Opinion
ROTH, Judge:[ 1 Jimmy D. Guard appeals from his conviction for child kidnapping. Guard asserts that the trial court abused its discretion when it excluded his expert's testimony on the reliability of eyewitness identification from trial, We vacate the conviction and remand for a new trial.
BACKGROUND
1 2 Guard's conviction for child kidnapping resulted from the following evidence. Between 3:00 and 4:00 pm. on November 15, 2004, a nine-year-old child was grabbed from behind while she was walking from the school bus to her home. The abductor put his hand over the child's mouth and told her that he had a knife. The child kicked the abductor in the shin, causing him to release her. The child then turned around and poked the abductor in the eye repeatedly for about twenty seconds. The child and the abductor fought before the child broke free and ran home. While she was running, the child turned back onee to see the abductor running in the opposite direction.
[ 3 When the child reached her home, she reported the incident to her mother. The child and her mother went out to look for the abductor, and when they could not find him, the mother called the police. First an officer, and later a detective, responded to the call,. The child described the abductor to the officer as a male who was wearing white shoes, jeans, and a hat with curly hair sticking out from underneath. She also told the officer that the abductor could have been Hispanic and had a shadow of a mustache or beard but that she did not get a good look at his face. At that point, the detective, who had more experience and training in interviewing children, took over and was able to elicit from the child that the abductor was tall and slightly chubby with a dark complexion and dark hair. When pressed on how tall he was, the child stated that he was taller than the officer but shorter than the detective, a range of approximately 57" to 6'1" tall. The child also gave a more specific description of the abductor's clothing, describing the hat as a black baseball cap with the letter "A" on it and stating that he wore tennis shoes and a black "Stone Cold" Steve Austin T-shirt. The detective asked the child if she would be able to recognize the *446abductor if she saw him again, and the child said that she would. At trial, the child provided much the same description of her abductor as she did to the police, with the exception of the shoes. On direct examination, she said that she told the police the shoes were blue, but on eross-examination, she clarified that his shoes were white. A schoolmate, who got off the bus with the child but was not walking with her, corroborated the child's account of the abduction and her description of the abductor's clothing. According to the schoolmate, a stranger grabbed the child but then let go, and the child then ran one direction, and the stranger ran the other way. The schoolmate could not identify the stranger, but she did see that he was wearing a hat, blue pants, a black shirt, and white shoes.
4 The day after the kidnapping, the detective brought to the child's school a six-photograph lineup, which included a photograph of Guard.1 While all six men had short dark hair and dark complexions and were a little overweight, Guard was the only suspect who had curly hair. Guard had no facial hair in his photograph, while four of the others did. The detective showed the child each photograph individually and asked her "to look at each one of them and tell [him] whether or not any of these people [was] the guy that tried to kidnap her." Although the detective did not tell the child that the abductor was in one of the six photographs, the child later testified that she understood that one of the photographs would be of the person who kidnapped her. Guard's photograph was the third one. The child eliminated the first two photographs, saying to each, "[NJo, that's not him." When the child was shown the third photograph-Guard's picture-though, the detective testified that "[hler eyes got big, she appeared excited and seared at the same time and she immediately said, 'That's him. - That's him. ... Yes, I'm sure that's him.'" The child also testified that she "was a hundred percent" sure that the person in the photograph was her abductor and she told the detective "that was the person ... 'for sure"" The child answered in the negative when asked about whether anyone in the remaining three photographs was the abductor.
15 Following the child's identification of Guard, the police obtained a search warrant for Guard's residence and began looking for additional witnesses. The search of Guard's home did not yield any of the clothing that the child described, but the police did find a pair of light blue running shoes. The police also located two people in the child's neighborhood who, after being shown Guard's picture, said they had seen a man who looked like Guard in the neighborhood on the day of the kidnapping. One neighbor reported that she "thought [she] had seen" Guard run past her house between 3:15 and 8:45 p.m. while she was in her yard waiting for her children to return home from school. She had found this behavior odd because the man was not wearing running clothes and because her neighborhood is not very popular with joggers due to the large number of dead-end streets. The second neighbor had seen a man who looked like Guard waiting at a Utah Transit Authority (UTA) bus stop across the street from the neighbor's house for approximately half an hour around mid-afternoon. The man was wearing denim pants and a dark shirt When the UTA bus arrived about 3:00 p.m., the man did not board the bus but instead stayed at the bus stop, which was near the school bus stop. The neighbor then saw the school bus drop off the children and the man follow three girls up the road. The neighbor did not see the man approach or grab any of the girls. The neighbor thought the UTA bus stop was "about 70 feet at the most" from his house but said that he is farsighted and can see well at a distance, even though his vision is poor "close up." An investigator for the defense measured the distance as 245 feet.
T6 Prior to trial, Guard gave notice that he intended to call Dr. David H.. Dodd to testify as an expert regarding the reliability of eyewitness identification, specifically "concerning the full range of cognitive processes associated with the eyewitness, including at*447tention, perception and memory." The State moved to exclude Dr. Dodd's testimony, asserting that it "would amount to a lecture to the jury and would infringe on the jury's responsibility to weigh the eredibility of the witnesses." Around this same time, Guard filed a motion to suppress the child's identification of Guard through the photograph lineup. In support of this motion, Guard cited several factors known to affect the reliability of eyewitness identification that were present in the cireumstances of the crime, including the short duration during which the child viewed her abductor, a stranger; her hyperactive state due to having to defend herself from attack; and the cross-racial nature of the identification because the child is African American and she described the abductor as Hispanic.2
{7 The court scheduled a hearing on Guard's motion to suppress (the motion hearing) during which Guard explained that recent studies and case decisions from around the country have "consistent[ly]" identified the same factors he cited in his memorandum as indicating problematic eyewitness identifications. Guard then illustrated how those factors were present in the case, particularly as they related to the photographic lineup. In this context, Guard agreed to a hearing pursuant to State v. Rimmasch, 775 P.2d 388 (Utah 1989), superseded by rule as stated in State v. Maestas, 2012 UT 46, ¶ 121 n. 134, 299 P.3d 892, where he would call Dr. Dodd to testify about the fallibility of eyewitness identification in support of his motion to suppress the child's photograph lineup identification. The purpose of that hearing would be to establish the reliability of Dr. Dodd's expert testimony under rule 702 of the Utah Rules of Evidence to determine its admissibility at trial. The court agreed that a Rim-masch hearing might be necessary for purposes of Guard's motion to suppress the child's photo-lineup identification but noted that if Dr. Dodd's testimony was "very generic ..., not going to this alleged victim, . but a general type of ... testimony regarding cognitive processes associated with eyewitnesses," the "Court could allow that expert testimony to come in." When the State insisted on a Rimmasch hearing because it was not clear whether Dr. Dodd would testify at trial generally about the fallibility of eyewitness identification or specifically about the child's reliability in identifying Guard, the court denied Guard's motion to suppress the child's photograph identification of Guard but indicated that it "w[lould] reconsider" if Guard "renew[ed] the motion at a Rimmasch hearing."
8 Guard renewed his motion to suppress, and the matter was set for a Rimmasch hearing two weeks later. Dr. Dodd's testimony at the Rimmasch hearing focused primarily on the reasons that the photograph-lineup identification should be suppressed.3 In the course of his testimony, however, Dr. Dodd also addressed the factors that affect the accuracy of eyewitness identification generally, and the court acknowledged that those factors were significant in assessing the reliability of an eyewitness identification. At the close of the hearing, the State inquired about whether defense counsel would "call [Dr. Dodd] as an expert to talk to the jury about eyewitness identification reliability" if "he doesn't prevail on that motion [to suppress]." Guard's attorney confirmed that was his intent and told the court, "I can give a two-page synopsis of that in the next day or so. ..." However, counsel never provided the promised synopsis.
T9 The case proceeded to jury trial. The State's entire case at trial consisted of the testimonies of the four eyewitnesses-the child, her schoolmate, and the two neighbors-and the officer and the detective who interviewed them. - Guard called several witnesses in support of an alibi defense, but Dr. Dodd did not testify. Although six months after trial the trial court issued a written decision explaining that it had granted a motion by the State to exelude Dr. Dodd's testimony about the general fallibility of eyewitness identification, both the motion and *448the court's resolution of it apparently occurred off-the-record: "At, or just prior to trial, this Court, having received no supplementary briefing by [Guard,4] granted the State's Motion to Exclude [Guard's] Expert Witness from the bench. A subsequent search of the record, however, indicates that a record was not made of that Ruling." As a consequence, the record provides no explanation of the cireumstances under which the court made its ruling to exclude Dr. Dodd's testimony, ie., it cannot be determined whether Guard simply did not call Dr. Dodd because the trial court had already excluded his testimony or if at trial Guard indicated his intent to call Dr. Dodd and the court then granted a motion by the State to exclude him.5 However it came about, the court refused to let Dr. Dodd testify and instead gave what has come to be known as a Long instruction (after State v. Long, 721 P.2d 483 (Utah 1986), in which it was first approved), cautioning the jury about the risks inherent in eyewitness identification and describing specific factors that jurors should consider in assessing the reliability of the child's identification of Guard. The jury convicted Guard of child kidnapping. Guard now appeals the trial court's decision to exclude Dr. Dodd's testimony about the reliability of eyewitness identification.
ISSUE AND STANDARD OF REVIEW
110 "The trial court has wide discretion in determining the admissibility of expert testimony." State v. Hollen, 2002 UT 35, ¶ 66, 44 P.3d 794 (citation and internal quotation marks omitted). Thus, we review a trial court's decision to admit or exclude expert testimony for abuse of discretion, which means we will not disturb that decision "unless [it] exeeeds the limits of reasonability." Id. (citation and internal quotation marks omitted).
ANALYSIS
€ 11 After Guard's trial, the Utah Supreme Court issued State v. Clopten, 2009 UT 84, 223 P.3d 1103, which addresses the precise issue Guard raises here. In Clopien, the supreme court held that eyewitness expert testimony should be routinely admitted in most stranger identification cases. Id. TT 80, 49. Although Clopten was a murder case, it involves eyewitness identification issues remarkably similar to those raised here. Clop-ten and Guard were each tried and convicted in 2006, barely three months apart. See id. T2. In both cases, the convictions resulted solely from the identification of the defendant by eyewitnesses who had never seen the defendant before the day of the crime. See id. 1T 2, 46. In Clopten, the principal eyewitness saw the defendant as the witness entered the club where the murder occurred, but she described the shooter to the police while she was "frantic" and identified Clop-ten later that same day while she was "still distraught." Id. In this case, the child viewed her abductor for a very short time while she was in an emotional and fearful state and described him to the police shortly thereafter. See id. The respective eyewitnesses subsequently identified the defendants from a lineup (an in-person lineup in Clopten and a photographic lineup in this case) where only the defendant possessed a *449prominent attribute of the person the eyewitnesses had described to the police-only Clopten was wearing a red hooded sweatshirt, as had been the shooter, and only Guard had curly hair, as did the abductor. See id. Although the primary eyewitness in Clopten picked Clopten out of the lineup within an hour of the shooting, id., while the child did not identify Guard until the next day, each identification was potentially influenced by police expectation. The witness in Clopten was reluctant to identify the shooter and did so only after "an officer told her to '[dlo it for [the victim].'" Id. (first alteration in original). The child testified that she understood that her abductor would be in one of the photographs, even though the detective had not said that he would be. The primary eyewitness in each case also described the perpetrator's race as different from her own.6 See id. 1 46.
1 12 In addition, there were concerns about the reliability of the other eyewitnesses in each case. In Clopten; one of the witnesses changed his story in response to the officer's ultimatum that he "'be a witness' or 'go to jail for many years'" and then disappeared prior to trial, while another "received a substantially reduced sentence in exchange for his testimony." Id. 142. A third witness saw a man in a red sweatshirt with whom she had spoken just prior to the shooting standing over the victim with a gun. Id. 145. Although she later identified Clopten as the shooter, she had told the police that the shooter had not been wearing red pants, as Clopten had been, and she had described a different red sweatshirt than the one Clopten was wearing when arrested that, in fact, matched the sweatshirt another suspect had been wearing prior to the lineup. Id. Here, the neighbors were shown Guard's picture before being asked if they had seen him the previous day. One neighbor reported that she "thought [she] had seen" Guard run by while she was in the front yard waiting for her children to return from school. That neighbor, who lived some blocks away from where the abduction occurred, did not witness the event itself. The other neighbor claimed to have observed Guard for some time but at a significant distance by his own measurement (and at a distance more than three times the neighbor's estimate, according to the defense investigator who measured it). And while the schoolmate witnessed the kidnapping and corroborated the child's de-seription of the abductor's clothing, she did not provide any information that would identify Guard as the perpetrator.
113 Furthermore, in both cases, defense counsel sought to have Dr. Dodd testify as an expert on the "various factors that can affect the accuracy of eyewitness identifications." See id. 18. And in each, the court excluded Dr. Dodd's testimony. In Clopten, the trial court excluded Dr. Dodd's expert testimony because it concluded that the potential problems with eyewitness identification could be adequately explained through the use of a jury instruction. Id. T4. Here, the court excluded the testimony because "[Guard] failed to meet the initial threshold requirement [of demonstrating that the scientific principles underlying Dr. Dodd's expert's opinion are reliable] in a Rimmasch analysis." It instead gave the jury a Long instruction.
1 14 We recognize that the facts in Clopten presented some additional concerns regarding the eyewitness identifications that are not present in this case and that even the existing analogous concerns are not entirely equivalent. Certainly, some aspects of the identification of Guard in this case were not as problematic as the identification of Clop-*450ten, but there are some eyewitness issues present here that are arguably of greater concern. - Overall, we believe that the facts of this case substantially parallel the facts that led the supreme court to be concerned about the conviction in Clopten in the absence of expert testimony to educate the jury on the fallibility of eyewitness testimony. This conclusion, coupled with the similarity of the issue presented, the commonality of the proposed expert, and the proximity of the trials, thus confronts us with the question of whether our resolution of Guard's case should follow Clopten.
I. The Unusual Cireumstances of This Case Require Application of the Same Analysis as in Clopten.
¶15 Following oral argument, we requested supplemental briefing from the parties to address the question of whether Clopten should apply to the decision before us. In response to our request, Guard asserts that when a procedural rule changes during a criminal case, the new rule is automatically retroactive to all nonfinal criminal cases, including those on direct appeal, unless the change is expressly declared to be prospective only. Because Clopten did not limit its application to future cases only, Guard argues that its rationale automatically applies to his case. The State counters that the issue is governed by the "clear break" exception, which provides that changes to procedural rules that are merely a clarification of existing rules are retroactive while changes that represent a clear deviation from prior practice are not. According to the State, because Clopten rejected the longstanding practice of instructing the jury on eyewitness reliability rather than routinely admitting this type of expert testimony, it "was a clear break with ... rulings in previous cases dealing with eyewitness expert testimony." (Citation and internal quotation marks omitted.)
116 In Griffith v. Kentucky, 479 U.S. 314, 107 S.Ct. 708, 93 L.Ed.2d 649 (1987), the United States Supreme Court discussed the application of the clear break exception to questions of retroactivity when a conviction is not yet final, meaning the appeal period has not expired and the appeal process has not been exhausted. Id. at 321 n. 6, 326-28, 107 S.Ct. 708. Griffith addressed whether the new rule establishing the burdens for making a claim of racial discrimination in selecting a jury announced in Batson v. Kentucky, 476 U.S. 79, 106 S.Ct. 1712, 90 L.Ed.2d 69 (1986), was applicable to two defendants-one of whom was tried in the same Kentucky court three months after Batson-whose cases were pending on direct appeal. Griffith, 479 U.S. at 316, 107 S.Ct. 708. The Court concluded that "a new rule for the conduct of criminal prosecutions is to be applied retroactively to all cases, state or federal, pending on direct review and not yet final, with no exception for cases in which the new rule constitutes a 'clear break' with the past." Id. at 328, 107 S.Ct. 708 (emphasis added). The United States Supreme Court and the Utah Supreme Court have since limited Griffith's holding to its context, cases involving rule changes of constitutional dimension. See id. at 322, 107 S.Ct. 708 (explaining that "failure to apply a newly declared constitutional rule to criminal cases pending on direct review violates basic norms of constitutional adjudication"); see also Teague v. Lane, 489 U.S. 288, 303-04, 109 S.Ct. 1060, 103 L.Ed.2d 334 (1989) (explaining Griffith's reasoning for applying constitutional rule changes retroactively); State v. Stilling, 770 P.2d 137, 143 (Utah 1989) (declining to apply the rule mandating a cautionary jury instruction on eyewitness identification adopted in State v. Long, 721 P.2d 483 (Utah 1986), retroactively because Long was decided "on neither federal nor state constitutional principles"); see also State v. Baker, 2010 UT 18, ¶ 24, 229 P.3d 650 (observing that Griffith "eliminated the 'clear break' exception to retroactive application of newly declared constitutional rules for cases pending on direct review").7 Clopten was not decided on a constitutional *451basis, and the reasoning of Griffith is therefore not controlling.
117 Where a rule change is not constitutional, Utah appellate courts have ordinarily not applied "a new rule ... retroactively if it constitutes a clear break with the past." 8 State v. Lovell, 2011 UT 36, ¶ 73, 262 P.3d 803 (citation and internal quotation marks omitted). "A new rule is a clear break with the past if it caused an abrupt and fundamental shift in doctrine as to constitute an entirely new rule which in effect replaced the older one." Id. (citation and internal quotation marks omitted). The State contends that Clopten establishes a new rule requiring routine admission of expert testimony on the reliability of eyewitness identification evidence that is a clear break from the prior practices that established "a de facto presumption against eyewitness expert testimony in Utah's trial courts." State v. Clopten, 2009 UT 84, ¶ 13, 223 P.3d 1103. In his reply brief, Guard expressly rejects any characterization of Clopten as a "clear break" from prior law. In supplemental briefing on the specific issue of retroactivity, however, Guard recognizes that Clopten may be "subject to a[] 'clear break' exception" and advocates for automatic retroactive application of all rule changes to cases pending on direct appeal. We assume, without deciding, that Clopten is a clear break from prior law. Thus, because Griffith does not control, the general rule precluding retroactive application of a rule change that amounts to a clear break with prior precedent applies here.
18 Despite our conclusion that Clop-ten is not retroactive, we believe that the unusual cireumstances of Guard's case nevertheless require application of Clopten 's analysis. Clopten was tried and convicted in February 2006, id. 12, and Guard's trial and conviction occurred in May 2006. Both filed timely appeals. Guard's initial appeal was dismissed due to defense counsel's failure to file a docketing statement, and his appeal rights were duly reinstated four years later. Had Guard's initial appeal gone forward in a timely manner, it would have been at roughly the same stage of proceedings on appeal as Clopten. Given the similarities of both the facts and the issues in each case, as well as both defendants' intention to call the same expert witness, it seems almost inevitable that the two cases would have been either consolidated on appeal or treated as companion cases. See, eg., State v. Davis, 2011 UT 57, ¶¶ 1, 5, 8, 11, 266 P.3d 765 (consolidating the interlocutory appeals of three cases arising in three consecutive years where the "three consolidated cases ... involve nearly identical facts and issues"); State v. Morrison, 2001 UT 73, ¶¶ 1-4, 31 P.3d 547 (considering the constitutionality of the sexual exploitation of a minor statute in a consolidated appeal where both defendants were convicted after downloading and printing child pornography, and each claimed that the statute "was unconstitutionally overbroad and vague on its face"); Stromquist v. Cokayne, 646 P.2d 746, 746 (Utah 1982) (consolidating the appeals from two separate cases that "raise[d] identical issues"); see also, eg., Wasatch Cnty. v. Okelberry, 2008 UT 10, ¶ 15, 179 P.3d 768 (adopting a bright line test for determining when an event interrupts the ten-year period of continuous use necessary to dedicate a road to the public); Town of Leeds v. Prisbrey, 2008 UT 11, ¶¶ 6-7, 179 P.3d 757 (applying the bright line test adopted in the companion case of Okelberry issued the same date because it concerned *452the same issue); Utah Cnty. v. Butler, 2008 UT 12, ¶¶ 14-17, 179 P.3d 775 (same). Under the circumstances, we conclude that had Guard's case proceeded on appeal as it would have without his counsel's default, the result would have been identical to the result in Clopten.
{ 19 Given that Guard's and Clopten's trials essentially paralleled each other and that the eyewitness issues in their cases were substantially similar, if not perfectly aligned, it seems inconsistent with the administration of justice to deny Guard the benefit of the supreme court's approach in Clopten where, but for the happenstance that delayed Guard's appeal, it appears to us that the same analysis would have been applied to both cases.9 Cf. Teague, 489 U.S. at 303, 109 S.Ct. 1060 (explaining that the United States Supreme Court has applied a new rule to companion cases even when it did not apply the new rule retroactively to cases then pending on appeal). Thus, we conclude that the same analysis that the supreme court applied in Clopten should be applied here. Applying that analysis, we conclude that Dr. Dodd's testimony should have been admitted.
II. Application of the Principles Adopted in Clopten Leads to the Conclusion that Dr. Dodd's Eyewitness Identification Testimony Should Have Been Admitted.
120 Appellate courts have long recognized that " 'the vagaries of eyewitness identification are well-known; the annals of criminal law are rife with instances of mistaken identification.'" State v. Long, 721 P.2d 483, 491 (Utah 1986) (quoting United States v. Wade, 388 U.S. 218, 228, 87 S.Ct. 1926, 18 L.Ed.2d 1149 (1967)); see also id. at 488 {noting that there is "no significant division of opinion" about the "unreliability of eyewitness identification"). But "(allthough research has convincingly demonstrated the weaknesses inherent in eyewitness identification, jurors are, for the most part, unaware of these problems." Id. at 490. The fallibility of eyewitness identification increases when the witness is "identifying a stranger and ... one or more established factors affecting accuracy are present." Clopten, 2009 UT 84, ¶¶ 15, 32, 223 P.3d 1103. Among the "well-documented factors that affect the reliability of eyewitness identifications" generally are the emotional state of the observer at the time of observation; "the circumstances of the observation"; "the expectations, personal experiences, biases, and prejudices" of the observer, particularly when identifying a person of another race; the confidence of the person making the identification, as "the accuracy of an identification is, at times, inversely related to the confidence with which it is made"; and the process by which the observation is reported, e.g., through interrogation, which tends to "influence what a witness 'remembers,' " or by narrative, which is more likely to produce only the facts that the eyewitness remembers and deems important. Long, 721 P.2d at 488-90; see also Clopten, 2009 UT 84, ¶¶ 15, 32 n. 22, 223 P.3d 1103.
21 It is in this kind of stranger identification-plus situation that the Utah Supreme Court determined that expert testimony is critical to ensuring that jurors understand the limitations of an eyewitness identification. As the court explained in Clopten, in the absence of expert testimony, defense attorneys have only two tools for conveying to the jury the possibility that an identification is mistaken: cross-examination and cautionary jury instructions. 2009 UT 84, ¶ 16, 223 P.3d 1103. These tools, however, "suffer from serious shortcomings when it comes to addressing the merits of eyewitness identifications" of a stranger because an eyewitness may appear confident even when his or her identification is mistaken, a juror may be unwilling to alter his or her belief that the eyewitness is reliable when a cautionary instruction is received only at the close of trial, *453and such instructions "tend to touch only generally on the empirical evidence ... that certain factors are known to influence perception and memory" without "explain[ing] how this occurs or to what extent." Id. Tf 15-16, 24. Expert testimony, on the other hand, "quantiffies}" the problems that can arise in eyewitness identifications and "teaches jurors about certain factors ... that have a strong but counterintuitive impact on the reliability of an eyewitness" while the evidence is still being presented, ensuring a better likelihood that "a jury ... is ... able to reach a just decision." Id. 120. In addition, such testimony can focus more specifically on cireumstances generally influencing the accuracy of identifications that are also present in the particular case.
122 In Clopten, the Utah Supreme Court recognized the shortcomings of the Long instruction for the first time and held that expert testimony is preferable in the stranger identification-plus situation. State v. Clopten, 2009 UT 84, ¶¶ 32-34, 223 P.3d 1103. In the course of its analysis, the court recognized that testimony on the factors that generally affect the accuracy of eyewitness identification is "sufficiently reliable" to be admissible under rule 702 of the Utah Rules of Evidence, id. 11 35, 38, provided that the witness is "qualified as an expert by knowledge, skill, experience, training, or education," 10 Utah R. Evid. 702(a). The supreme court explained that the current version of rule 702 requires a threshold showing that expert testimony is reliable, a showing that can be made in one of two ways: through a demonstration that the " 'principles or methods underlying the testimony ... (i) are reliable, (i) are based upon sufficient facts or data, and (iii) have been reliably applied to the facts of the case,'" Clopten, 2009 UT 84, ¶ 35, 223 P.3d 1103 (quoting Utah R. Evid. 702(b)), or by a demonstration that the "underlying principles or methods 'are generally accepted by the relevant expert community," id. (quoting Utah R. Evid. 702(c). The court also explained that at the time of Clopten's 2006 trial, rule 702 required a showing of reliability according to the three-part Rimmasch test but that "the current version of rule 702 incorporates an updated reliability analysis for expert testimony" that "subsume[s]" the Rimmasch test. Id. 1187-88; see also State v. Maestas, 2012 UT 46, ¶ 121 n. 134, 299 P.3d 892 (citing Clopten, 2009 UT 84, ¶ 38, 223 P.3d 1103). The court therefore concluded that under either test, the reliability of eyewitness identification expert testimony is so widely acknowledged that it should be considered routinely admissible in cases where the evidence would be helpful to the jury. Clopten, 2009 UT 84, ¶¶ 35-38, 223 P.3d 1103 ("The phenomena that eyewitness experts seek to explain have been reviewed and replicated many times in recent decades.... [Utah appellate courts have recognized] that it was appropriate to take judicial notice of 'general acceptance' of those principles in the community of researchers that specialize in the study of eyewitness identification."). And "[iln cases where an eyewitness is identifying a stranger and in which various factors that can affect accuracy are present"-the stranger identification-plus situation-"eyewitness expert testimony is helpful to the jury and thus admissible." Id. 138.
1 23 The cireumstances of Guard's case fit the - stranger - identification-plus - model. Guard was unknown to the child and the other witnesses prior to their identifying him as the child's abductor, and there were additional factors that raised questions about the accuracy of the identification. For instance, the child's opportunity to view her abductor was of very short duration while she was also attempting to escape by fighting him off. When the child later identified Guard as the abductor, he was the only suspect in the photograph lineup who had curly hair, a prominent attribute in her description of the abductor. The child testified that upon seeing Guard's picture, she told the detective that "that was the person ... 'for sure."" The presence of this factor is important because "juries seemed to be swayed the most by the confidence of an eyewitness, even *454though such confidence correlates only weakly with accuracy." See id. 115. Yet the jury did not receive any information regarding this weak correlation, even in the Long instruction.
{24 The State contends, however, that Guard cannot take advantage of the supreme court's acknowledgment of the reliability of eyewitness expertise. According to the State, Guard did not ask the court to take judicial notice of Dr. Dodd's testimony in the trial court and, instead, agreed to a Rim-masch hearing. Thus, Guard voluntarily took on the burden of establishing reliability, a burden the trial court said he did not meet.
125 That argument misses the mark. First of all, Guard could not be expected to foresee that a future decision would determine that the reliability of this type of testimony is subject to judicial notice given that, until Clopten, the widespread approach in Utah courts was to the contrary. And, in any event, at the motion hearing, Guard made what was, in all but name, an argument for judicial notice. Specifically, Guard identified the same factors outlined in Clopten as tending to affect the reliability of an eyewitness identification and asserted that "those items are well established in not only Utah case law but throughout the country." See 2009 UT 84, ¶¶ 15, 32 & n. 22, 223 P.3d 1103 (discussing the consensus of research that certain factors affect the accuracy of eyewitness identification).
126 In fact, before trial, the trial court seemed to acknowledge that Dr. Dodd's testimony was based on principles "generally accepted by the relevant expert community," Utah R. Evid. 702(c). For instance, at the motion hearing, the trial court said that if Dr. Dodd's testimony was limited to "very generic ... testimony regarding cognitive processes associated with eyewitnesses," it "could allow that expert testimony to come in," presumably based on the court's own understanding that the reliability of this type of testimony was well established. Further, the court seemed to recognize a distinction between testimony that amounts to "generic education of the jury as it relates to fallibility of eyewitness [identification]" and testimony that "reachfes] conclusions and ... some opinion[s] relative to the reliability of this child." The Rimmasch hearing itself focused not on the reliability of the general principles underlying the fallibility of eyewitness testimony but principally on the particular issues related to the cireumstances of the photograph lineup and Guard's motion to suppress it as unreliable. In fact, at the hearing the court itself described the factors that are usually associated with mistaken identifications as if they were both well known and uncontroversial. While the court seemed to take a different view six months after the trial when it issued its memorandum decision justifying the decision to exclude Dr. Dodd's testimony, it is difficult to reconcile that decision with its earlier statements, and the court itself did not do so. In light of Clopten's definitive resolution of the question of the reliability of the same type of eyewitness expert testimony that Guard offered below (from the same expert witness), it would seem pointlessly technical to conclude that Guard is somehow foreclosed from raising the issue of judicial notice in this appeal because he acceded to a Rimmasch hearing and did not manage to put on enough evidence at that hearing to show the reliability of fundamental principles that the Utah Supreme Court has now recognized as the subject of widespread consensus. Thus, we do not find persuasive the State's argument that this case is procedurally distinct enough from Clopten to overcome its strong similarities.
127 Because Guard's conviction is based solely on eyewitnesses who identified him as the abductor, the reliability of those identifications is "of paramount importance." See State v. Clopten, 2009 UT 84, ¶ 48, 223 P.3d 1103. Yet several circumstances are present that potentially undereut the reliability of the child's identification of Guard as her abductor, including the child's limited opportunity to view the abductor, a stranger; her focus on defending herself from his attack; the possible implications of a cross-racial identification; the photograph lineup only containing one suspect-Guard-who had curly hair, a prominent attribute of the child's abductor; and the child's belief that her abductor would be in the lineup. The concern about reliability is underscored by *455the fact that the child was the only eyewitness to the kidnapping who could identify the perpetrator. The schoolmate witnessed the kidnapping and could describe the person who kidnapped the child but could not identify Guard. And the two neighbors, neither of whom had witnessed the kidnapping or knew Guard, claimed to have seen Guard in the neighborhood only after being shown his picture. Yet both neighbors' opportunity to view the person they identified as Guard had limitations: one neighbor's primary focus was on the arrival of her children from school and the other had only seen the person he thought was Guard at a significant distance. In the absence of independent corroborating evidence to support the conviction, we are persuaded that there is a reasonable likelihood that had the jury heard Dr. Dodd's testimony, it may have assessed the reliability of the eyewitnesses' identifications differently. See id. (concluding that there was a reasonable likelihood of a more favorable result for Clopten where the primary eyewitnesses were strangers to Clop-ten, other factors affecting the reliability of the identifications were present, and there was no independent corroborating testimony). At the very least, in the absence of Dr. Dodd's testimony, our "confidence in the verdict ... is undermined." See State v. Kohl, 2000 UT 35, ¶ 17, 999 P.2d 7 ("We will reverse an erroneous evidentiary ruling ... if, absent the error, there is a reasonable likelihood that there would have been a more favorable result for the defendant. A reasonable likelihood of a more favorable outcome exists when the appellate court's confidence in the verdict actually reached is undermined." (citations and internal quotation marks omitted)). Accordingly, we vacate Guard's conviction and remand for a new trial.
CONCLUSION
128 As the supreme court acknowledged in Clopten,
[wle are always reluctant to reverse a jury's decision to convict, particularly when the crime in question is as serious as this one. The seriousness of the crime, however, makes it only more imperative that the jury's decisionmaking abilities are supported by the best information available. If unreliable identifications are not addressed properly at trial, then there exists an unacceptable risk of the innocent being punished and dangerous criminals remaining at large.
2009 UT 84, 1 49. For the foregoing reasons, we vacate Guard's conviction and remand for a new trial in accordance with this decision.
. Guard was a person of interest in a similar kidnapping incident in another city. Because Guard fit the description of the person who had kidnapped the child, the detective thought Guard might be a suspect in this case.
. When asked about his ethnicity at trial, Guard responded, "My mother ... is Haitian and my father is American, Caucasian."
. Following the Rimmasch hearing, the court again denied Guard's motion to suppress the identification from the photograph lineup. Guard does not appeal that decision.
. The only additional material Guard promised the court on the record at the Rimmasch hearing was a two-page synopsis of Dr. Dodd's general testimony regarding eyewitness identification that Guard intended to present if the motion to suppress was denied. But because at least a portion of the proceedings after the Rimmasch hearing relating to the admission of Dr. Dodd's testimony took place off-record, we cannot determine whether the court, in its written decision to grant the State's motion to exclude Dr. Dodd's testimony, was referring to the synopsis or some other promised or requested briefing, the nature of which is not apparent. Despite the court's statement that Guard had failed to provide the supplemental material he promised, the court explained that the motion was not granted on that basis but because "[Guard had] failed to meet the initial threshold requirement [of demonstrating that the scientific principles underlying the expert's opinion are reliable] in a Rim-masch analysis." Thus, even if the court was referring to Guard's promised synopsis, the failure to provide it was not a factor in its decision and we do not address it further.
. The State suggests a third alternative, that Guard's failure to provide the synopsis may indicate that he decided simply to withdraw his notice to call Dr. Dodd. In the face of a barren record, this is simply an invitation to speculate, which we must decline.
. In this case, the child, who is African American, identified her abductor as "possibly Hispanic." The record establishes that Guard's mother "is Haitian" and his father "is American, Caucasian" and that Guard emigrated from Haiti to the United States as a teenager. The State argues that Haitians are of African descent and that because Guard was of the same race as the victim, her identification was not cross-racial. The record, however, does not provide a clear picture of Guard's racial status; thus, the question of whether the identification is cross-racial in reality or simply in perception cannot be re-soived here. That the victim perceived Guard as of a different race, though, suggests that the fundamental problem with cross-racial identification may still have been present. The concern raised by the child's possible misidentification of Guard's race presents a cross-racial identification issue similar to the one that has been widely recognized as a reliability concern and that was present in Clopten.
. Guard cites State v. Belgard, 615 P.2d 1274 (Utah 1980) (per curiam), and State v. Norton, 675 P.2d 577 (Utah 1983), overruled on other grounds by State v. Hansen, 734 P.2d 421 (Utah 1986), in support of his position that Utah treats all procedural rule changes as applicable retroactively to cases that are not yet final. Belgard and Norton both pre-date Griffith, and we do not analyze their continuing viability because we resolve the case on different grounds.
. Prior to the adoption of an automatic retroac-tivity rule, the United States Supreme Court applied a three-part test to determine if a rule change may be retroactively asserted to challenge a conviction, regardless of whether the conviction was final or pending on appeal. Griffith v. Kentucky, 479 U.S. 314, 320-21, 107 S.Ct. 708, 93 L.Ed.2d 649 (1987) (discussing the analysis adopted by Linkletter v. Walker, 381 U.S. 618, 85 S.Ct. 1731, 14 L.Ed.2d 601 (1965). That same test has been adopted and applied by the Utah Supreme Court in cases involving final convictions. See, eg., Labrum v. Utah State Bd. of Pardons, 870 P.2d 902, 912 (Utah 1993); Andrews v. Morris, 677 P.2d 81, 91 (Utah 1983). But we have not located any cases where the Utah appellate courts have applied that test to non-final convictions, and the case law our supreme court has developed in the meantime simply applies the clear break rule described above. See, e.g., State v. Lovell, 2011 UT 36, ¶ 73, 262 P.3d 803; State v. Hoff, 814 P.2d 1119, 1123 (Utah 1991). Furthermore, the parties do not address the three-part test in their briefings. For these reasons, we do not consider whether the three-part test is applicable here.
. Indeed, the Supreme Court in Griffith took note of what appear to be comparable concerns in its analysis of why disparity in treatment between "similarly situated defendants" undercuts the administration of justice. Griffith v. Kentucky, 479 U.S. 314, 327, 107 S.Ct. 708, 93 L.Ed.2d 649 (1987). There, the defendant in Batson v. Kentucky, 476 U.S. 79, 106 S.Ct. 1712, 90 L.Ed.2d 69 (1986), and one of the defendants in Griffith had been tried in February and May 1984, respectively, by the same Kentucky state court. Id. The same prosecutor exercised the peremptory challenges that became the subject of appeal. Id. And it was due simply to the "fortuities of the judicial process'" that the United States Supreme Court heard Batson before Griffith. Id.
. There is no dispute that Dr. Dodd, who was also the expert in Clopten, is qualified as an expert in eyewitness identification.