dissenting. In November, 1983, a jury convicted the petitioner, Lawrence J. Miller, of viciously assaulting a young girl, and boy in Danbury on August 20, 1981. His conviction was later upheld on appeal. State v. Miller, 202 Conn. 463, 522 A.2d 249 (1987).
In 1996, the trial judge in Miller’s habeas corpus hearing ordered him released and granted him a new trial on the basis that he was innocent. The majority upholds the habeas court while announcing a novel and unique standard of reviewing habeas corpus determinations. Because of the far-reaching effect of this decision, I respectfully disagree and dissent.
I
The majority adopts a rule of habeas corpus review of independent claims of actual innocence that is not found in any other jurisdiction. In the ease with which new trials may be ordered, Connecticut will now lead the nation. The court now holds that our standard will be: (1) there was clear and convincing evidence of innocence; and (2) in the circumstances, no rational trier, drawing the same inferences as the habeas judge did, would find the habeas petitioner guilty. See part II A of the majority opinion. This standard leaves every criminal conviction subject to retrial by a habeas judge.
As to the standard’s first part, “clear and convincing evidence” means merely that the claim is “highly proba*819bly” true. The majority states that it may not be found if the evidence is “loose, equivocal or contradictory.”
The majority here seeks to comfort society by stating that Miller is not released outright, but may be faced with a new trial. A trial sixteen years after the event, however, with its continued devastating effect upon the victims and the difficulty of collecting and analyzing evidence is not something that should be ordered in the absence of evidence that unerringly demonstrates the defendant was innocent. In re Clark, 5 Cal. 4th 750, 766, 855 P.2d 729, 21 Cal. Rptr. 2d 509 (1993).
The second component of the majority’s standard concerns the effect of the evidence supporting a claim of actual innocence upon a juror or judge. The standard emerging in the United States Supreme Court and in the other states is that upon such evidence no reasonable trier of fact could find proof beyond a reasonable doubt or would do so. Herrera v. Collins, 506 U.S. 390, 429, 113 S. Ct. 853, 122 L. Ed. 2d 203 (1993) (White, J., concurring) (“no rational trier of fact could find proof of guilt beyond a reasonable doubt” [emphasis added; internal quotation marks omitted]); Ex parte Elizondo, 947 S.W.2d 202, 209 (Tex. Crim. App. 1996) (“[t]he petitioner must show . . . that no reasonable juror would have convicted him in light of the new evidence” [emphasis added]). This court adopts the less restrictive “would” standard and goes further by giving controlling weight to the effect of the evidence upon the mind of the habeas judge. Under this standard we are to be bound by the inferences the habeas judge drew from the evidence. We would review the record to ascertain whether there is substantial evidence to support such inferences. The court thereby puts the ultimate issue in the hands of a single judge applying “all of his or her powers of intellect, common sense, judgment, reason, and knowledge of human nature . . . .” This effectively blocks our appellate review of the habeas court’s *820conclusions. I do not believe the United States Supreme Court or any other jurisdiction has established such a free-ranging commission whether to set aside criminal convictions. I do not agree that such a standard gives due consideration to the public safety, to an innocent convicted person’s rights, or to public confidence in our criminal justice system.
II
The facts of this case are these. On August 20, 1981, at about 8:30 p.m., fifteen year old Elizabeth and sixteen year old William were assaulted in Danbury. During the assault, Elizabeth had a face-to-face view of the assailant as he fought with her and she scratched his face, until he beat her unconscious with a handgun. William was handcuffed, then beaten and strangled to unconsciousness. When William regained consciousness, he did not see Elizabeth bloody, half naked and unconscious, and he ran for help. Elizabeth was later found nearby. She suffered a fractured jaw, fractures of both hands, a lacerated liver as well as other head and facial wounds. William suffered many head and facial wounds.
At the scene the police found the bandanna the assailant had worn on his face and had lost during his struggle with Elizabeth and some metal objects identified as parts of a Marksman air pistol. They also had the handcuffs the assailant had put on William.
During the investigation, the police showed to Elizabeth photographs of Miller that she identified on the third occasion as “99 percent sure.” Subsequent efforts to have her identify Miller in person at church services were not availing of any identification. On July 12,1982, however, while at a department store in Danbury, Elizabeth saw and recognized Miller as her attacker. According to Elizabeth, when Miller noticed her in the *821store he tried to hide and he avoided eye contact. Elizabeth followed him from the store, took his license plate number and reported this to the police. Miller’s arrest followed. Elizabeth testified to these facts at Miller’s trial and identified him as the one who had assaulted her and William. She testified she had scratched Miller and hit him in his face during the struggle.
Miller was employed in August, 1981, as a prison guard at the federal correctional institute in Danbury. During his workshift from midnight to 8 a.m. on the night of the assault, a fellow guard noticed abrasions on both his hands, a scratch on his neck and a red line down his face as if he had been in a fight. As a federal prison guard, Miller would also have had access to the Peerless handcuffs found on William. These handcuffs had been sold to the federal government.
Miller claimed to have been at his sister’s house in nearby Brewster, New York, at the time of the assault, getting ready to leave with his wife the next day for Atlantic City, New Jersey. He produced testimony that he and his wife had left their children at his sister’s for the weekend before he went to work. Miller denied that he was scratched or marked up that night. He claimed that he left for Atlantic City the next morning upon leaving work. He admitted that in applying for emergency leave to take the trip, he gave a false excuse. Miller did produce an Atlantic City hotel receipt as evidence of his stay there on August 21. The reservation, however, was made by the wife of a present prison inmate at the federal correctional institute, and room service at the room was signed for by Maryann Raiko, a person not known to Miller or his wife.
The jury deliberated a short time and found Miller guilty.
At the habeas hearing in 1995 and 1996, Miller produced the testimony of Daniel Johnston. Johnston testi*822fied that he had attacked Elizabeth and William on the night in question. Johnston claimed that he had sexual intercourse with Elizabeth while she was conscious and after she agreed to remove her clothes and have sex with him. He also testified that while managing an apartment complex in Norfolk, Virginia, he had stolen a Marksman air pistol and a pair of Peerless handcuffs from the tenants’ rooms. Johnston also testified that he had known Miller for many years, that Miller had been a deputy sheriff in Putnam County, New York, and that he had had many contacts with Miller and Miller’s wife after Miller had been arrested for the assaults. Johnston also testified that he, as did Elizabeth and William, had developed poison ivy after the incident.
At the hearing Elizabeth repeated her testimony that Miller had attacked her that night in 1981. She also testified that Johnston was not her assailant.
The habeas court resolved the conflicts between Elizabeth’s testimony and that of Johnston in favor of Johnston. Relying on circumstances surrounding Johnston’s claims, some corroboration of Johnston’s claim that he stole a Marksman pistol in Virginia, the sexual assault committed on Elizabeth of which she was not conscious, and the poison ivy reaction, the trial court concluded that the mosaic of evidence would render Miller’s conviction at a trial where Johnston testified unlikely.
The habeas judge, and the majority finds evidence to uphold him, considered Elizabeth’s identification of Miller as based “on her sincere belief, ” but found factors tending to erode confidence in her initial and continuing identification of Miller as her assailant. Thereupon, the habeas court concluded that Miller’s continued incarceration was unjust and that he should be released.
*823Johnston had further testified that he was arrested in New York in December, 1985, for the murder of Isabelle Dell of New York, whose body had been found in Danbury. He testified that he was tried, convicted and sentenced to thirty-eight and one-half years to life imprisonment in New York state. See People v. Johnston, 147 App. Div. 2d 589, 537 N.Y.S.2d 882 (1989). His complaint was that he should have been tried in Connecticut where the body had been found.
Johnston also testified that in 1991 or 1992 he was visited in prison by two Connecticut state troopers who told him that the fingerprint evidence against him in the Dell case had been fabricated. He stated that the officers then offered Johnston $100,000 in stolen drug moneys and seven years to serve on the Dell murder if he did not go to Connecticut on the Miller case. He testified with respect to that meeting in these words:
“[A trooper] said, ‘We don’t want you to come back to Connecticut.’ He says ‘It’s too embarrassing.’ So I said, ‘Well, you’re not making sense here.’ He says, ‘Why do you think we didn’t prosecute you for your crime in the state of Connecticut?’ I says, T still don’t follow you.’ He says, ‘Well, we know your crime happened in the state of Connecticut.’ He says, ‘We just deferred on it. We didn’t want to prosecute you in the state of Connecticut. We didn’t want you in the state of Connecticut.’ ”
Johnston further stated: “I believe they did not want me back, just as I believe the Danbury police department did not want to have the Dell case in the state of Connecticut. I believe they wanted me far away from the state of Connecticut.”
It is difficult to credit Johnston’s testimony when all the evidence was that he first confessed to the 1981 attack in 1995.
*824Johnston testified that he wanted to come to Connecticut to expose the corruption in the Dell murder case. In this manner, Johnston attempted to weave his own case into his Miller testimony. The habeas court’s observation, therefore, that Johnston’s confession would be against his penal interest does not stand scrutiny. Johnston admitted to an offense well beyond the statute of limitations and did so while incarcerated for a life term in New York state with a first parole eligibility date of 2021. It could be believed by a habeas court that his testimony in the Miller case was designed to aid his own claim that he should have been prosecuted in Connecticut. Rather than finding it adverse to his penal interest, some habeas judge could find Johnston’s testimony was in support of his quest to overturn his own New York murder conviction.
The habeas court relied on corroborative testimony to find Johnston’s evidence persuasive. First, the habeas court focused on the pellet gun parts found at the scene. When Johnston contacted Miller after Miller’s arrest, the Marksman parts had been publicly identified at the crime scene. Only sometime in 1982 when the tenant Jack Lee Zollars moved from his apartment in Norfolk, where Johnston had lived and worked, did Zollars notice that his Marksman pellet gun was missing. He could not tell the court whether it had been stolen from his apartment, and he believed it had been lost in his moving out. It, therefore, could not be proven the gun had been stolen before the 1981 assault. Ffirthermore, before 1981 there had been over three million Marksman pellet guns distributed in the United States. There was no evidence, apart from Johnston’s testimony, that Zollars’ gun, if it had been stolen by Johnston before the assaults, was the source of the parts left at the crime scene.
With respect to the initials on the pellet gun parts, Johnston prefaced his statement about them by surmis*825ing that a crime scene investigator had marked the gun slide, and he made a similar remark about the bandanna holes surrounded by marks made by the crime laboratory.1 This hardly corroborated his testimony.
Poison ivy reactions are so common as to prove nothing. When Johnston was admitted to a Virginia hospital on the morning of August 23, 1981, he suffered from acute tonsillitis, pneumonia and a poison ivy rash. At admission, his body showed no signs of a struggle with Elizabeth. A physical examination revealed that his skin was essentially negative except for a poison ivy rash. The record of the physical examination, although noting the location of his rash and old scars over a fractured bone and an operation site, listed no other injuries to his face, neck and hands.
With respect to a sexual assault, the manner of the assaults and the condition of Elizabeth where she was found, well known to Johnston when he gave his statement, strongly suggested one did take place.
Lastly, Johnston gave explicit details of his vaginal, anal and oral rape of Elizabeth. His recitation, vital to his testimony, included conversations with her and her aiding him in intercourse. This the habeas court itself found unbelievable.
It is certainly conceivable that another habeas judge using such intellect, common sense, judgment, reason and knowledge of human nature would conclude that, even should Johnston testify, Miller was likely to be convicted. Such a judge might find credible the testimony of the victim and eyewitness, Elizabeth, who has maintained throughout that Miller was her assailant. Such a judge might consider the signs of a fight Miller *826exhibited at work after the assault and his access to the government handcuffs as well. As to Johnston, an experienced habeas judge might dismiss his testimony altogether and find he had no credibility. His record, which includes murder, his having nothing to lose, his stories of sexual acceptance by Elizabeth and his claims about an offer of $100,000 and time served for his promise not to return to Connecticut to testify would clearly support such a conclusion. As to the corroboration found in the Marksman pistol parts, Johnston’s poison ivy, and the sexual assault, none of those would require a habeas court to disbelieve the victim and accept Johnston’s testimony.
In sum, the granting or denial of Miller’s release under the rule now adopted by the majority depends entirely and finally upon the individual habeas judge’s assessment of the testimonies of Elizabeth and Johnston.
I believe this court should play a different role in reviewing a habeas corpus petition claiming actual innocence after conviction upon a trial. We should bring to bear on the issue, because of our number, our diverse experiences to measure the question of the weight and effect of all the evidence upon the jury’s finding of guilty. In doing so, we should recognize human fallibility and the value of collective jury verdicts.
If we, as a reviewing court, were to play a role in such determinations of innocence, I believe we should measure the evidence by the standard whether it points unerringly to the innocence of Miller and, as announced by Justice White, it is such that no rational trier of fact could find Miller guilty if Johnston were to testify. Herrera v. Collins, supra, 506 U.S. 429 (White, J., concurring). Here this cannot be said.
Accordingly, I respectfully dissent.
Abraham Stolman, then the state’s toxicologist, testified at the trial that his office had tested the bandanna and had group typed the human blood found near the knotted end. Since the early 1980s, courts have recognized and admitted much more specific blood identification evidence.