Ben Gersten v. Daniel Senkowski, Superintendent of Clinton Correctional Facility, Eliot L. Spitzer, Attorney General of the State of New York

*591 POOLER, Circuit Judge. Respondent-appellant Daniel Senkowski, Superintendent of Clinton Correctional Facility, appeals from a memorandum, judgment, and order of the United States District Court for the Eastern District of New York (Jack B. Weinstein, Judge), entered January 21, 2004, granting petitioner-appellee Ben Gersten’s petition for a writ of habeas corpus under 28 U.S.C. § 2254, ordering that he be released from custody unless new state criminal proceedings were commenced within 60 days, and staying judgment until completion of respondent’s appeal before this Court. See Gersten v. Senkowski, 299 F.Supp.2d 84, 106 (E.D.N.Y.2004). BACKGROUND Petitioner was charged in an indictment issued July 22, 1999, by a grand jury in Nassau County, New York, with six counts of first degree sodomy under N.Y. Penal Law § 130.50, two counts of first degree sexual abuse under N.Y. Penal Law § 130.65, and one count of endangering the welfare of a child under N.Y. Penal Law § 260.10. The indictment alleged that between approximately March 15 and March 31, 1995, petitioner forcibly inserted his penis into the mouth and anus, and forcibly placed his mouth on the vagina of, his then nine year old daughter, and that on or about December 13, 1998, petitioner forced his then thirteen year old daughter to rub his penis, and forcibly touched her vagina. Pretrial Proceedings Contrary to the advice of his attorney, petitioner, a law school graduate who apparently passed the New York State Bar Examination but was never admitted to practice as an attorney, waived his right to a jury trial and requested a bench trial. The prosecution requested a pretrial hearing pursuant to People v. Ventimiglia, 52 N.Y.2d 350, 438 N.Y.S.2d 261, 420 N.E.2d 59 (1981), to seek permission to offer evidence of uncharged crimes. In particular, the prosecution sought to admit evidence that petitioner sexually abused his daughter, including forcibly compelling her to perform oral sex on him, beginning in 1990, when she was between the ages of five and seven, and that he had forced her to have vaginal intercourse with him on two occasions in November 1998, when she was thirteen years old. The court ruled that the evidence from the early 1990s could be admitted on the prosecution’s direct case solely for the purpose of enabling the court to understand the daughter’s testimony, and the evidence from 1998 could be admitted solely for the purpose of allowing the court to understand the daughter’s state of mind, medical evidence and the timing of the daughter’s disclosure of the abuse. The Trial At trial the prosecution presented five witnesses: (1) petitioner’s daughter, the alleged victim; (2) Elaine Gersten, the alleged victim’s mother and petitioner’s ex-wife; (3) Aleksander Fester, the alleged victim’s ex-boyfriend; (4) Dr. Bella Silece-hia, a medical expert who had examined the alleged victim after her revelation of the alleged abuse; and (5) Dr. Donald J. Lewittes, a child psychologist who testified as an expert on children’s psychological reactions to sexual abuse. The defense presented no witnesses and no evidence. 1. The Daughter’s Testimony Petitioner’s daughter and alleged victim, then fourteen years old, testified as follows. In 1990, when she was five years old, petitioner’s daughter lived with petitioner and her mother in a two-story house in Málveme, New York. She was in kindergarten at the time. She and her parents each slept in a second story bedroom, at *592 opposite ends of the house. At some point that year, according to petitioner’s daughter, petitioner began coming into her room late at night and touching her breasts and vagina with his hands. After doing this for some time, he also began to forcibly insert his penis into her mouth and anus and to place his mouth on her vagina. He frequently ejaculated into her mouth or anus. When he first did this, he called her a “little slut” and told her that this activity was normal and would be happening from now on. For three years, petitioner did this almost every night while the alleged victim’s mother was sleeping in the other bedroom. While petitioner was not working, and he and the alleged victim were alone together for much of the day on most days, petitioner abused her only at night while her mother was in the house, and never abused her while her mother was away from the house. Petitioner’s daughter testified that the abuse continued almost every night from when she was five until her parents separated and petitioner moved out of their house in Málveme when she was nine. Around March 15, 1995, petitioner and his wife told petitioner’s daughter that they would be separating and getting divorced. Every night between March 15 and 31, petitioner entered his daughter’s room late at night, entered her bed, said dirty words to her, touched her breasts and vagina, and forcibly inserted his penis into either her mouth or her anus. When he placed his penis into her mouth he ejaculated. He also placed his mouth on her vagina. In addition he called her his “little slut,” told her that her mother knew and approved of what he was doing, and threatened to kill her if she told anybody. Petitioner moved out of the house soon thereafter, around April 1. In November 1998, petitioner and his wife had separated and petitioner was living in an apartment in the Chelsea neighborhood of Manhattan. Petitioner’s daughter visited him there on the weekend of November 7 and 8. On one of those nights, around midnight, petitioner inserted his penis into her vagina. She had gone to bed around ten o’clock, and at around twelve he entered her room, told her he was going to make her into a “real woman,” undressed, and forced her to undress. When she complained, he called her a “little slut.” He then got on top of her, pushed her legs open with his hands, and forced his penis into her vagina. After ejaculating, he dressed and left the room. Petitioner’s daughter next saw petitioner later that month over Thanksgiving weekend. She visited him at his apartment again. One night, after she went to bed, petitioner entered her room, began to undress, and forced her to undress. He got on top of her and forced his penis into her vagina, at the same time telling her that he would kill her mother if she ever told anybody. After ejaculating, he dressed and left the room. On December 13, 1998, petitioner visited his daughter and ex-wife at their home in Syosset, New York. While her mother watched television in another room, petitioner’s daughter was with petitioner in the kitchen. Petitioner allegedly forced his daughter to put her hand into his pants and touch his penis. He called her a “bad girl” and a “little slut” and then unbuttoned her jeans, placed his hand into her underpants, and touched her vagina. Petitioner’s daughter testified that she and Aleksander Fester had engaged in sexual touching but that he had not penetrated her vagina or anus. During the entire period of the alleged sexual abuse, petitioner’s daughter never complained about it to her mother, boyfriend, therapists, or anybody else. She was an excellent student throughout the entire period, receiving very good grades until after she *593 revealed the abuse, when she began to miss a large amount of school. Petitioner’s daughter was diagnosed with attention deficit disorder. She began seeing therapists in the third grade, and continued for the rest of her life. 2. Elaine Gersten’s Testimony Elaine Gersten testified that after giving birth to petitioner’s daughter in 1985 she went back to work full time and would often come home late. Petitioner attended law school from 1986 to 1990 and after graduating in 1990, studied for the bar. He was not employed during this period and spent a lot of time in class or studying. During this period the Gerstens hired a sitter to watch their daughter during the day while they were not at home. Sometime in 1990, the year that petitioner’s daughter started kindergarten, Elaine noticed a change in her sleeping habits. Whereas before, she had gone to sleep with relative ease, she began to appear very fearful about going to bed. She would beg her mother not to leave her, or to leave the light on, and would physically resist being put to bed, and would wake up having had nightmares. Petitioner’s daughter began seeing a psychiatrist for therapy sessions, but this behavior continued. After 1995, when Elaine and petitioner separated, petitioner’s daughter continued to have some problems sleeping, though they were apparently less severe. At some point around the same time that her erratic sleep habits first began, petitioner’s daughter was diagnosed with attention deficit disorder. Nevertheless, petitioner’s daughter received excellent grades in school throughout the entire period of the alleged abuse. When her daughter was five, six, and seven years old, Elaine noticed that when she bathed her daughter, her daughter’s vagina appeared red and her daughter would become very upset when Elaine would attempt to touch it, and would even slap Elaine’s hand away. Elaine told her daughter’s pediatrician about this, and the pediatrician told her that the redness did not seem to be anything abnormal and suggested application of an ointment. Petitioner’s daughter continued seeing various therapists during this period, and continued to have nightmares and trouble sleeping. After having passed the bar in 1990, petitioner remained unemployed until 1995, when he and Elaine decided to get a divorce. The two had what Elaine characterized as an “amicable” divorce, and ultimately agreed that petitioner’s daughter would visit petitioner every other weekend and on Jewish holidays. Petitioner moved out of the family house in Málveme in Spring 1995. Petitioner’s daughter continued to have problems sleeping and to attend therapy sessions after petitioner moved out. Between then and 1998, petitioner’s daughter consistently expressed anger and fear at the prospect of visiting petitioner, and on coming back from seeing him would express hostility. In fall of 1998 petitioner’s daughter became increasingly fearful of having to see petitioner. Elaine encouraged petitioner’s daughter to discuss her fears with her therapist. In early December 1998 petitioner’s daughter began to tell her therapists and her mother that petitioner had physically abused her in various nonsexual ways, such as slapping. On December 18, 1998, petitioner visited his daughter at Elaine’s home. Elaine, aware of the allegations of physical abuse, allowed the two to be alone together while she waited in an adjacent room separated by an open doorway. Despite the fact that she was only about 20 feet away and occasionally got up to check on them, Elaine did not perceive any inappropriate touching. On December 30, 1998, as Elaine was encouraging petition *594 er’s daughter to get ready to leave the house and go with her to the post office to mail an application to take early SATs, petitioner’s daughter “bolted into her room” and sat down on the floor of her room, rocking back and forth and repeatedly saying, “I can’t take it any more.” Elaine asked what was wrong, and petitioner’s daughter began screaming. She then recounted her story of petitioner’s alleged sexual abuse in graphic detail. Elaine then called her daughter’s therapist, who reported the abuse to Child Protective Services. While Elaine’s testimony did not make clear which of the allegations of abuse petitioner’s daughter disclosed in her outburst and which were only disclosed later, it is clear that petitioner’s daughter did not mention in her outburst the abuse that had allegedly occurred only two weeks earlier. This alleged incident of abuse was only revealed later, in interviews with law enforcement. Elaine admitted that during all of the years that the abuse allegedly occurred, she never saw petitioner touch his daughter inappropriately. S. Dr. Bella Silecchia’s Testimony Dr. Bella Silecchia, the director of pediatric services and the Suspected Child Abuse and Neglect Evaluation Program at the Nassau County Medical Center, testified about the results of her physical examination of the alleged victim. One of Dr. Silecchia’s duties is to conduct physical examinations of children in cases where sexual abuse is alleged. In her examinations Dr. Silecchia uses various instruments including a colposcope, which is a gynecological tool that magnifies and photographs areas where injuries or trauma may be visible. Dr. Silecchia examined the alleged victim on February 4, 1999. She first examined the alleged victim’s genitalia visually, discovering no trauma or discharges and nothing out of the ordinary for a girl of her age. Dr. Silecchia then used the colpo-scope to examine the alleged victim’s vagina, detecting “neovascularization of ... the posterior fourchette.” The “posterior fourchette” refers to the location where the labia majora come together, and “neo-vascularization” occurs when blood vessels break and reform. Dr. Silecchia explained that this was indicative of “chronic irritation, chronic manipulation, chronic rubbing,” which caused the labia at this location to become “inflamed and irritated.” Dr. Silecchia next separated the alleged victim’s labia and used the colposcope to examine her hymen. She observed “a thickened, fimbriated, redundant type of hymen,” which “is a hymen where the tissues are thickened and sort of obscure the opening into the vaginal canal.” She clarified that the hymen “had thickened rolled edges,” which she explained would not be normal in a prepubsecent child, whose hymen should have “fairly sharp edges; if not sharp, at least velamentous, thin looking, not thick and rolled back as you would see in a hymen that has either been stretched out or tom.” Using the colposcope to look into the vaginal canal, Dr. Silecchia testified that she observed “clefts in the 5 o’clock, 9 o’clock and 11 o’clock positions.” “Clefts,” which Dr. Si-lecchia stated are also called “notches,” are V-shaped “interruptions in the normal contours of the hymenal tissue.” Dr. Silecc-hia stated that clefts indicate tearing of the hymen at those locations. Dr. Silecchia had the alleged victim change positions, so that her hymenal tissue would be more relaxed, and she went around the hymen with a Q-tip to verify the clefts. Dr. Si-lecchia’s professional opinion was that “there had been trauma to the hymenal tissue of a penetrating type.” The neovas-cularization had been caused by some kind of “stretching, irritation, chronic manipulation, possibly a rubbing,” and the clefts *595 and notches had been caused by penetration. Dr. Silecchia next conducted an anal examination. Using the colposcope she noticed “some smoothening, not totally flattened but smoothed out rectal mucosa or ruga or puckering around the anus. The normal anal folds that were there showed some slight amount of smoothening out. And ... in the 10 o’clock position we saw a healed fissure with a scar.” This indicated to Dr. Silecchia that something had passed through the alleged victim’s anus that was large enough to tear it. Dr. Silecchia also saw “marked venous congestion of the anal mucosa around the opening of the anal canal,” which is “often ... due to ... stretching, pressure on the vessels in that area.” Dr. Silecchia stated that in her professional opinion, her findings “were highly suggestive of penetrating trauma to the hymen and chronic irritation of the ... posterior fourchette and perihymenal tissues; and tearing of the mucosa of the rectum due to the stretching out of the rectal mucosa-of the anal mucosa.” She stated that the findings could not be explained by blunt, nonpenetrating trauma such as “fall[ing] onto the bar of a bike or something,” and that the trauma to the hymen could not be explained by masturbation because a child masturbating for pleasure would tend to focus on the clitoris and not the hymen, and in any event to explain the penetrating trauma that Dr. Silecchia found would require masturbation to the point of pain and bleeding, “and before [a child] got to that point she would stop.” On cross-examination, defense counsel apparently intended to question Dr. Silecc-hia about the alleged victim’s sexual history to demonstrate that she had engaged in sexual behavior with her ex-boyfriend, but Dr. Silecchia revealed that she did not take the alleged victim’s sexual history because her staff took it before she began the examination. Counsel thus moved on to ask about the “notches” that Dr. Silecchia reported having discovered, asking whether they could have been caused by “sexual activity.” Dr. Silecchia replied that “[t]hat could be one of the penetrating types of trauma.” Counsel asked whether it would be consistent with simultaneous penetration by multiple fingers of a teenager, to which Dr. Silecchia replied that it would not, “[b]ecause penetration with fingers usually shows a type of trauma to the hymen which is between the 9 and 3 o’clock position.” On the other hand, “[pjenile penetration usually causes damage [between] the 3 to 9 o’clock position.” Dr. Silecchia stated that this was the “classic” pattern for penile penetration, as established by “numerous literature studies.” Dr. Silecchia admitted that it was “possible” that the clefts she observed were caused by fingers, “[i]f they were large enough and numerous enough in insertion.” Counsel obtained Dr. Silecchia’s concession that she could not establish when the damage to the hymen occurred, except that it was more than a week before the examination. Moving on to discuss the anal examination, counsel obtained Dr. Silecchia’s concession that she also could not establish when the damage to the rectum occurred, except that it was more than a week before the examination. Dr. Silecchia stated her opinion that the damage was “chronic” and would have produced bleeding. Counsel obtained Dr. Silecchia’s admission that if repeated constipation was in the alleged victim’s medical history, then it could be the cause of the damage she observed to the alleged victim’s anus. Counsel also obtained Dr. Silecchia’s admission that it was not the case that “the only way this tear would be there would be from a male penis being inserted into the anus,” and *596 that there were nonsexual ways in which it could have been caused. Defense counsel apparently did not even ask to be allowed to examine the slides of the photographs taken during Dr. Silecc-hia’s examination using the colposcope during pretrial disclosure, and counsel did not once refer to these slides during cross-examination. On redirect, the prosecution established that Dr. Silecchia had not ascertained any chronic constipation in the alleged victim’s medical history. k- Dr. Donald J. Lewittes''s Testimony Dr. Donald J. Lewittes, a clinical psychologist specializing in child and adolescent trauma including child sexual abuse, testified primarily about Child Sexual Abuse Accommodation Syndrome. He credited the origins of the concept of this syndrome to a “Dr. Summit,” who also coined the term. He described it as an explanation of the ways in which children are typically emotionally and psychologically affected by sexual abuse. While every child is unique and may have individual and idiosyncratic reactions to such abuse, generally, the theory postulates, children’s reactions proceed through five identified stages. These are secrecy, helplessness, delayed disclosure, entrapment and accommodation, and blurred memories. Thus, Dr. Lewittes opined that because children perceive the need for incestuous sexual contact to be kept secret, and feel trapped by the familial structure which provides no avenue for avoiding such contact when it is forced on them, they are likely to disclose such contact only after a period of delay during which they react by instead becoming anxious or depressed or having trouble concentrating. The child may overcompensate for the extremely uncomfortable feelings that she experiences during the abuse by focusing on achieving a sense of normalcy in other aspects of her life. Thus, she may focus on schoolwork and getting good grades, attempting to build normal relationships at school with other children or teachers. When disclosure ultimately occurs, there is likely to be a “blurring of distinct elements or borders between one event to the next,” because of the child’s attempts while the abuse is occurring to “space out,” and attempts af-terwards to repress the memory. The child is also not likely to recall peripheral details, such as what an abuser was wearing or the specific date or time on which abuse occurred, because during traumatic experiences such as abuse children typically focus their attention on the abuse itself and the physical and emotional pain that it causes. The memory “blending effect” also makes it difficult for children to recall such specific details, particularly when there were numerous incidents over a period of time. Dr. Lewittes analogized the memory process of a sexually abused child to “tak[ing] the different sexual acts and ... throwing] them into a blender and they all kind of combine in some way so that you don’t have the element of date and time for each distinct act[ ].” Dr. Lewittes opined that a child will likely not disclose sexual abuse until she has reached her “saturation point,” when she “can’t absorb the stress any longer,” and will feel she has to talk about it. He opined that it is common for disclosure to be delayed until after a child reaches adolescence, because “[o]ften when a child reaches adolescence, they have much more awareness of their own sexuality, they have much more understanding about the nature or the consequences of a sexual act.” Dr. Lewittes opined that good performance in school is not an indicator of whether sexual abuse has occurred. Some children may react to sexual abuse by losing the ability to concentrate and perform in school, whereas other children react by applying themselves to their studies and *597 performing well. Dr. Lewittes opined that attention deficit disorder is sometimes misdiagnosed where a child is actually suffering from anxiety or post-traumatic stress disorder as a result of physical or sexual abuse. On cross-examination, Dr. Lewittes conceded that he had no knowledge of the facts of the instant case and could not offer any opinion specific to the case as to, for example, whether the alleged victim was correctly diagnosed with attention deficit disorder. To determine whether such a diagnosis was correct, one would have to know all of the relevant environmental factors, including any emotionally traumatizing events such as sexual abuse. In response to counsel’s question whether “in your professional career is it your opinion that everyone was telling the truth?” Dr. Lewittes responded that “alleged perpetrators ... often lie,” and he knew this because perpetrators who initially denied any involvement in sexual abuse had later admitted it to him. Counsel then asked whether any children had ever admitted having lied about sexual abuse, to which Dr. Lewittes responded that “[tjhere have been children who have told me that they have, for example, stated that they were abused and then recanted it,” but he stated that there were “very, very few,” and that the “vast majority” both in his personal experience and according to “the research literature” “would not he about something like sexual abuse.” Dr. Lew-ittes conceded that “[fit’s possible” that a “small minority” of children would lie about sexual abuse for one reason or another. Defense counsel did not ask Dr. Lew-ittes any questions about the supposed scientific basis for his testimony regarding Child Sexual Abuse Accommodation Syndrome, or his opinion that disclosure of sexual abuse is more likely after the onset of adolescence, nor did counsel ask any questions about the “research literature” that purportedly supported his conclusion that children are very unlikely to lie about sexual abuse. On redirect, Dr. Lewittes testified that in determining whether a child was lying about sexual abuse it would be important to determine whether there was any motivation to lie, such as involvement in a custody dispute. 5. Aleksander Fester’s Testimony Aleksander Fester, who was then thirteen years old, testified that he had been friends with the alleged victim for two or three years and that for a year they were boyfriend and girlfriend. When they were romantically involved, they did not do anything “hard core,” but they did engage in “a mutual masturbation kind of thing” in which “she would touch [his] genitalia, [he] might touch hers sometimes, but it was rarely beneath the clothes, if anything.” They never engaged in vaginal or anal intercourse and he never inserted anything into her vagina. On cross-examination defense counsel established that the sexual relationship between the alleged victim and Mr. Fester ended in February 1999. Mr. Fester also testified that the alleged victim did not tell him about her father’s abuse until after she had already told her mother and others, and that she continued to engage in sexual activity with him after those revelations. 6. Summation, Verdict, and Sentencing After the prosecution rested, defense counsel moved to dismiss the charges, arguing that the prosecution had not presented evidence sufficient to establish that the charged events occurred on the dates alleged in the indictment. The motion was denied. The next day, the defense rested without presenting any evidence. In summation, counsel argued that the alleged *598 victim was either lying or engaged in fantasy. But counsel repeatedly disclaimed any intention to offer any explanation for why the alleged victim might lie about her father sexually abusing her. Counsel, rather, stated that the court “may never have an answer” to the question of why the alleged victim would have lied, and “I can’t submit to this Court why, nor am I going to try, as to why [the alleged victim] would have a reason or a basis to lie, so I won’t.” Rather, counsel simply argued that the alleged victim’s testimony did not make sense in light of her mental health issues, her excellent record in school, and the implausibility that several years of daily abuse could occur without her mother ever finding out and without her ever telling anybody. While counsel argued that both of the prosecution’s experts were evasive when questioned by the defense and appeared biased in favor of the prosecution, counsel conceded the validity of their scientific conclusions, arguing only that they were “consistent with innocence” as well as with guilt. Counsel made no attempt to argue that the physical or psychological evidence was inconsistent with guilt. Counsel conceded that Dr. Silecehia’s examination found evidence of penetration. Counsel argued only that the penetration could be explained by the sexual relationship with Mr. Fester. After summations concluded, the court immediately rendered a guilty verdict. At sentencing on January 5, 2000, petitioner continued to maintain his innocence. The alleged victim and her mother Elaine made statements regarding the impact of his alleged abuse. The prosecution requested that the court impose the maximum sentence authorized by law, which would be consecutive terms of the maximum allowable sentence for each of three of the sodomy counts, with other sentences to run concurrent. Defense counsel, amazingly, informed the court that the petitioner felt that “if I was guilty of this ... then I should go to jail for the rest of my life.” The court, explaining to the defendant that it “fe[lt] that your actions have ruined the lives of many people,” ultimately sentenced petitioner to three consecutive indeterminate sentences of twelve and a half to twenty-five years imprisonment on three of the sodomy charges, concurrent indeterminate terms of twelve and a half to twenty-five years on the other three sodomy counts, concurrent determinate terms of seven years imprisonment on each of the two sexual abuse charges, and one year imprisonment on the charge of endangering child welfare. The court thus imposed the maximum possible sentence. State Direct Appeal On direct appeal, the Appellate Division found that the sentence was unauthorized by law, and modified it to reflect the statutory requirement that the minimum sentence imposed be one third of the maximum, resulting in sentences of eight and a third to twenty-five years imprisonment on the six sodomy convictions, with the first three to run consecutively to one another. The Appellate Division approved the imposition of consecutive sentences and otherwise affirmed the convictions, rejecting various evidentiary arguments as unpre-served. In rejecting petitioner’s ineffective assistance of counsel claim, the court found that “under the totality of circumstances existing at the time of representation, the defendant received meaningful representation.” People v. Gersten, 280 A.D.2d 487, 719 N.Y.S.2d 900, 900 (2d Dept. 2001). Leave to appeal to the New York Court of Appeals was sought and denied. People v. Gersten, 96 N.Y.2d 901, 730 N.Y.S.2d 799, 756 N.E.2d 87 (2001). *599 Motion Under N.Y.Crim. Proc. Laid Article W While his state direct appeal was pending, petitioner filed a motion in the trial court under New York Criminal Procedure Law § 440.10 to vacate the judgment of conviction on the ground that he was deprived of the effective assistance of trial counsel. He alleged that trial counsel had conducted an inadequate pretrial investigation, improperly failed to call expert witnesses to rebut the prosecution’s experts, improperly neglected to request a hearing to challenge the admissibility of the expert testimony, failed to show a motive for the alleged victim to lie, improperly waived a jury trial, and failed to use a diary and information concerning his daughter’s use of the Internet during cross-examination. In support of his motion petitioner submitted affidavits from two experts: Dr. Jocelyn Brown, a medical doctor, and Dr. John C. Yuille, a psychologist. 1. Dr. Jocelyn Brown’s Affidavit Dr. Jocelyn Brown, a doctor specializing in pediatric medicine and the founding and then current director of the Child Advocacy Center at Columbia Presbyterian Hospital, submitted an affidavit reviewing the medical evidence. Dr. Brown’s work primarily concerns forensic medicine in the area of child sexual abuse, and her practice frequently involves close cooperation with law enforcement. Dr. Brown had testified as an expert witness in New York state criminal proceedings in approximately 80 cases. In 78 of these she testified on behalf of the government. In preparing her affidavit Dr. Brown reviewed Dr. Si-lecchia’s testimony as well as the alleged victim’s medical records and the slides of the magnified photographs taken during the examination through the colposeope. Regarding Dr. Silecchia’s finding that “neovascularization” of the “posterior four-chette” tended to indicate “chronic irritation, chronic manipulation, and chronic rubbing,” Dr. Brown stated that “such a finding is no longer considered of significance in the forensic community when evaluating children suspected of being sexually abused as the condition of ‘neo-vascu-larization’ can arise from enumerable factors unrelated to sexual abuse.” Dr. Brown noted that the American Professional Society on the Abuse of Children (“APSAC”) had “directed that ‘the use of this term should be discouraged.’ ” Regarding Dr. Silecchia’s finding that the hymen’s “thickened, rolled edges” indicated stretching or tearing, Dr. Brown stated bluntly that “Dr. Silecchia’s opinion on this issue is erroneous.” Dr. Brown explained that at the time of the examination the alleged victim was thirteen years old and post-pubertal, and a “curled, rolled, or ‘redundant’ hymenal edge is commonly found in post-pubertal children and is not indicative of penetrating trauma.” Regarding the clefts reported by Dr. Silecchia at the 5, 9, and 11 o’clock positions on the alleged victim’s hymen, Dr. Brown noted a number of erroneous aspects of Dr. Silecchia’s testimony. First, Dr. Brown stated that Dr. Silecchia was incorrect in stating that “clefts” and “notches” referred to the same thing. Unlike clefts, “ ‘notches’ are not ‘v-shaped’ interruptions of the hymenal tissue.” Moreover, “[njotches are commonly found on the hymenal tissue of adolescents and are not indicative of penetrating trauma to the hymen.” On the other hand, clefts, which are “V-shaped” and are “transsec-tions in the hymenal tissue beginning from its edge, or rim, to its base,” “are indications of penetrating trauma.” However, Dr. Brown’s review of the colposcopic slides did not reveal any clefts at the reported locations. Rather, these magnified photographs of the alleged victim’s hymen revealed only one notch, at the 4 o’clock position, and no clefts. *600 Regarding the anal examination conducted by Dr. Silecchia, Dr. Brown reviewed the magnified photographs of the alleged victim’s anus taken through the colposcope and found that they did not reveal the “tearing of the mucosa” or the “smoothed out rectal mucosa or ruga or puckering around the anus” reported by Dr. Silecchia. Moreover, Dr. Brown stated that “the association of these findings with sexual abuse remains uncertain as such findings are not currently recognized in the scientific community as dispositive of penetrating trauma to the anus.” Dr. Brown further stated that the “healed fissure with a scar” discovered by Dr. Silecc-hia was “of no clinical significance” because such a fissure “can be caused by any number of factors unrelated to penetrating trauma of the anus,” and “typically heals ... within a one week period.” Thus, in essence, Dr. Brown concluded that the physical evidence did not appear in any respect to be indicative of penetrating trauma to the alleged victim’s vagina or anus, and thus none of the medical evidence corroborated the allegations of abuse or the alleged victim’s testimony. Dr. Brown stated that had she been consulted by the defense prior to trial, she would have offered the opinions contained in her affidavit, and had she been called to testify, she would have given the testimony contained in her affidavit. 2. Dr. John C. Yuille’s Affidavit Dr. John C. Yuille, a forensic psychologist and professor at the University of British Colombia, submitted an affidavit reviewing Dr. Lewittes’s testimony. Dr. Yuille is a member of various professional associations for forensic psychologists and the author of over seventy published journal articles, six texts, and twenty-one text chapters on topics related to the psychology of child sexual abuse. His work has included training law enforcement in evaluation of the credibility of allegations of child sexual abuse, and he has testified as an expert on the psychology of child sexual abuse in numerous judicial proceedings in the United States and Canada. In the vast majority of these cases he testified on behalf of the government. Dr. Yuille stated that “[i]n point of fact, what was once known as ‘Child Sexual Abuse Accommodation Syndrome’ is no longer regularly accepted in the child sexual abuse research community.” He reported that the inventor of the theory, Dr. Roland Summit, “has retreated from his position of classifying it as a ‘syndrome,’ ” and that “[rjesearch in the field reveals that the ‘symptoms’ which were the alleged components of the syndrome do not occur regularly enough in those who truly have been the victims of sexual abuse to call it a syndrome.” Dr. Yuille concluded that “Dr. Lewittes’ testimony with regard to the child sexual abuse accommodation syndrome and its alleged five components has no validity and is not regularly accepted in the scientific community.” Regarding Dr. Lewittes’s testimony as to the impact of traumatizing events on memory-the “blurring effect”-Dr. Yuille stated bluntly that, after a distinguished career focused on understanding “issues of suggestibility, memory and child sexual abuse,” he could state with some assurance that “Dr. Lewittes’ opinion that trauma ‘blurs’ memory is erroneous. The best research in the field of child sexual abuse, memory and suggestibility uniformly reveals that trauma, has varying effects on memory, ranging from memory enhancement to diminishment.” Regarding Dr. Lewittes’s testimony that the onset of adolescence could trigger disclosure of child sexual abuse because it increases a child’s awareness of her own sexuality, Dr. Yuille reported that “[i]n my years of research in this field, I have never encountered this proposition. In short, *601 this ‘theory’ has no scientific validity in the field of child sexual abuse.” Dr. Yuille concluded that had he been consulted by the defense prior to trial, he would have offered the opinions contained in his affidavit, and had he been called to testify, he would have given the testimony contained in his affidavit. S. The County Court’s Decision The trial court denied the motion, finding that “[u]nder the totality of the circumstances the defendant did receive effective assistance of counsel.” The court elaborated, explaining that “[a]t trial defense counsel vigorously cross-examined the People’s expert witnesses. Counsel neutralized the impact of Dr. Silecchia at trial.” The court opined that notwithstanding counsel’s failure to consult with experts before trial, he was still able to put forth his theory of the case by getting the Doctor to admit that what she observed about the victim could have come from non-sexual abuse related activity. The theory of the defense was that the victim was either lying or was delusional and that the medical testimony was inconclusive. Consequently the defense counsel didn’t need to consult with or call expert witnesses to rebut the People’s experts who were neutralized by cross-examination. The court stated that the same analysis applied to counsel’s failure to consult a psychological expert. The court viewe[ed] trial counsel’s decision not to call independent experts as nothing more than a tactical decision since he was able to support his defense theories through cross-examination of the People’s experts. In view of counsel’s thorough cross-examination of the People’s experts which cross-examination obtained concessions that the vaginal and anal fusions [sic] could be caused by non-sexual abuse reasons or by sex between the victim and her boyfriend from Dr. Silecchia and that some children lie about sex abuse from Dr. Lewittes, the defense attorney was able to promote his defense theory. This trial tactic can not now be second guessed by the Court. The court summarized by stating that “[t]he court feels that the defendant received meaningful representation. The tactics chosen by defense counsel were reasonable and he provided zealous representation of his client. The defendant now is merely second guessing his failed trial strategy and this is not a basis to vacate the Judgment of Conviction.” The court distinguished this Court’s decision in Lindstadt v. Keane, 239 F.3d 191 (2d Cir.2001), because unlike in Lindstadt, (1) this case involved only a failure to consult or call expert witnesses while Lindstadt involved the cumulative effect of such a failure alongside other deficiencies in counsel’s performance, (2) the experts in the instant case did not offer their opinions as to the ultimate question of whether abuse occurred, rendering their testimony less significant, and (3) the experts in the instant case “did not rely on any studies or texts in reaching [their] conclusions.” As noted above, the latter was false, as both experts made reference in their testimony to unnamed studies in the research literature of their respective fields. The court thus rejected the ineffective assistance claim. It also rejected all of petitioner’s other claims. Petitioner’s application to appeal this decision to the Appellate Division was denied on June 27, 2001. Federal Habeas Corpus Proceedings On July 11, 2002, petitioner filed a federal petition for a writ of habeas corpus pursuant to 28 U.S.C. § 2254 in the United States District Court for the Eastern Dis *602 trict of New York, alleging deprivation of his right to the effective assistance of counsel by trial counsel’s failure to obtain medical discovery materials or to have them reviewed by an expert, or to consult with or call a medical or psychological expert to testify at trial. In support of this claim, petitioner submitted the same affidavits from Dr. Brown and Dr. Yuille previously submitted to the County Court in connection with his Article 440 motion. Petitioner presented a number of additional arguments in support of his ineffectiveness claim, as well as a claim of violation of due process. In opposition to petitioner’s federal ha-beas petition, respondents presented affidavits from Judge Donald E. Belfi, the County Court judge who presided over petitioner’s trial, and Dennis M. Lemke, the attorney who represented petitioner at trial. Finding that the “only substantial issue” raised was “whether trial counsel was adequate,” and finding a hearing on this issue necessary, the district court held a hearing at which Dr. Silecchia and Mr. Lemke testified. 1. The County Court Judge’s Affidavit Judge Belfi stated in his affidavit that Mr. Lemke “has always impressed” him, “as he did in this case, as extremely competent and professional,” and that his “reputation within the legal community is that of an excellent defense attorney.” The Judge stated that he based his guilty verdict against petitioner “on the testimony of the victim, whom [he] found to be credible beyond a reasonable doubt,” and added that he had “extensive experience evaluating the credibility of witnesses” and “believed, beyond any doubt, that the victim was telling the truth concerning the acts committed against her by the petitioner.” Explaining his decision to deny petitioner’s Article 440 motion, Judge Belfi stated: I now reaffirm my statement that the testimony of the People’s expert witnesses was neutralized by the cross-examination conducted by defense counsel in this case. The expert testimony was not significant or important to my determination of petitioner’s guilt or innocence of the charges. I reviewed the affidavits of petitioner’s expert witnesses submitted with his motion to vacate the judgment of conviction. The information contained therein did not undermine the testimony of the victim in this ease and, if it had been presented at trial, would not have changed my verdict. 2. Defense Counsel’s Affidavit Donald M. Lemke, the attorney who represented petitioner at trial, submitted an affidavit stating that: In preparation for the trial, I reviewed [petitioner’s daughterj’s medical records. Rather than calling any experts, I chose to advance the defense theories of the case through cross-examination of the People’s witnesses. Based on what I learned during my pre-trial investigation of the facts of the case, I was confident that I would be able to adduce innocent explanations for the observations and/or theories of the People’s experts. Petitioner informed me that he wished to resolve all the charges against him in the speediest manner possible. Against my advice, Petitioner chose to waive a jury and to proceed to trial non-jury.... It was our belief that his daughter ... would not appear for trial as she had been institutionalized during the pen-dency of these cases and speedy trial time was running. The latter was apparently a misstatement, for counsel later expressed understanding of the fact that the New York Speedy Trial statutes, N.Y.Crim. Proc. Law §§ 30.20-30, would toll speedy trial time while the victim was institutionalized. Counsel did *603 not more specifically elaborate on what his “pre-trial investigation of the facts of the case” entailed. 3. Petitioner’s Affidavit Petitioner submitted an affidavit to respond to statements made in Mr. Lemke’s affidavit. Petitioner maintained that Mr. Lemke exaggerated the extent to which he directed his own defense, stating that while he posed questions and advised his attorney about facts, he “relied entirely on Mr. Lemke’s professional advice” when it came to “important decisions.” Petitioner stated that he “did not agree with Mr. Lemke’s” decision not to consult any expert witnesses and instead to attempt to offer innocent explanations of the purported medical and psychological evidence through cross-examination. He stated that Mr. Lemke told him that it would not be possible to view the colposeopic slides or have an independent expert examine them before trial. Petitioner stated that while he urged Mr. Lemke to consult with experts on “false claims of child sexual abuse,” Mr. Lemke did not do so. Finally, petitioner stated that while Mr. Lemke had informed him prior to the trial that the alleged victim “had been hospitalized and he therefore believed that she would not be able to testify at trial,” petitioner “did not share his belief ... and told him so,” and would have preferred to have been fully prepared to “meet [his] daughter’s false accusations at trial.” k- The District Court’s Evidentiary Hearing The hearing before the district court was held on January 12, 2004. While petitioner had attempted to have Dr. Brown and Dr. Yuille available to testify, they were not available, and the court ruled that it would therefore rely on their affidavits. a. Dr. Bella Silecchia’s Testimony Respondent called Dr. Silecchia to rebut the statements in Dr. Brown’s affidavit. Dr. Silecchia testified that the slides of the photographs taken through the colposcope “rarely reflect everything that you see,” because “[i]t’s a difficult camera to get into focus.” However, she maintained that the handwritten notations on the written report of the exam reflected what was actually seen even when the slides did not seem to reflect the same thing. On cross-examination, Dr. Silecchia conceded that it is common for doctors working together to review one another’s reports against what the colposeopic slides reveal and to alter findings to accord with what can be validated with the slides. It was thus important, to ensure that the slides revealed all of the trauma that was observable. Dr. Silecchia stated that she had reviewed the colposeopic slides as soon as they were available, a couple of weeks after the examination, but did not record any observation of discrepancies between what the slides showed and what she had seen during her examination. Discussing her particular findings, Dr. Silecchia conceded that “a thickened and rolled edge hymen” is a “nonspecific finding” for sexual abuse but is, rather, indicative only of the onset of puberty. She also conceded that “neovascularization of the posterior fourchette” is “a nonspecific finding for sexual abuse.” She maintained that, despite the fact that APSAC had directed that the term should not be used in evaluating trauma, she nevertheless “like[d] the term because it is quite descriptive and tells exactly what I’m seeing.” Dr. Silecchia agreed that APSAC had directed that “only [clefts] ... in the posterior rim from the 4:00 to 8:00 o’clock position are ... usually indications of sexual abuse.” But she maintained that this was not always true and that the alleged *604 victim in this case was an exception, because her hymen was “annular” in shape, and clefts would be expected all around if the hymen was penetrated by an object larger than it. She stated that she disagreed with APSAC’s conclusion that clefts at the 9 and 11 o’clock positions were nonspecific for sexual abuse, and noted in any event that she had seen the 5 o’clock cleft. Dr. Silecchia apparently agreed with APSAC that, as a general matter, “venous congestion is of unknown significance in evaluating trauma to the anus,” that “flattened anal skin folds has an unknown relationship to sexual abuse and is a common result in a child being relaxed,” and that “a healed rectal fissure is a nonspecific finding with regard to sexual abuse.” Dr. Silecchia agreed that she had not seen any evidence of “anal dilatation,” which is a relaxation of anal musculature that may be caused by penetration. b. Trial Counsel’s Testimony Mr. Lemke testified by telephone. Pri- or to his representation of petitioner, Mr. Lemke had practiced primarily as a criminal defense attorney for about twelve years, and had tried between 200 and 300 cases, including fifty or sixty felonies. Mr. Lemke stated that after he was retained by petitioner, he reviewed the alleged victim’s medical records, and “consulted with a friend who [he] went to law school with ... who had been in the medical field .... ” However, this was a brief conversation, and “[biased on [his] conversations with Mr. Gersten and the defense that we were putting forward, strategically,” he was concerned only with determining “whether or not [the medical records] could have been consistent with any sexual relationship with a boyfriend at the time.” Mr. Lemke explained that he “did not find anything that would be helpful, or that [he] felt to be helpful at the time,” in the medical records, “based upon [the] defense.” This defense was “[t]hat [petitioner’s] daughter had some relation, sexual relation, with a friend that would be consistent,” and “that if there had been any findings consistent with any penetration, that it was as a result of the daughter’s relationship with her friend, male friend, at the time.” This was based on information that petitioner provided to Mr. Lemke to the effect that he had inquired with some people who informed him that his daughter was involved in a sexual relationship with her boyfriend. Mr. Lemke was unable to interview the ex-boyfriend, Mr. Fester, before trial because Mr. Fester refused to speak with him. On cross-examination Mr. Lemke clarified that the friend he consulted about the medical records was a registered nurse and not any kind of medical expert, and that their conversation was limited to “the defense that we were moving forward with, nothing independent, as far as whether I would take that avenue of defense differently.” Rather, his “defense right from the start was that [the alleged victim] was involved with this other individual, and anything that was consistent with that was the way [he was] going to proceed with.” Thus he did not obtain, or attempt to obtain, any information about clinical indicia of child sexual abuse, such as whether “neovascularization of the posterior fourchette,” or “flattening of mucosa around the entrance of the anus,” or any of the other indicia relied on by Dr. Silecchia, in offering her opinion that the examination revealed evidence of penetration of the vagina and anus were actually accepted as such by any scientific community. He simply did not believe that an understanding of such indicia was necessary for the defense given the theory that he had adopted from the very beginning. Other than his registered nurse friend, Mr. Lemke did not consult with anybody who had anything even approaching medical expertise in preparing for petitioner’s trial. *605 Mr. Lemke did not recall requesting to see the colposcopic slides during discovery, and conceded that he did not see them until trial. Mr. Lemke maintained that he did familiarize himself prior to trial with research literature discussing the Child Sexual Abuse Accommodation Syndrome. However, Mr. Lemke did not recall whether he had done any research into the scientific validity of that “syndrome,” stating only that he “hoped” he would have done so but that his cross-examination of Dr. Lewittes “speaks for itself.” Mr. Lemke explained that he had taken various actions in an attempt to expedite petitioner’s trial because of his belief that the alleged victim “had been institutionalized and for medical reasons was not going to come in to testify as a result of her suicide attempts.” He stated that even though the government asserted that it was ready for trial, he knows from experience that sometimes the government says it is ready for trial when it actually is not, and he believed this to be such a case. He added that he did not believe that the alleged victim would testify even if she was not institutionalized and was available to do so, because “[i]t was Ben Gersten’s belief that his control over his daughter would be that when his daughter appeared before him, if she ever appeared, she would not say a word against him, would break down and leave.” 5. The District Court’s Memorandum, Judgment, and Order The district court filed its memorandum, judgment, and order on January 21, 2004, granting petitioner’s petition for a writ of habeas corpus and ordering petitioner’s release from custody unless new state criminal proceedings against petitioner were commenced within sixty days, and staying that order pending appeal. The court found that defense counsel’s “performance was, in several respects, constitutionally deficient.” Gersten, 299 F.Supp.2d at 102. First, the court faulted defense counsel’s failure “to conduct an adequate pre-trial investigation into critical medical evidence.” For example, counsel “never obtained copies of the colposcopic slides” and “never arranged for their evaluation by an independent expert.” Id. at 103. The court found this error to be “most substantial” because it “limited [counsel’s] ability to effectively cross-examine Dr. Silecchia,” and counsel’s cross-examination in fact demonstrated that counsel possessed “little understanding of the medical fundamentals and how they might impact an expert’s evaluation of the critical question in this case.” Id. The court found that “[h]ad counsel consulted a medical expert, it is likely he would have been able to present expert testimony ... to largely rebut the conclusions of Dr. Silecchia.” Id. This would have “cast considerable doubt on Dr. Silecchia’s testimony,” and supported an inference “that no sexual abuse had occurred [and] ... that no penetrating sexual activity whatsoever had taken place.” Id. at 104. Second, the court faulted defense counsel for deciding on a theory of defense without having “conducted] an adequate investigation into alternative, complementary defenses.” Id. The court found that failing to discover and present potentially exculpatory evidence such as that now offered by Dr. Brown could not be explained by any reasonable trial strategy. This was particularly the case because the exculpatory evidence that counsel failed to discover would have been entirely consistent with his chosen defense theory that penetration, if there was any, was caused by somebody other than petitioner. Id. Moreover, a defense backed by expert medical testimony would have been “con *606 siderably more compelling than a simple denial of sexual abuse,” because it would have not only rebutted the prosecution’s medical evidence but would also “have cast considerable doubt on all of the daughter’s testimony.” Id. Third, the court faulted defense counsel for failing to consult with or call a psychological expert, finding this failure to be “an independent and sufficient indication of deficiency.” Id. at 105. Again, counsel’s cross-examination of the prosecution’s psychological expert “was pallid, showing no grasp of the scientific predicates for the testimony.” Id. This lack of grasp of the science resulted in a failure to challenge significant errors in Dr. Lewittes’s opinions and thus deprived petitioner of “another critical opportunity to damage the daughter’s credibility,” and such a deficiency could not be justified by any reasonable trial strategy. Id. Finally, the court faulted counsel for relying on his “belief that the daughter ... would be unable to confront her father in open court,” finding that “[t]here was no valid basis for such a conclusion” and that this assumption on counsel’s part could not excuse his failure to prepare an adequate defense. Id. at 102. The court found that these errors of counsel prejudiced petitioner because, “[gjiven the significance of [the medical] expert testimony,” “[t]here is a reasonable probability that petitioner would not have been convicted had defense counsel conducted an adequate investigation into the medical evidence and called an expert to testify.” Id. at 104. Counsel’s failures as to the psychological expert testimony, taken “[t]ogether with the failure to consult with a medical expert,” rendered the case for ineffectiveness all the more “serious.” Id. at 105. The County Court’s “failure ... to consider the importance of this omitted expert testimony in denying petitioner’s motion to vacate the judgment of conviction” was “an unreasonable application of the Strickland standard.” Id. at 104. Respondent timely appealed. Finding the district court correct on all points, we affirm. DISCUSSION I. Standard of Review We review de novo a district court’s decision to grant a petition for a writ of habeas corpus. Jackson v. Edwards, 404 F.3d 612, 618 (2d Cir.2005). Because petitioner’s ineffectiveness claim was considered and rejected by the County Court in adjudicating petitioner’s motion under New York Criminal Procedure Law § 440.10, this case is subject to the deferential standard of review of state court adjudications under the Antiterrorism and Effective Death Penalty Act (AEDPA), 28 U.S.C. § 2254(d). See Lindstadt v. Keane, 239 F.3d 191, 197-98 (2d Cir.2001). Therefore, the petition should have been granted only if the County Court’s decision was “contrary to, or involved an unreasonable application of, clearly established Federal law, as determined by the Supreme Court of the United States,” or “was based on an unreasonable determination of the facts in light of the evidence presented.” 28 U.S.C. § 2254(d). The “unreasonable application” language applies where the state court identified the correct governing legal principle in the Supreme Court’s decisions but applied it unreasonably. Williams v. Taylor, 529 U.S. 362, 413, 120 S.Ct. 1495, 146 L.Ed.2d 389 (2000); Torres v. Berbary, 340 F.3d 63, 68 (2d Cir.2003). This is the applicable provision here, because the Supreme Court has not confronted a set of facts materially indistinguishable from those in the instant case, which is the type of situation that would invoke the “contrary to” *607 language. See Williams, 529 U.S. at 412-13, 120 S.Ct. 1495; Torres, 340 F.3d at 68. Unreasonableness is determined by an “objective” standard. Williams, 529 U.S. at 409, 120 S.Ct. 1495. It is not sufficient that “a federal habeas court ... concludes in its independent judgment that the relevant state-court decision applied clearly established federal law erroneously or incorrectly.” Id. at 411, 120 S.Ct. 1495. Rather, to be “unreasonable,” the state court’s application of federal law must reflect “[sjome increment of incorrectness beyond error,” although that “increment need not be great.” Henry v. Poole, 409 F.3d 48, 68 (2d Cir.2005) (quoting Francis S. v. Stone, 221 F.3d 100, 111 (2d Cir.2000)). Where a petitioner challenges his conviction based on ineffective assistance of counsel, the question is whether the state court unreasonably applied Strickland v. Washington, 466 U.S. 668, 687, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984). See Williams, 529 U.S. at 390, 120 S.Ct. 1495. II. Ineffective Assistance of Counsel In order to prevail on a Sixth Amendment ineffectiveness claim, a petitioner must prove (1) that counsel’s representation “fell below an objective standard of reasonableness” measured under “prevailing professional norms,” Strickland, 466 U.S. at 688, 104 S.Ct. 2052, and (2) that “there is a reasonable probability that, but for counsel’s unprofessional errors, the result of the proceeding would have been different,” id. at 694, 104 S.Ct. 2052. “[Sjtrategic choices made by counsel after thorough investigation ... are virtually unchallengeable,” id. at 690, 104 S.Ct. 2052, and there is a strong presumption that counsel’s performance falls “within the wide range of reasonable professional assistance,” id. at 689-90, 104 S.Ct. 2052. But counsel has a duty to make reasonable investigations, and a decision not to investigate will be reasonable only “to the extent that reasonable professional judgments support the limitations on investigation.” Id. at 690-91, 104 S.Ct. 2052. In sexual abuse cases, because of the centrality of medical testimony, the failure to consult with or call a medical expert is often indicative of ineffective assistance of counsel. See Eze v. Senkowski, 321 F.3d 110, 127-28 (2d Cir.2003); Pavel v. Hollins, 261 F.3d 210, 224 (2d Cir.2001); Lindstadt, 239 F.3d at 201. 1 This is particularly so where the prosecution’s case, beyond the purported medical evidence of abuse, rests on the credibility of the alleged victim, as opposed to direct physical evidence such as DNA, or third party eyewitness testimony. See Eze, 321 F.3d at 128; Pavel, 261 F.3d at 224. 111. Petitioner’s Ineffectiveness Claim A. Defense Counsel’s Failure to Consult with or Call a Medical Expert or to Review or Challenge the Medical Evidence of Penetration Here, defense counsel failed to call as a witness, or even to consult in preparation for trial and cross-examination of the *608 prosecution’s witnesses, any medical expert on child sexual abuse. Counsel essentially conceded that the physical evidence was indicative of sexual penetration without conducting any investigation to determine whether this was the case. As Dr. Brown’s affidavit demonstrates, had counsel conducted such an investigation, counsel would likely have discovered that exceptionally qualified medical experts could be found who would testify that the prosecution’s physical evidence was not indicative of sexual penetration and provided no corroboration whatsoever of the alleged victim’s story. Counsel could thus have presented a strong affirmative case that the charged crime did not occur and the alleged victim’s story was incredible in its entirety. As the district court noted, “[i]n light of the daughter’s allegations of continuing rape and sodomy over a period of years, the absence of physical indicia of such abuse would necessarily have been troubling to a trier of fact.” 299 F.Supp.2d at 91. Dr. Brown testified that the photographic record of the alleged victim’s physical condition revealed no hyme-nal clefts, which would have been evidence of vaginal penetration, and Dr. Silecchia conceded that there was no indication of anal dilatation, which would have been evidence of anal penetration. Dr. Brown testified that all of the other physical observations by Dr. Silecchia that she relied on in concluding that penetration had taken place in fact provided no support for that conclusion. Such a lack of physical evidence of penetration would cast serious doubt on the alleged victim’s story of daily forcible penetration over a period of years. Were the trier of fact to credit expert testimony of the type petitioner offered in his state collateral attack and in the instant habeas action, the prosecution’s case would arguably amount to the uncorroborated and demonstrably incredible testimony of the alleged victim. The prosecution’s case rested centrally on the alleged victim’s testimony and its corroboration by the indirect physical evidence as interpreted by the medical expert. The medical expert testimony was central not only because it constituted the most extensive corroboration that any crime occurred, but because to undermine it would undermine the alleged victim’s credibility and thus the entire prosecution case as to all charges. If a medical examination of the alleged victim failed to reveal any evidence clinically indicative of sexual penetration, that failure would constitute strong affirmative evidence that forced sexual penetration did not oecur-particu-larly forced sexual penetration of the frequency alleged by petitioner’s daughter. As noted above, we have explained that medical expert consultation or testimony is particularly critical to an effective defense in sexual abuse cases where direct evidence is limited to the victim’s testimony. See Eze, 321 F.3d at 128; Pavel, 261 F.3d at 224; Lindstadt, 239 F.3d at 201. The situation may be different in a case where objective evidence exists implicating petitioner in a crime, such as bodily fluids identified as the petitioner’s, or where the prosecution offered third party eyewitness testimony. But in a case where the only direct evidence that any crime occurred or that, if it did, the petitioner committed it, was the testimony of the alleged victim, for defense counsel to simply concede the medical evidence without any investigation into whether it could be challenged was performance that the state court could not reasonably find to be objectively reasonable. As we stated in Eze: A lesson to be learned from Lindstadt and Pavel is that when a defendant is accused of sexually abusing a child and the evidence is such that the case will turn on accepting one party’s word over the other’s, the need for defense counsel to, at a minimum, consult with an expert *609 to become educated about the Vagaries of abuse indicia’ is critical. The importance of consultation and pre-trial investigation is heightened where, as here, the physical evidence is less than conclusive and open to interpretation. 821 F.3d at 128 (internal citations omitted). As the district court correctly noted, “[t]here is no per se rule that requires trial attorneys to seek out an expert.” 299 F.Supp.2d at 100-01. We do not even mean to hold that expert consultation is always necessary in order to provide effective assistance of counsel in child sexual abuse eases-we need not address the issue in such generality, but do note the Supreme Court’s cautionary observation that “[a] standard of reasonableness applied as if one stood in counsel’s shoes spawns few hard-edged rules .... ” Rompilla v. Beard, - U.S. -, 125 S.Ct. 2456, 2462, 162 L.Ed.2d 360 (2005). However, it is clear that in this case such a failure was not justified as an objectively reasonable strategic choice. Here, no facts known to defense counsel at the time that he adopted a trial strategy that involved conceding the medical evidence could justify that concession. Particularly troubling in this case is counsel’s failure to examine prior to trial the eolposcopic slides that were made to record the physical evidence of trauma purportedly observed by Dr. Silecchia. As a direct photographic record of the trauma, or lack thereof, to the alleged victim, it would be difficult to exaggerate the significance of these slides to petitioner’s case. Defense counsel’s apparent failure even to request to examine them was a serious dereliction of his duty to investigate the facts and circumstances of petitioner’s case. See ABA Standards for Criminal Justice 4-4.1 (3d ed. 1993) (“Defense counsel should conduct a prompt investigation of the circumstances of the case and explore all avenues leading to facts relevant to the merits of the case .... The investigation should include efforts to secure information in the possession of the prosecution and law enforcement authorities.”); see also Rompilla, 125 S.Ct. at 2466 (finding ABA Standards useful “guides to determining what is reasonable”) (quoting Wiggins v. Smith, 539 U.S. 510, 524, 123 S.Ct. 2527, 156 L.Ed.2d 471 (2003)). Appellant contends that defense counsel made a reasonable strategic decision based on the information known to him at the time, and based in part on information provided by petitioner, that challenging the medical evidence would be futile, and that therefore the most effective defense would be to suggest that the physical evidence of penetration could be explained by consensual sexual conduct between the victim and her ex-boyfriend. The County Court, similarly, concluded that defense counsel made a tactical decision, instead of mounting any challenge to the medical or psychological evidence, to focus on obtaining concessions that the medical and psychological evidence was consistent with innocence. And the County Judge continued to maintain in his affidavit before the district court not only that such a tactical decision was permissible, but that it ultimately blossomed into an effective cross-examination of the prosecution’s experts. Appellant argues that the County Court’s findings were reasonable applications of Strickland. In considering this argument, we judge the reasonableness of the purported “strategic decision” on the part of defense counsel “in terms of the adequacy of the investigations supporting” it. Wiggins, 539 U.S. at 521, 123 S.Ct. 2527; accord Strickland, 466 U.S. at 690-91, 104 S.Ct. 2052. Appellant’s argument fails because defense counsel decided that it would be futile to challenge the medical and psychological *610 evidence without having reasonably investigated whether that was in fact the case, and lacked sufficient information reasonably to determine that such an investigation was unnecessary. Thus, this is not a case where counsel made a reasonable decision to cease further investigation as a result of having “discovered... evidence ... to suggest that” challenging the prosecution’s medical or psychological evidence “would have been counterproductive, or that further investigation would have been fruitless.” Wiggins, 539 U.S. at 525, 123 S.Ct. 2527. Nor is this a case of “diligent counsel ... drawing] a line when they have good reason to think further investigation would be a waste.” Rompilla, 125 S.Ct. at 2463. Rather, counsel here never discovered any evidence to suggest one way or another whether such challenges would be counterproductive or such investigation fruitless, nor did counsel have any reasonable basis to conclude that such investigation would be wasteful. This was because counsel settled on a defense theory and cut off further investigation of other theories without having first conducted any investigation whatsoever into the possibility of challenging the prosecution’s medical or psychological evidence. Because counsel never investigated that alternative approach at all, counsel did not have any reasoned basis to conclude that such an approach would be fruitless-and, in fact, subsequent counsel have found it quite fruitful. It was deficient performance on counsel’s part to adopt a weak position that evidence was “consistent with innocence” without having first made any attempt to investigate whether a case could be made that the evidence was not consistent with guilt. As it turned out, petitioner’s counsel was unable to present any evidence that the victim engaged in consensual sexual conduct involving penetration. Counsel thus failed to present any explanation for the evidence other than petitioner’s guilt. This was not the result of a reasonable strategic decision, but only of counsel’s failure to conduct an adequate pre-trial investigation of the evidence. Similarly, petitioner’s counsel did not make a reasonable strategic choice by deciding to forego any investigation of the medical evidence in favor of relying on his unsupported assumption that the victim would be unwilling or unavailable to testify. Even granting, for the sake of argument, that counsel was entitled to credit, without any independent investigation, and with no opportunity for any, his client’s hearsay information that the victim had a consensual sexual partner, counsel was not entitled to assume, without conducting a reasonable investigation, including consulting with someone capable of interpreting the medical evidence, that the medical evidence was indicative of or consistent with forced sexual penetration. Even assuming counsel’s knowledge that the alleged victim had, as a matter of fact, been penetrated, this could not excuse counsel’s complete failure to investigate whether the prosecution would be capable of proving that penetration had occurred. Knowledge of such penetration is not equivalent to knowledge that an investigation of the evidence of penetration would be fruitless. Defense counsel may not fail to conduct an investigation and then rely on the resulting ignorance to excuse his failure to explore a strategy that would likely have yielded exculpatory evidence. For counsel to forego even an investigation of the possibility of challenging the physical evidence, in favor of an unsupported theory that someone other than the defendant penetrated the victim, or an unsupported assumption that the victim would not testify, was not an objectively reasonable performance. There was nothing strategic about a decision to concede the physical evidence, with no educated basis for doing *611 so, in favor of uninvestigated and uninves-tigatable theories about the victim’s relationship with her boyfriend and her ability or willingness to testify against her own father. As the district court noted, failing to present exculpatory evidence is not a reasonable trial strategy. 299 F.Supp.2d at 104. The County Court’s holding otherwise, which was not supported by any analysis or argument to justify a conclusion that counsel’s decision to settle on a defense without having conducted an adequate pretrial investigation of possible alternative strategies and defenses was objectively reasonable, was an unreasonable application of Strickland. B. Defense Counsel’s Failure to Consult with or Call a Psychological Expert As the district court noted, 299 F.Supp.2d at 105, the prosecution offered its psychological expert testimony in order to bolster the alleged victim’s credibility, primarily by explaining her failure to reveal the abuse earlier and her inability to provide a consistent and detailed account of the abuse. Defense counsel failed to consult or call an expert on the psychology of child sexual abuse, or to educate himself sufficiently on the scientific issues. Therefore he was unable to mount an effective cross-examination, and missed an opportunity to rebut this attempt at bolstering the alleged victim’s credibility. The prosecution’s entire case rested on the credibility of the alleged victim’s testimony. Indeed, the County Judge in his affidavit affirms that he also viewed the entire case as one of testing the credibility of the complaining witness. It would appear from Dr. Yuille’s affidavit that even a minimal amount of investigation into the purported “Child Sexual Abuse Accommodation Syndrome” would have revealed that it lacked any scientific validity for the purpose for which the prosecution utilized it: as a generalized explanation of children’s reactions to sexual abuse, including delayed disclosure and blurred memory. Similarly, it would appear that had counsel investigated the possibility of challenging the prosecution’s psychological expert, he would have discovered that exceptionally qualified experts could be found who would challenge the scientific validity of the prosecution expert’s other theories about, for example, adolescence prompting disclosure of sexual abuse. Defense counsel’s lack of preparation and failure to challenge the credibility of the key prosecution witness could not be based on a sound trial strategy, and it was an unreasonable application of Strickland for the County Court to hold otherwise. C. Prejudice As noted, under the prejudice prong of Strickland, petitioner was required to show that “there is a reasonable probability that, but for counsel’s unprofessional errors, the result of the proceeding would have been different,” Strickland, 466 U.S. at 694, 104 S.Ct. 2052. “A reasonable probability is a probability sufficient to undermine confidence in the outcome.” Id. In evaluating prejudice, we look to the cumulative effect of all of counsel’s unprofessional errors. Lindstadt, 239 F.3d at 204. We must keep in mind that “a verdict or conclusion only weakly supported by the record is more likely to have been affected by [counsel’s] errors.” Strickland, 466 U.S. at 696, 104 S.Ct. 2052. Conversely, where there is overwhelming evidence of guilt, even serious errors by counsel will not warrant granting a writ of habeas corpus. See Lindstadt, 239 F.3d at 204. Again, it must be noted that the prosecution’s entire case rested on the credibility of the alleged victim. All other evidence presented by the prosecution was indirect evidence offered to corroborate aspects of the alleged victim’s story. Defense counsel’s failure to investigate the *612 prosecution’s evidence led him to decide not to challenge what was clearly the most significant corroborative evidence-the medical expert testimony that the physical condition of the alleged victim supported a conclusion that penetration had taken place. Counsel’s decision not to consult with or call an expert precluded counsel from offering a potentially persuasive affirmative argument that the alleged victim’s condition was not indicative of or consistent with forced sexual penetration. Counsel’s failure also prevented him from challenging the psychological evidence offered to explain the victim’s delay in coming forward and failure to recall events in detail as the result of something other than a lack of credibility-which would as a general matter be a common inference to draw from such shortcomings. The County Court’s finding that the medical expert testimony was insignificant to the outcome of the trial and had been effectively neutralized by cross-examination, and that there was thus not a “reasonable probability” that counsel’s failures affected the ultimate result, was unreasonable. Had counsel investigated and discovered the expert witnesses that petitioner’s current counsel has discovered, and called them to testify, and had they been credited, there is a reasonable probability that the trier of fact would have rejected the entirety of the alleged victim’s narrative as not credible, because her assertion that petitioner penetrated her almost daily over a period of years would be inconsistent with the lack of any medical evidence of penetration. There is also a reasonable probability that the trial court would credit such expert testimony offered by petitioner. The County Judge’s affidavit is neither dispositive nor more than minimally probative on this issue, and we do not find that it casts doubt on our finding of prejudice. As a preliminary matter we note our doubt as to whether the affidavit is entitled to any evidentiary weight whatsoever, in light of petitioner’s lack of opportunity to test the County Judge’s testimony through cross-examination. See Hollis v. Smith, 571 F.2d 685, 694 (2d Cir.1978). In Hollis we found that “[i]f inquiry was to be made in 1977 with respect to the state judge’s mental processes in 1969, this should have been done in open court with Hollis having the right of cross-examination, not by an affidavit solicited by the State .... ” Id We thus “disregarded] the state judge’s affidavit.” Id Similarly here, the County Judge’s 2003 affidavit regarding his mental processes in 1999 is arguably entitled to no weight as evidence of those processes in light of the lack of opportunity for cross-examination at which petitioner’s counsel could have solicited the Judge’s explanation of how he reconciled his unflappable belief in the alleged victim’s credibility with a complete lack of any objective evidence to support her story. Even were we to consider the affidavit as evidence of the County Judge’s thought process, the Judge’s perception after the fact of his thought process at the time of trial, rendered after he had already become committed to a position by rendering a guilty verdict, would be entitled to minimal evi-dentiary weight as proof of his actual thought process four years earlier. Even were we to consider the affidavit as more than minimally probative of the County Judge’s thought process at the time of trial, we would still find the County Judge’s apparent finding in denying petitioner’s Article 440 motion that petitioner was not prejudiced by counsel’s errors to be an unreasonable finding of fact in light of the record, and we find the Judge’s statements in his affidavit to be irrelevant to our analysis of this question. As the district court correctly noted, “[i]n light of the daughter’s allegations of continuing rape and sodomy over a period of years, *613 the absence of physical indicia of such abuse would necessarily have been troubling to a trier of fact.” 299 F.Supp.2d at 91. The County Judge’s affidavit says nothing that addresses this question. The Judge’s statement that petitioner’s new expert testimony “did not undermine the testimony of the victim in this case” misses the point that the expert testimony, while not itself an attack on the victim’s credibility, would have made possible a potentially persuasive attack based on the absence of physical indicia of abuse that a reasonable trier of fact would expect to be present in light of the alleged victim’s story. The County Judge’s other statements-that his guilty verdict was based exclusively on the alleged victim’s testimony, which he believed at the time beyond any doubt, and not at all on the medical or psychological evidence, which he found to be neutralized by defense counsel’s cross-examination-whether or not objectively persuasive, are also irrelevant as evidence of the County Judge’s subjective state of mind. Whether or not the County Judge disregarded the prosecution’s affirmative medical evidence of penetration, he was not presented at trial with an affirmative argument, supported by medical expert testimony, that the alleged victim’s story was contradicted by the objective physical evidence. It is the effect of such an argument on the trier of fact that is relevant to prejudice, and the County Judge’s affidavit does not address such an argument. Respondent characterizes the evidence against petitioner as “overwhelming.” On the contrary, the evidence presented by the prosecution to establish petitioner’s guilt here was no stronger than it was in Lindstadt, which we characterized as “a case of underwhelming evidence.” 239 F.3d at 205. In Lindstadt, as here, the prosecution presented the victim’s eyewitness testimony, corroborative testimony from the victim’s mother and the medical expert who examined her, and bolstering testimony from an expert on child psychology. Id. at 195. 2 In Lindstadt, as here, the only direct evidence against petitioner consisted of the eyewitness testimony of the victim, which was not corroborated by testimony from any other eyewitnesses. Here, other than the prosecution’s questionable medical expert testimony, the prosecution offered no objective evidence to support an inference that any crime took place at all, and presented no physical evidence linking petitioner to any crime that occurred. No semen or other bodily fluids were recovered that could be connected to petitioner. Rather, all of the prosecution’s evidence, with the exception of the alleged victim’s testimony and the prosecution’s questionable medical expert testimony-such as the mother’s testimony about the victim’s emotional outburst on revealing the abuse, and the psychological expert’s explanation of the victim’s delay and lack of detailed memory-was only indirectly and weakly corroborative of the victim’s story or was offered to bolster her credibility. Not only was the evidence against petitioner relatively thin, but most of it could have been, but was not, effectively challenged by defense experts or an informed cross-examination. The victim’s credibility, the psychological expert’s bolstering, and the medical expert testimony concluding that penetration had taken place, all could have been seriously undermined had petitioner’s counsel offered the expert testimony that was available but that he failed to discover. As noted, where the record evidence in support of a guilty verdict is thin, as it is here, there is more likely to be prejudice. This is even more true where counsel’s failures go to *614 something as important as the medical evidence in this case-the only objective evidence that a crime occurred and the only evidence directly corroborating any aspect of the victim’s story. See Lindstadt, 239 F.3d at 204. Here, as in Lindstadt, defense counsel failed to satisfy the requirement of objectively reasonable performance by failing to investigate critical prosecution evidence. Counsel was consequently unable and unprepared to challenge, either with expert testimony or an effective cross-examination, the key evidence of petitioner’s guilt and the only evidence that any crime had taken place at all. In a case where the overall body of evidence against petitioner was “underwhelming,” as in Lindstadt, petitioner was prejudiced by counsel’s errors, and the County Court’s finding that he was not prejudiced was an unreasonable application of Strickland. See Lindstadt, 239 F.3d at 205. IV. Spillover Respondent argues that even if counsel was ineffective in failing to investigate the medical expert testimony, because that testimony only concerned evidence that the alleged victim was penetrated, petitioner was only prejudiced as to two of the six sodomy counts, and the district court erred in granting a writ of habeas corpus as to the convictions on the other sodomy counts and the counts of sexual abuse and endangering child welfare. Respondent argues, similarly, that even if counsel was ineffective in failing to investigate the psychological expert testimony, because the only purpose served by that testimony was to explain the alleged victim’s delay in reporting the abuse, petitioner was not prejudiced by that failure as to the two convictions for sexual abuse, which related to incidents that allegedly occurred only weeks before she revealed them. In evaluating the “spillover” effect of counsel’s errors, we generally consider three factors: (1) whether the evidence on the vacated counts was inflammatory and likely to inflame the jury; (2) whether the evidence on the vacated counts was similar to that required to prove the remaining counts; and (3) the strength of the prosecution’s case on the remaining counts. Lindstadt, 239 F.3d at 205 (quoting United States v. Naiman, 211 F.3d 40, 50 (2d Cir.2000)). Respondent suggests that the inflammatory nature of the evidence is not relevant in a bench trial. We do not necessarily agree, because a bench trial is merely a jury trial before one rather than twelve. However, we need not reach the question of whether judges are impervious to inflammatory evidence because, based on the other two factors, we find that counsel’s errors spilled over and prejudiced petitioner as to all counts. First, as to the strength of the prosecution case: In a case where the evidence against petitioner was generally “underwhelming,” the counts that respondent argues were insulated from prejudice were the weakest of all. See Lindstadt, 239 F.3d at 206 (“Count 5 was the weakest of the lot; it relied solely on the word of Lindstadt’s wife and daughter, and was not buttressed by physical evidence or expert testimony of any kind.”). Second, regarding the similarity of the evidence: The primary evidence as to all counts was not only similar, it was the same-the alleged victim’s testimony. Contrary to respondent’s argument, it is not true that counsel’s errors impacted only the evidence supporting penetration and explaining the victim’s delay in reporting. Counsel’s errors resulted in counsel foregoing presentation of evidence that could have cast serious doubt on the veracity of *615 the alleged victim’s testimony in its entirety. Counsel’s failure to investigate the medical evidence of penetration resulted in counsel’s inability to present expert testimony, or at least to suggest on cross-examination, that the medical evidence was not indicative of penetration. A natural inference from the contrary medical expert testimony presented by petitioner on habe-as, if it is credited, is that no penetration took place, and that the alleged victim was lying or mistaken in her testimony that she was penetrated. The latter inference infects all of the counts of which petitioner was convicted. The alleged victim’s entire story is brought into doubt if it is demonstrated that no penetration took place. The trier of fact clearly relied on a finding of the alleged victim’s credibility in reaching convictions for all of the counts, not only the ones involving penetration. The credibility finding clearly relied on the fact that the alleged victim’s testimony was consistent with that of the prosecution’s medical expert, and would be undermined if that testimony was rejected and the alleged victim was found to have lied or been mistaken with regard to penetration. Counsel’s errors therefore infected all of petitioner’s convictions. See id. (“[T]o the extent that the evidence differs, there is less to support Count 5, and all of it is infected by the same errors affecting the other counts.”). CONCLUSION For the foregoing reasons, the district court’s order granting petitioner-appellee Ben Gersten’s petition for a writ of habeas corpus, and directing that petitioner be released from custody unless new state criminal proceedings against him are commenced within sixty days, is affirmed. Footnotes [1] Respondent argues that the district court erred in relying on these cases because under the AEDPA deferential standard of review, a writ of habeas corpus cannot be granted based on a state court's unreasonable application of precedents of the Court of Appeals. But in Eze and Lindstadt we held that New York courts unreasonably applied Strickland in finding counsel's performance effective. See Eze, 321 F.3d at 124-38; Lindstadt, 239 F.3d at 197-98, 205. The district court was bound to apply this Court's precedents governing when applications of Strickland are "unreasonable,” and in doing so did not violate, but rather, effectuated the AEDPA standard of review. While Pavel did not apply the deferential AEDPA standard, 261 F.3d at 215-16, it elaborated on the principles explicated in Lindstadt, see Pavel, 261 F.3d at 221-25 & n. 15. [2] In fact, the psychological expert testimony at issue in Lindstadt was given by the same expert that testified for the prosecution in the instant case, Dr. Lewittes. 239 F.3d at 195.