Appeal by the defendant from a judgment of the Supreme Court, Queens County (Hollie, J.), rendered January 6, 2009, convicting him of robbery in the second degree (two counts) and robbery in the third degree, upon a jury verdict, and imposing sentence.
Ordered that the judgment is reversed, on the law and in the exercise of discretion, and a new trial is ordered.
The defendant allegedly stole the victim’s cell phone in a park *784in Queens. He was identified by the victim at a police-arranged showup less than 30 minutes after the alleged robbery occurred, approximately three blocks from the park. The victim was the only eyewitness who identified the defendant.
The Supreme Court improvidently exercised its discretion in denying the defendant’s motion in limine to present at trial the testimony of a professor of psychology as an expert witness with respect to several factors that might affect the accuracy of an identification. “[W]here the case turns on the accuracy of eyewitness identifications and there is little or no corroborating evidence connecting the defendant to the crime, it is an abuse of discretion for a trial court to exclude expert testimony on the reliability of eyewitness identifications if that testimony is (1) relevant to the witness’s identification of the defendant, (2) based on principles that are generally accepted within the relevant scientific community, (3) proffered by a qualified expert, and (4) on a topic beyond the ken of the average juror” (People v LeGrand, 8 NY3d 449, 452 [2007]; see People v Santiago, 17 NY3d 661, 669 [2011]). Here, there was little or no corroborating evidence connecting the defendant to the crime. The fact that the victim was confronted by the defendant on a clear, sunny day, and had an unobstructed view of the defendant at close range, does not constitute corroborating evidence of the identification for purposes of determining whether expert testimony regarding the accuracy of an eyewitness identification is admissible (see People v Cordes, 71 AD3d 912, 913 [2010]; cf. People v Zohri, 82 AD3d 493, 494 [2011]; People v Fernandez, 78 AD3d 726, 727 [2010]; People v Austin, 46 AD3d 195, 200 [2007]). Further, the testimony of the police officer who responded to the scene of the incident, as well as that of the police officer who stopped the defendant shortly after a radio transmission was sent out describing the assailant, failed to harmonize with the victim’s testimony in such a manner as to furnish the necessary connection between the defendant and the crime (cf. People v Novak, 212 AD2d 740, 741 [1995]; People v Keindl, 117 AD2d 679 [1986]).
Moreover, the proposed expert testimony was (1) relevant to the victim’s identification of the defendant, (2) based on principles that are generally accepted within the relevant scientific community, (3) proffered by a qualified expert, and (4) on a topic beyond the ken of an average juror (see People v Santiago, 17 NY3d at 672; People v Abney, 13 NY3d 251, 268 [2009]; People v LeGrand, 8 NY3d at 452). Thus, the Supreme Court should have granted the defendant’s motion in limine to present the proposed expert testimony at trial (cf. People v Muhammad, 17 NY3d 532, 546 [2011]).
*785The Supreme Court’s errors were not harmless. “Trial error is only harmless when there is overwhelming proof of the defendant’s guilt and no significant probability that the jury would have acquitted the defendant were it not for the error” (People v Santiago, 17 NY3d at 673-674; see People v Crimmins, 36 NY2d 230, 242 [1975]). Here, the proof of the defendant’s guilt was not overwhelming, and thus, the error was not harmless.
In light of our determination, the defendant’s remaining contentions regarding the Supreme Court’s alleged failure to suspend jury deliberations, and those concerning alleged improper remarks made by the prosecutor on summation, have been rendered academic.
Accordingly, we reverse the judgment and order a new trial. Eng, EJ., Mastro, Florio and Sgroi, JJ., concur.