, 113 S.Ct. 853, 122 L.Ed.2d 203 (1993)
. 470 U.S. 68, 83-87, 105 S.Ct. 1087, 84 L.Ed.2d 53 (1985).
. Henderson v. State, 962 S.W.2d 544, 551-57 (Tex.Crim.App.1997).
. Sehlup v. Bela, 513 U.S. 298, 115 S.Ct. 851, 130 L.Ed.2d 808 (1995).
. Ex parte Henderson, 246 S.W.3d 690, 692 (Tex.Crim.App.2007).
. 33R.R. 854.
. 33 R.R. 868.
. Ex parte Henderson, 246 S.W.3d at 696-97 (Keasler, J., dissenting).
. Ex parte Brown, 205 S.W.3d 538, 545 (Tex. Crim.App.2006); Ex parte Robbins, 360 S.W.3d 446, 464 (Tex.Crim.App.2011) (Price, J., concurring).
. Ex parte Franklin, 72 S.W.3d 671, 678 (Tex.Crim.App.2002).
. Ex parte Elizondo, 947 S.W.2d 202, 209 (Tex.Crim.App. 1996).
. Ex parte Robbins, 360 S.W.3d at 458 (citing Ex parte Elizondo, 947 S.W.2d at 206).
. Findings of Fact and Conclusions of Law on First Subsequent Application for Post-Conviction Writ of Habeas Corpus at 11, No. 94-2034 (299th Dist. Ct., Travis County, Tex. May 17, 2012) (hereinafter "Findings of Fact and Conclusion of Law”).
.Ex parte Reed, 271 S.W.3d 698, 727 (Tex. Crim.App.2008).
. 31 R.R. at 347.
. 33 R.R. 903.
. Findings of Fact and Conclusions of Law at 7.
. Colella v. State, 915 S.W.2d 834, 839 n. 7 (Tex.Crim.App.1995) (citing Foster v. State, 779 S.W.2d 845, 859 (Tex.Crim.App.1989)); see also Hernandez v. State, 939 S.W.2d 173, 178 (Tex.Crim.App.1997); Valdez v. State, 623 S.W.2d 317 (Tex.Crim.App.1981); Holloway v. State, 525 S.W.2d 165 (Tex.Crim.App.1975).
. Findings of Fact and Conclusions of Law at 3.
. 360 S.W.3d at 458.
. Id. at 450.
. Id. at 454.
. See id. at 457-58.
. Id. at 458.
. See id. at 459.
. Id.
. See id. at 466 (Price, J., concurring).
. Daubert, 509 U.S. 579, 113 S.Ct. 2786; Kelly, 824 S.W.2d 568.
. Nenno v. State, 970 S.W.2d 549 (Tex.Crim.App.1998).