Crawford v Washington (541 US 36 [2004]) is a fresh precedent, the contours and limits of which are still indistinct. At this early stage, it is difficult to predict whether the Supreme Court will apply Crawford—universally or in some cases or with limitations or at all—to hearsay used as a basis for expert testimony, much less the exact implications of any such holding. Still, the majority’s analysis of this case in relation to Crawford is reasonable, and I do not disagree with it. Further, I wholeheartedly agree with section II of the majority opinion, which points out a New York law issue that is, in my view, significant, but which we need not reach here. I respectfully dissent, however, because, even assuming that Crawford precludes admission of the four remarks made by interviewees to the People’s expert, I cannot agree that defendant suffered any harm as a result.
As an initial matter, we should keep in mind that this trial presented no issue regarding guilt per se. Defendant indisputably killed Kendra Webdale by thrusting her into the path of an oncoming subway train. The only issue at trial was whether defendant had established, by a preponderance of the evidence, his affirmative defense that he should not be held criminally responsible for his conduct because he suffered from a mental disease or defect at the time of the killing (see Penal Law § 40.15). Specifically, defendant wanted the jurors to accept that he killed Ms. Webdale while in the throes of what his primary *133expert, Dr. Spencer Eth, during direct examination called a “sudden psychotic act.”
According to Dr. Eth, defendant murdered Ms. Webdale “when he was suffering an acute exacerbation of sudden intensification of severe psychotic symptoms and his brain was not functioning[, which meant that] his motor control—he could walk, he could thrust out his arms, he could see, but he couldn’t think, he couldn’t plan, he couldn’t intend, he couldn’t know as we understand what know means what he was doing or that it was wrong.” During cross-examination, Dr. Eth denied describing the incident as a “sudden psychotic moment,” and declined to characterize it as a “sudden psychotic attack,” although he essentially had testified to this effect. As an alternative description of his diagnosis, he offered that, at the moment of the killing, “a ferocious torrence [sic] of symptoms overwhelmed [defendant’s] mind,” thereby precluding him from planning or “executing] an action with reason and intent.”
Another of defendant’s experts, Dr. Wilfred Van Gorp, provided similar testimony, telling the jurors that he agreed with Dr. Eth’s conclusion that defendant suffered a “transient episode of extreme psychotic symptomology that destroyed his capacity to appreciate the nature and consequences of his conduct and to appreciate that his conduct was wrong.” Both experts opined that the “transient episode” essentially began directly before and ended almost immediately after the killing—a “symptomatic exacerbation while on the station platform.”
But defendant’s expert testimony, which hinged on defendant’s supposed inability to intend or plan, ran directly counter to the prior testimony of witnesses to the crime, who described how defendant had engaged in seemingly meticulous planning. According to eyewitness testimony, before approaching the blond Ms. Webdale and immediately after being rebuffed by another blond woman, defendant walked to the front of the subway platform, the best vantage point from which to see a train entering the curved station. Bending over, defendant peered up the tracks, concededly looking for the headlights of an oncoming train. He then walked back towards Ms. Webdale, who was standing near the platform’s edge, and asked her the time. Following Ms. Webdale’s response, defendant positioned himself directly behind her, standing with his back to the wall of the station. Notably, he chose to stand behind Ms. Webdale rather than the other blond woman, who was taller and heavier. As the *134train proceeded through the station, defendant rushed forward and shoved Ms. Webdale at the precise moment when she would pitch headlong into the train’s path without any chance to save herself or be rescued. Despite having moved with great force (he pushed off from the back wall), defendant had sufficient presence of mind to break his momentum and avoid sharing Ms. Webdale’s fate by twisting his body away from the platform’s edge. As the trial prosecutor told the jurors during summation, “actions speak louder than words.”
This description of events by the People’s fact witnesses, which defendant did not challenge, contradicted the defense experts’ assertions that a theorized “acute exacerbation” prevented defendant from planning or intending. Another facet of the attack also undermined defendant’s supposed inability to comprehend that his conduct was wrong. Immediately after killing Ms. Webdale, he announced “I’m sick,” and asked to be taken to a hospital, which displayed his understanding that he indeed had done something wrong and needed an excuse to negate his blameworthiness.
In addition to the facts of the murder, the People also refuted defendant’s affirmative defense with the testimony of their rebuttal experts. These experts attacked the legitimacy of the defense theory that defendant had acted while experiencing a fleeting psychotic disorder or “transient episode,” and they did so independently of the four challenged observations. Specifically, Dr. Angela Hegarty explained to the jurors that, according to the Diagnostic and Statistical Manual of Mental Disorders (DSM), the standard classification of mental disorders used by mental health professionals in the United States, psychotic symptoms simply do not rapidly appear and disappear as defendant’s supposedly did while he stood on the subway platform. Rather, the DSM makes clear that the shortest duration for a brief psychotic disorder is one day. Dr. William Bryon Barr, the People’s other expert, offered similar testimony.
Although defendant adduced expert testimony that consumed thousands of pages of trial transcript (as did the People), that, standing alone, does not compel a conclusion that he proffered such a strong case that the error here could not have been harmless. As we have noted,
“before constitutional error . . . may be found to be harmless, it is not necessary that the untainted evidence on which the verdict in the case must be sup*135ported demonstrate undisputable guilt. Rather, the reasonable doubt standard, extremely high though it is, still leaves room for judgmental determination of harmlessness” (People v Schaeffer, 56 NY2d 448, 455 [1982]).
In other words, a court engaging in harmless error review should center its analysis not on the quantity of the evidence adduced, but rather on its quality (see id. [“because consideration of whether an error is harmless requires an evaluation not only of the tainted matter, but of the strength of the case absent the taint, the court must focus on the reliability and persuasiveness of the untainted matter and its source. ... In short, neither side of the evidentiary equation may be ignored; in the end, the picture must be seen as a whole”]).
Significantly, the trial court instructed the jurors to engage in a similar inquiry, explaining that defendant had to prove that he “was not criminally responsible” by “a preponderance of the credible evidence [, which] is evidence which you find worthy of belief.” The court amplified its instruction by stating that defendant would meet his burden if “you, the jury, are satisfied that the evidence of lack of criminal responsibility, from whatever source, outweighs and is more convincing than the evidence that he was criminally responsible when he committed the crime.” Finally, the jurors were informed that “[a]s with any other factual issue, it is the quality of the evidence which controls, not the number of witnesses on one side or the other.” Ultimately, defendant’s criminal responsibility does not seem to have presented a close case in the minds of the jurors. They deliberated for no more than two hours, during which they apparently lunched as well. The jurors sent no notes. Clearly, no problematic issues arose during these brief deliberations. I find it impossible to believe that in those two hours, the jurors were only able to dismiss defendant’s “transient episode” theory by focusing on the four isolated comments cited by the majority. Far more likely, the jurors readily rejected what they must have viewed as a rather outlandish defense theory, unsupported as it was in fact or professional literature.
Evaluating the “importance” of the four comments to the case, however, the majority concludes that the People “drew some significant support” from them (majority op at 129, 130). But even assuming that the jurors, like the majority, zeroed in on these four comments, they were generally duplicated or corroborated by other evidence in the record, thereby diminishing *136their hypothesized potential significance to the jurors’ deliberations. For example, the majority asserts that the record contains “no clearer proof’ of defendant’s “habit of announcing his mental illness the moment he got in trouble” (majority op at 130) than John E’s statement that, after attacking a woman at a Waldbaum’s supermarket, defendant repeated “I’m sick, I’m sick, I’m schizophrenic.” As the majority acknowledges, however, there is evidence in the record that defendant made similar statements following other instances of aggressive behavior. And John E’s statement is not disputed. Defendant’s primary expert, Dr. Eth, acknowledged on cross-examination that defendant announced to John E that he was sick and a psychiatric patient, and begged him not to call the police. As for Kimberly D.’s apparent* observation that Kendra Webdale bore a “rather remarkable” resemblance to another woman, Stephanie H., who had teased defendant, the jury heard from another witness that both women were blond, and that Stephanie H. had teased defendant. Moreover, as the majority points out, Stephanie H.’s and Kendra Webdale’s photographs were admitted into evidence. As a result, the jurors were free to make their own judgment about any resemblance between the two women, assuming they found this important, and either to credit or discount Dr. Hegarty’s testimony accordingly. It is also worth noting that Dr. Hegarty testified that while interviewing defendant, she mentioned Stephanie H. inadvertently, and defendant became visibly sexually aroused. This speaks far more powerfully to “inappropriate behavior that reflected defendant’s sexual frustration” (majority op at 131) than does the third chal*137lenged observation by Serita G. that defendant, naked on a bed, did not cover up when her maid entered his room. Finally, Issac V’s statement that, close in time to the killing, defendant appeared “a little weird,” is, if anything, innocuous. True, the People’s experts opined that defendant had a “relatively mild” mental disorder. But defendant’s “transient episode” theory also called for him to act relatively normally immediately preceding the attack.
In short, I see no possibility that the four hearsay comments caused the jurors to reject defendant’s affirmative defense. Rather, his defense was subverted by the incredible nature of his psychiatric theory coupled with his uncontested actions, which contradicted any “transient” loss of control or comprehension. Because any Crawford error that occurred here was harmless beyond a reasonable doubt (see Chapman v California, 386 US 18, 24 [1967]; People v Crimmins, 36 NY2d 230, 240-241 [1975]), defendant’s conviction should be affirmed.
Chief Judge Kaye and Judges G.B. Smith, Ciparick, Rosenblatt and Graffeo concur with Judge R.S. Smith; Judge Read dissents and votes to affirm in a separate opinion.
Order reversed, etc.
From the transcript, it is not entirely clear to me whether Dr. Hegarty was testifying that Kimberly D. told her that Stephanie H. and Kendra Web-dale resembled each other, or was offering her own view on this subject. After Dr. Hegarty testified that Kimberly D. told her that Stephanie H. teased defendant, defense counsel objected based on the “right to confrontation,” and the trial court overruled the objection. Then the prosecutor asked Dr. Hegarty two questions about Stephanie H. (her last name and occupation). After Dr. Hegarty answered these two questions, the prosecutor asked her “And did you learn anything about [Stephanie H.’s] general appearance?” to which Dr. Hegarty responded “Yes. She bears a rather remarkable similarity and appearance to Kendra Webdale.” In context, the prosecutor might having been asking Dr. Hegarty if she had learned this information about Stephanie H.’s general appearance from Kimberly D., but this is far from certain, especially since Dr. Hegarty surely had access to the photographs of both women. Defense counsel did not object to the question or answer about Stephanie H.’s “general appearance,” which were followed by Dr. Hegarty’s testimony about defendant’s reaction when she inadvertently mentioned Stephanie H. to him during an interview.