Easterly, Associate Judge:

Marlon Williams was arrested and prosecuted for the shooting death of Min Soo Kang. As no eyewitnesses to the crime were discovered and as Mr. Williams had no known relationship with Mr. Kang, it took a number of investigative steps for the police to connect Mr. Williams with.the crime: after finding Mr. Kang’s body, the police located his car; after examining fingerprints recovered from Mr. Kang’s car, the police identified Mr. Williams as a potential suspect; and after searching Mr. Williams’s apartment, the police recovered a gun that, when test-fired, left markings on the bullets that appeared to match the markings on bullets recovered from Mr. Kang’s car. This evidence, in conjunction with the testimony of an individual to whom Mr. Williams had made incriminating statements while they were in the courthouse cellblock, formed the bulk of the government’s case. After considering this evidence, a jury convicted Mr. Williams of first-degree felony murder while armed,1 attempt - to commit robbery while armed,2 two counts of possession of a firearm during a crime of violence (PFCV),3 and carrying .a pistol without a license.4 He received an aggregate sentence of 480 months’ imprisonment.

On appeal Mr. Williams primarily attacks the firearms and toolmark evidence presented against him, arguing among other things that, although defense counsel never objected, the examiner should not have been permitted to testify that the markings on the bullets recovered from Mr. Kang’s car were “unique” to the gun recovered from Mr. Williams’s apartment and thus that he did not have any doubt of their source. Because, to date, this court has only assumed without deciding that such testimony of absolute certainty is impermissible, we conclude that Mr. Williams has failed to establish. that it was plain error for the trial court to permit the jury to hear it. We discern no other error warranting reversal, although we agree that Mr. Williams’s attempted robbery conviction and associated PFCV conviction merge with his felony murder conviction and must be vacated.

I. Facts

In the early morning hours of September 13, 2010, the bullet-riddled body of Min Soo Kang was discovered lying on the side of the road in Southeast D.C. The Metropolitan Police Department (MPD) began investigating and learned that Mr. Kang drove a Cadillac Escalade equipped with OnStar, a service that could remotely disable the vehicle. At MPD’s request, OnStar disabled Mr. Kang’s Escalade by the evening of September 13 and directed MPD officers to the vehicle’s location in Northeast D.C,

An MPD officer inspected the Escalade. He found no damage to the exterior of the *346car but discovered what he suspected were bullet holes in the backrest of the driver’s seat. The officer cut into the seat and recovered three bullets. He also collected fingerprints from the Escalade.

An MPD fingerprint examiner entered the fingerprints lifted from the Escalade into the national Automated Fingerprint Identification System (AFIS), which connects unknown prints to known prints in a digital database, AFIS identified Mr. Williams as a possible source of the fingerprints. Based on the fingerprint examiner’s preliminary conclusion that the prints on the Escalade belonged to Mr. Williams, MPD applied for and was granted a search warrant for Mr. Williams’s residence. Executing this' warrant, MPD officers recovered a High Point brand firearm from Mr. Williams’s bedroom.

At trial,5 the government relied almost exclusively on forensic evidence, presenting expert testimony from a fingerprint examiner and a firearms and toolmark examiner.6 The fingerprint examiner testified to his conclusion that the prints recovered from the Escalade belonged to Mr. Williams,- The firearms and toolmark examiner, Luciano Morales, testified on direct examination that when a bullet is fired from a particular gun, .the gun leaves “unique” identifying marks, “similar to a fingerprint,- basically.” He then testified that he had compared the markings on the bullets recovered, from Mr. Kang’s ear with the markings on the bullets test-fired from the gun recovered from Mr. Williams’s apartment (manufactured by High Point and admitted as Exhibit No. 58), and he had concluded that the bullets were fired by the same gun. On redirect, when the prosecutor asked whether there was “any doubt in [his] mind” that the bullets recovered from Mr. Kang’s Esca-lade were fired from the gun found in Mr. Williams’s room, the examiner responded, “[n]o sir.” He elaborated that “[t]hese three bullets were identified as being fired out of Exhibit No. 58. And it doesn’t matter how many firearms High Point made. Those markings are unique to that gun and that gun only.” The prosecutor then asked the examiner whether, “judging from the markings that you find in 58, it’s your conclusion that those three bullets were fired from 58?” The examiner was unequivocal: “Item Number 58 fired these three bullets.”

. Counsel for Mr. Williams did not object to any of this testimony. The jury also heard stipulations that a print lifted from the gun did not match Mr. Williams and that the blood and DNA recovered from the gun did not match Mr. Kang or Mr. Williams. The jury convicted Mr. Williams on all charges.

II. Analysis

A. Sufficiency of the Evidence

We first address Mr. Williams’s argument that the government did not present sufficient evidence to support his felony murder conviction because it failed *347to establish the underlying ’felony of attempted robbery, and specifically failed to prove that Mr, Williams, and. not another person, had stolen.Mr. Kang’s Escalade. Reviewing the sufficiency of the evidence de novo, Nero v. United States, 73 A.3d 153, 157 (D.C.2013), we disagree. As Mr. Williams acknowledges in his brief, the. government presented the following evidence t(o support an. attempted robbery conviction: (1) testimony by the fingerprint examiner that the fingerprints lifted from both the exterior and interior of Mr. Kang’s Escalade matched Mr. .Williams; (2) eyewitness testimony that a person consistent with Mr. Williams’s physical description was seen opening and closing the hood of the Escalade around the time it. was disabled; and (3) testimony by the firearms and toolmark examiner that the bullets recovered from the Escalade matched bullets fired from Mr. Williams’s gun. From this evidence, drawing all reasonable inferences in favor of the government as we must, Nero, 73 A.3d at 157, we conclude that the- jury- reasonably could have determined that Mr. Williams stole Mr. Kang’s car, and thus necessarily committed the crime of attempted robbery.7 See Ray v. United States, 575 A.2d 1196, 1199 (D.C.1990) (“Every completed criminal offense necessarily includes an attempt to commit that offense.”). But see (Richard) Jones v. United States, 124 A.3d 127, 132-34 (D.C.2015) (Beckwith, J., concurring) (highlighting conflicting precedent from this court indicating that for general intent crimes, an attempt conviction requires proof of a higher mens rea than conviction for the completed offense).

B. The Firearms and Toolmark Examiner’s Opinion Testimony

Mr. Williams argues that the firearms and toolmark examiner should not have been able to testify that the markings on the bullets recovered from Mr. Kang’s Escalade were unique or that he was without “any doubt” that these bullets were fired from the gun found in Mr. Williams’s room. Because Mr. Williams did not object at trial to this testimony, we review only for plain error. See (John) Jones v. United States, 990 A.2d 970, 980-81 (D.C.2010). To prevail under this test, it is not enough for an appellant to demonstrate error; the appellant must also show that the error is plain, i.e., that the error is “so egregious and obvious as to make the.trial judge and prosecutor derelict in permitting it, despite the defendant’s failure to object.” Id. at 981. We attribute such dereliction to the trial court only when an error is “clear under current law.”8 Conley v. United States, 79 A.3d 270, 289 (D.C.2013) (quoting United States v. Olano, 507 U.S. 725, 734, 113 S.Ct. 1770, 123 L.Ed.2d 508 (1993)). Applying this standard, we cannot say the trial court plainly erred by permitting the jury to hear the examiner’s certainty statements.

There is no precedent in this jurisdiction that limits a toolmark and firearms *348examiner’s testimony about the certainty of his pattern-matching conclusions. The closest this court has come to addressing this issue was in (.Ricardo) Jones v. United States, 27 A.3d 1130 (D.C.2011). In that case the defense argued inter alia that toolmark and firearms examiners could not “stat[e] their conclusions with ‘absolute certainty excluding all other possible firearms.’ ” Id. at 1138, In response, the government assured this court, both in its appellate brief and at oral argument, that it was the government’s policy not to present such testimony. “In light of the government’s representation,” this court “assume[d], without deciding, that such experts should not be permitted to testify that they are 100% certain of a match, to the exclusion of all other firearms.” Id. at 1139. The court then determined that any such error was harmless. Id. Jones did not plainly bar the toolmark examiner in this case from testifying as he did and does not provide a foundation for a determination of plain error.

Nor can we say that the weight of non-binding authority outside this jurisdiction is a sufficient foundation for a determination that the -trial court “plainly” erred by not sua sponte limiting the tool-mark examiner’s testimony. See Euceda v. United States, 66 A.3d 994, 1012 (D.C.2013) (holding that error cannot be plain where neither this court nor the Supreme Court has decided the issue, and other courts are split on the issue). We are aware of only one state supreme court decision9 and no federal appellate decisions limiting the opinion testimony of firearms and toolmark examiners. Indeed, as one federal district court judge has observed, “[although the scholarly literature is extraordinarily critical” of toolmark pattern-matching, it appears that courts have made little effort to limit or qualify the admission of such evidence.10 United States v. Green, 405 F.Supp.2d 104, 122 (D.Mass.2005).

Mr. Williams refers us to the policy representation made by the government in Jones. The government concedes that, at Mr. Williams’s trial, it violated its policy “to only elicit firearms examiners’ opinions to a reasonable degree of scientific certainty.” But this concession cannot serve as the sole foundation for a determination of plain error. The government’s internal policy does not-constitute binding law11— let alone a “clear” or “obvious” rule — that a trial court should be presumed to know.12 *349 Cf. Rose v. United States, 49 A.3d 1252, 1256, 1258 (D.C.2012) (holding .that a trial court’s error could not be plain when there was “no clear case law” in our jurisdiction and that a published concurrence from a judge of this court, while on point, “is not the law of our jurisdiction”).

Since Mr. Williams has not shown that the state of the law is such that the trial court plainly should have sua sponte precluded or struck the certainty statements of the firearms and toolmark examiner in this case, Mr. Williams’s unpreserved challenge to these certainty statements cannot prevail under our test for plain error.

C. Confrontation Clause and Hearsay Challenges to the Firearms and Toolmark Evidence

Regarding the firearm and toolmark evidence presented in this case, Mr. Williams also challenges the admission, over objection, of two “worksheets” documenting the analysis of the bullets. These worksheets were signed by the firearms and toolmark examiner who testified at trial, Mr. Morales, but they also bore the signature and initials of his colleague, the “lead examiner on that particular case,” Rosalyn Brown.13 The government did not call Ms. Brown to testify because she had since been fired. On appeal, Mr. Williams argues that the admission of the worksheets violated his Sixth Amendment right to confrontation.

The Confrontation Clausé of the Sixth Amendment, U.S. Constamend. VI, prohibits the government from introducing “testimonial” hearsay at a criminal trial, unless the declarant isunavailable and the defendant has had a prior opportunity to cross-examine him. Crawford v. Washington, 541 U.S. 36, 53-54, 124 S.Ct. 1354, 158 L.Ed.2d 177 (2004). A hearsay statement is considered testimonial if it is “ ‘a solemn declaration or affirmation made for the purpose of establishing or proving some fact’ ... in the prosecution or investigation of a crime.” Young v. United States, 63 A.3d 1033, 1039-40 (D.C.2013) (quoting Crawford, 541 U.S. at 51, 124 S.Ct. 1354). Forensic evidence is also subject to the Confrontation Clause, which means a defendant must have an opportunity to cross-examine the analyst who actually conducted or observéd the forensic testing. Id. at 1039.

Assuming. the ballistics worksheets contained Ms. Brown’s testimonial hearsay statements, we conclude that their erroneous admission was harmless. See Duvall v. United States, 975 A.2d 839, 843 (D.C.2009) (applying the test for harmless error under Chapman v. California, 386 U.S. 18, 87 S.Ct 824, 17 L.Ed.2d 705 (1967) to admission of a lab report in violation of the Confrontation Clause). To begin with, the jury never heard any testimony about Ms. Brown’s observations and conclusions in Mr. Williams’ case and thus *350had no reason to think that the worksheets might document her examination of the bullet and firearm -evidence. On the contrary, Mr. Morales testified (without “any doubt”, see supra Part II.B) only as to his own observations and conclusions. Meanwhile, the prosecution made no reference to another examiner in closing or rebuttal. Lastly, nothing on the worksheets themselves indicated that they reflected the independent conclusions of another, absent examiner. Thus, at most, the jury saw an ambiguous extra signature at the bottom of a document that Mr. Morales had testified reflected his work product. Based on these particular facts, we cannot discern any harm to Mr. Williams from admission at his trial, of these worksheets.14

D. Other Issues

With one exception, Mr. Williams’s remaining arguments fail. His unpreserved challenge to the admission of fingerprint evidence fails the third prong of the test for plain error where trial counsel conceded,'both in opening and in closing, that the fingerprints on the Escalade belonged to Mr. Williams.15 Mr. Williams’s new argument that he is entitled to a Franks hearing16 also fails; the trial court did not plainly err by overlooking the discrepancy between the affidavit in support of the search warrant for Mr. Williams’s apartment, which cited fingerprint evidence as a basis for probable cause, and the fingerprint examiner’s testimony that he reviewed the prints and linked them to Mr. Williams on a date after the search warrant was executed. Instead, given other documentation indicating that the fingerprint examiner was asked to analyze the latent prints before the police sought and obtained the warrant, it would have been reasonable for the trial court to conclude that the examiner was simply mistaken as to the date on which he first examined the latent prints and connected them to Mr. Williams.

Mr. Williams prevails on his argument that this court must merge his attempted' robbery and corresponding PFCV conviction with his felony murder conviction. “[A] person cannot be convicted of both felony murder'and the underlying felony that supported the felony murder conviction.” Matthews v. United States, 13 A.3d 1181, 1191 (D.C.2011). Accordingly, we remand the case with instructions for the trial court to vacate Mr. *351Williams's convictions for attempted robbery and the associated count of PFCV. See Morris v. United States, 622 A.2d 1116, 1130 (D.C.1993) (holding that when two predicate crimes for PFCV merge into one, the PFCV offenses also merge).

In all other respects, we affirm the judgment of the trial court.

So ordered.

Concurring opinion by Associate Judge EASTERLY.

EASTERLY, Associate Judge,

concurring:

In our adversarial system, we do not expect trial courts to “recognize on [their] own” that an expert’s testimony is “scientifically unorthodox or controversial.” (John) Jones v. United States, 990 A.2d 970, 980-82 (D.C.2010). In the absence of any objection at Mr; Williams’s trial to the admission of the firearms and toolmark examiner’s certainty statements, we could only reverse if the law were clear that the expert could not make these statements. See supra Majority Opinion, Part II.B. As discussed above, the law in this jurisdiction does not clearly preclude a firearms and toolmark examiner from testifying with unqualified, absolute certainty.1 But it should.

A statement that markings on a bullet are “unique” to a particular gun is a statement that the probability of finding another gun that can create identical bullet markings is zero. If purportedly unique patterns on bullets are declared a match, that declaration likewise negates the possibility that more than one gun could have fired the-bullets — it is a statement of unqualified certainty that the bullets were fired from a specific gun to the exclusion of all others. Here the firearms and tool-mark examiner testified that he had identified matching “unique” patterns; .he also declared that he did not have “any doubt” that the bullets recovered from Mr. Kang’s car had been fired by the gun recovered from Mr. Williams’s apartment.

The government has a policy, admittedly violated here, not to elicit such certainty statements. This court was advised of the government’s policy in Jones. At oral argument in that case, in November 2011, counsel for the government stated that, as “concede[d]” in its brief, it was the government’s “position that practitioners should not state their conclusions to 100% scientific certainty.” The government further noted that it had “conceded in every hearing, starting two to three years ago when we first started having Frye hearings on this issue, that firearms examiners should not state their conclusions with absolute certainty.”2 Id. Which ráises the question: why did the government adopt a policy to limit the opinion testimony of firearms and toolmark examiners? What happened “two to "three” years'before the Jones oral argument that prompted the creation of this policy?

In 2008, a committee of scientists and statisticians assembled by. the National Research Council (NRC),3 which was in *352turn acting at the behest of the Department of Justice, issued a report on bullet pattern-matching analysis, Ballistic Imaging. 4 Although the NRC Committee’s charge was to assess the feasibility and utility of establishing “a national reference ballistic image database ... that would house images from firings of all newly manufactured or imported firearms,” it recognized that the “[underlying ... question” is “whether firearms-related toolmarks are unique: that is, whether a particular set of toolmarks can be shown to come from one weapon to the exclusion of all others.” Ballistic Imaging, supra note 3, at 1, 3. The NRC Committee determined that there was no data-based foundation to declare, with any certainty, individualization based on toolmark pattern matching.

Specifically, the NRC Committee made a “finding” that the “validity of the fundamental assumptions of uniqueness and reproducibility of firearms-related toolmarks has not yet been fully demonstrated.” Ballistic Imaging, supra note 3, at 3, 81. The NRC Committee noted that “derivation of an objective, statistical basis for rendering decisions [about matches] is hampered by the fundamentally random nature of parts of the firing process. The exact same conditions — of ammunition, of wear and cleanliness of firearms parts, of burning of propellant particles and the resulting gas pressure, and so forth — do not necessarily apply for every shot from the same gun.” Id. at 55. The NRC Committee concluded that “[a] significant amount of research would be needed to scientifically determine the degree to which firearms-related toolmarks are unique or even to quantitatively characterize the probability of uniqueness.” Id. at 3, 82.

The NRC Committee further expressed concern that, notwithstanding the absence of data and the corresponding statistical unknowns, firearms and toolmark examiners “tend to cast their assessments in bold absolutes, commonly asserting that a match can be made ‘to the exclusion of all other firearms in the world.’” Ballistic Imaging, supra note 3, at 82. The NRC Committee denounced this sort of testimony, explaining that “[sjuch comments cloak an inherently subjective assessment of a match with an extreme probability statement that has no firm grounding and unrealistically implies an error rate of zero.” Id. “[Shopping short of commenting on whether firearms toolmark evidence should be admissible” in court, the NRC Committee determined that “[c\onclusions drawn in firearms identification should not be made to imply the presence of a firm statistical basis when none has been demonstrated.” Id. (emphasis in original).

In a subsequent report commissioned by Congress and issued in 2009, Strengthening Forensic Science in the United States: A Path Forward,>5 another NRC Commit*353tee published similar words of warning regarding firearms and toolmark evidence.6 This Committee explained that “[individual patterns from manufacture or from wear might, in some cases, be distinctive enough, to suggest one particular source.” Id. at 154 (emphasis added). But “[b]ecause not enough is known about the variabilities among individual tools and guns,” the Committee was “not able to specify how many points of similarity are necessary for a given level of confidence in the result.”7 In other words, there is currently no statistical basis to declare with any degree of certainty that tool-marks on a bullet connect that bullet to a particular gun or “match” the markings on other bullets fired from that gun.8

Against this backdrop, there is only one permissible answer to the question left undecided'in Jones regarding firearms and toolmark examiners’ assertions of certainty in their pattern-matching conclusions: the District of Columbia courts should not allow them. It is well established that expert opinion evidence is admissible if “it will not mislead the jury and will prove useful in understanding the facts in issue.” *354 Clifford v. United States, 532 A.2d 628, 632 (D.C.1987) (citing Dyas v. United States, 376 A.2d 827, 831 (D.C.1977)); of Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 589, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993) (“[T]he trial judge must ensure that any and all scientific testimony, or evidence admitted is not only relevant, but reliable.”). Certainty statements such as those .elicited by the government in.this case.are misleading and lack any legitimate utility in criminal trials; they express a solid statistical foundation for individualization that does not currently (and may never) exist.

The government states in its brief to this court that it is “regrettable” that its expert was permitted to state his pattern-matching conclusion with absolute certainty. It is more than regrettable.. It is alarming. We know that faulty forensic evidence, and in particular, objectively unfounded statements of certainty regarding forensic analysis, can contribute to wrongful convictions. See Strengthening Forensic Science, supra note 5, at 45; Brandon L. Garrett, Judging Innocence, 108 Colum. L.Rev, 55, 83-84 (2008).

Take the case of Donald Gates, who was wrongfully convicted of rape and murder and needlessly served twenty-seven years in prison.9 To- persuade a jury of Mr. Gates’s guilt, the government relied on the similarly subjective pattern-matching analysis of hair evidence. The hair examiner in Mr. Gates’s case testified- with only slightly more restraint than the firearms and toolmark examiner in this case, • acknowledging that “it cannot be said that a hair came from one person to the exclusion of all others,” but nonetheless asserting that it was “ ‘highly unlikely’ that the hair found on the victim came from someone other than [Mr. Gates].” Brief for Appellee at 8, Donald E. Gates v. United States, 481 A.2d 120 (D.C.1984) (transcript citations omitted). But, just as in this case, there was no data-based foundation for the expert’s expression of certainty in his opinion.10

The use of these subjective certainty statements not only implicates the government’s “duty to refrain from improper methods calculated to produce a wrongful conviction,”11 it also calls into question the *355“fairness, integrity [and] public reputation of judicial proceedings.”12 Courts are oiir society’s chosen forum for ascertaining guilt in criminal cases. Our justice system can only function if it maintains the trust of the community. We rely on judges-r-as the umpires in our adversarial system — to prohibit the admission of evidence that is clearly without foundation. As matters currently stand, a certainty statement regarding toolmark pattern matching has the same probative value as the vision of a psychic: it reflects nothing more than the individual’s foundationless faith in what he believes to be true. This is not evidence on which we can in good conscience rely, particularly in criminal cases, where we demand proof — real proof — beyond a reasonable doubt, - precisely because the stakes are so high. To uphold the public’s trust, the District of Columbia courts must bar the admission of these certainty statements, whether or not the government has a policy that prohibits their elicitation. We cannot be complieit in their use.