(concurring in part and dissenting in part).

I join in the concurrence and dissent of Justice Russell A. Anderson.

(concurring in part and dissenting in part).

I join in the DNA sections of the concurrence and dissent of Justice Russell A. Anderson.

(concurring in part and dissenting in part).

I join in the Post-Miranda Statement and the DNA: Due Process sections of the concurrence and dissent of Justice Russell A. Anderson.

(concurring in part, dissenting in part).

I join in the concurrence and dissent of Justice Meyer.

OPINION

Appellant Billy Daymond Bailey was convicted of first-degree murder while committing criminal sexual conduct in connection with the May 1984 death of a 69-year-old neighborhood resident, Agnes Fafrowicz. On appeal to this court, he makes ten claims of error. Because prejudicial statements made by Bailey to police were erroneously admitted in evidence in violation of Miranda v. Arizona, 384 U.S. 436, 86 S.Ct. 1602, 16 L.Ed.2d 694 (1966), we reverse and remand to the district court for a new trial. To provide guidance for the new trial, we also address the other evidentiary issues that are likely to recur.

A. The Initial Investigation

On Sunday, May 20, 1984, Virginia Golden entered the house of her mother, Agnes Fafrowicz, and discovered her mother’s body on the living room floor. When police arrived, they found evidence that Faf-rowicz had been robbed and sexually assaulted by someone who broke into her home. Investigators concluded that the crime had occurred on Wednesday, May 16. They took fluid samples from Fafrow-icz’s mouth, vagina and anus. Lab tests detected sperm in the vaginal sample.

On May 22, police learned that two checks “had come into the bank on Ms. Fafrowicz’ account.” When Detectives Ronald Snobeck and Robert Nelson acquired the checks, they saw that one of them, dated May 17,1984, was made out to “Bill Vollmar-Bailey” for the sum of $230.00. The back of that check indicated that it had been cashed by Gopher Liquor, a store next to Bailey’s apartment building. ■

B. The 198i Interrogations and Indictment

While they were at Gopher Liquor questioning the store’s owner, Detectives Sno-beck and Nelson spotted Bailey walking across the adjacent parking lot and entering his apartment building. According to Detective Snobeck, when they saw him reemerge and begin walking toward the corner, the detectives returned to their car and drove close to Bailey, cutting off his path. The detectives exited the car, shouted to Bailey to stop and put his hands up, and approached him, one of them with his gun drawn and the other with his hand on *386his gun. They identified themselves as police, put Bailey against their car for a pat-down weapons search, handcuffed his hands behind his back and put him in the backseat of the car. The detectives interrogated Bailey in the car without providing any Miranda warning and then told him they would take him to the station.

At the station, the same two detectives continued to interrogate Bailey without providing any Miranda warning. Midway through the station interrogation, they read Bailey his Miranda rights.1

When the detectives testified at the hearing conducted on Bailey’s motion to suppress both the warned and unwarned statements, neither of them provided any explanation of why they did not give a Miranda warming in the car or on arrival at the station. According to Detective Nelson, Bailey told the detectives in the car that Fafrowicz made out the check to him on May 18 for work that he had done on her lawn and the brakes of her car on the previous day. After they brought Bailey to the police station but before he was informed of his Miranda rights, he elaborated: he claimed that he had done “a brake job” on Fafrowicz’s car on May 17, cut her grass on May 14, and at some point “cleaned up her battery posts on her car” and “cleaned the engine compartment up.”

As the detectives realized, Bailey’s story was riddled with doubtful assertions. Bailey claimed that Fafrowicz had written him the check on May 18, whereas there was considerable evidence that Fafrowicz had died on May 16. The detectives noticed several inaccuracies in Bailey’s description of Fafrowicz’s car and lawn mower, and they believed his claims to have cut her grass and cleaned her car battery to be demonstrably false.

The “culmination” of the police-station interrogation, according to the detectives, came when Bailey misidentified the side of Fafrowicz’s car on which the battery was located. At this point, the detectives told Bailey that he was under arrest for murder, and Detective Nelson gave Bailey the standard Miranda warning.2 He agreed to waive his Miranda rights. The interrogation continued immediately, with no significant pause. Bailey largely recounted his story but added the further detail that he had arranged to paint Fafrowiez’s house. Challenged by the officers to explain how Fafrowicz could write him a check on May 18, when she was already dead, Bailey responded, “That’s a good question.”

Bailey was charged under Minn.Stat. § 609.185(2) (1982)3 with first-degree mur*387der while committing criminal sexual conduct. He was indicted on June 6, 1984. About 6 months later, the state dismissed the indictment under' Minn. R.Crim. P. 30.01, stating:

[S]inee the Grand Jury indictment, all of the physical evidence has been processed by the Bureau of Criminal Apprehension laboratory. The results of that examination tend to negate some of the evidence upon which the Grand Jury indicted this defendant. This recently obtained evidence makes it highly unlikely that this case could be proved beyond a reasonable doubt.

The state took no further action on the case for several years. , ;

C. The 1992 Destruction of Evidence

In February 1992, police lieutenant Brad Johnson, then head of the homicide division of the Minneapolis Police Department, authorized the destruction and release of several pieces of physical evidence relating to Fafrowicz’s murder. Photographs and photocopies of some of the destroyed and released evidence (such as the $230 check) were retained. The district court later allowed trial witnesses to describe the destroyed evidence, and the court allowed the defense to attack the reliability of that testimony based on the fact that the underlying evidence had been destroyed.

D. The 2000 DNA Tests and Trial

When the case was reopened in 2000, investigators procured from the Medical Examiner’s Office two slides that reportedly contained the vaginal and oral samples taken from Fafrowicz’s body during the autopsy. Both slides were protected by cover slips that were adhered by a mounting medium. Bureau of Criminal Apprehension (BCA) scientist Catherine Knutson recognized that the tests she intended to perform on the DNA samples would exhaust those samples. She advised the prosecutor of this fact by letter approximately six weeks before the tests were performed, citing Minn. R.Crim. P. 9.01, subd. 1(4), and stating in part:

This case is identified as a homicide case involving Bill Daymond Bailey and Agnes Mary Fafrowicz * * *. Please be advised that this testing “may preclude any further tests or experiments” within the meaning of the'amendment to Minnesota Rules of Criminal Procedure 9.01, subdivision 1(4), January 1, 1990. For example, it may affect the evidence in our possession in the following manner: all of the evidence will be consumed in the analysis regarding the oral slides (Item 41) and vaginal slides (Item 42), both said to have been collected from Agnes Mary Fafrowicz.

The decision whether to disclose this information to any other party is your responsibility as the prosecuting attorney.

The prosecutor did not notify Bailey or his previously appointed attorney before Knutson performed the tests.

In October 2000, using a Profiler Plus kit, Knutson tested the samples against a blood sample taken from Bailey during the 1984 investigation. To gain.access to the sample on the slides, Knutson applied heat to the slides with a Bunsen burner until the mounting medium began to boil. Bailey claims that this process was in violation *388of approved laboratory protocols and had not been the subject of any validation studies by the BCA. The vaginal sample yielded interpretable DNA results at six of the ten tested loci, including the amelogenin (the sex gene). The resulting profile matched the DNA from Bailey’s blood sample. The Profiler Plus test exhausted the DNA sample. As a result, the BGA was not able to also test the sample with the Cofiler kit, which typically examines an additional three loci, and no sample was available for independent testing by Bailey.

With this new evidence, police once again charged Bailey with first-degree murder while committing criminal sexual conduct. A grand jury indicted him in December 2000. At that time, Bailey was serving time in federal prison in Oklahoma for unrelated offenses.

At trial, over Bailey’s objections, the district court allowed the state to present (1) evidence regarding Bailey’s inculpatory statements made after the Miranda warning; (2) witness testimony describing and drawing conclusions from the destroyed or released physical evidence; (3) testimony from Knutson and other state experts providing the results of the DNA testing; and (4) evidence of three subsequent burglaries committed by Bailey. On February 28, 2002, the jury found Bailey guilty of first-degree murder. He was sentenced to a consecutive term of life imprisonment.4

E. Issues on Appeal

On appeal to this court, Bailey makes ten claims of error, six of them concerning the DNA evidence that was introduced against him. First, he argues that the district court improperly admitted his in-culpatory statements made during police interrogation after he was given the Miranda warning. Second, he complains that the court improperly allowed the prosecution to solicit testimony regarding the evidence that the state destroyed and released in 1992. Third, he argues that the court improperly admitted the results of DNA tests because (1) the chain of custody for the sample was unsubstantiated; (2) the state solicited inadmissible testimony purporting to interpret the results of DNA testing at the other four loci even though those results did not reach the BCA’s normal threshold for interpretation; (3) the state’s expert impermissibly used the “product rule” in calculating probability statistics; (4) the state violated Minn. R.Crim. P. 9.01, subd. 1(4), by failing to notify Bailey before state experts conducted a test that used up the DNA sample; (5) the use of Bunsen burner heat to remove the cover slip and take the DNA sample from its slide was not a “validated technique” and violated laboratory standards; and (6) the PCR-STR method for testing DNA, as applied in this case, is unreliable and violates due process. Next, he argues that prior-crimes evidence was improperly admitted. Finally, he argues that the prosecutor engaged in courtroom misconduct that deprived Bailey of a fair trial.

I.

The parties agree that the statements Bailey made to police prior to being informed of his Miranda rights were inadmissible. The district court held that the statements Bailey made to police after receiving the Miranda warning were voluntary and admissible under the United States Supreme Court’s decision in Oregon *389v. Elstad, 470 U.S. 298, 105 S.Ct. 1285, 84 L.Ed.2d 222 (1985).

In Elstad, the Court concluded that the traditional “taint” analysis does not apply to Miranda violations, stating that

[i]t is an unwarranted extension of Miranda to hold that a simple failure to administer the warnings, unaccompanied by any actual coercion or other circumstances calculated to undermine the suspect’s ability to exercise his free will, so taints the investigatory process that a subsequent voluntary and informed waiver is ineffective for some indeterminate period. Though Miranda requires that the unwarned admission must be suppressed, the admissibility of any subsequent statement should turn in these circumstances solely on whether it is knowingly and voluntarily made.

Elstad, 470 U.S. at 309, 105 S.Ct. 1285 (emphasis added). Elstad did note, however, that the failure to give a Miranda warning does create “a presumption of compulsion.” Id. at 307, 105 S.Ct. 1285.

The rationale of Elstad may have been weakened somewhat by subsequent decisions. The rejection in Elstad of a taint analysis was based in part on the conclusion that a Miranda warning is not constitutionally required. Id. at 306-07, 105 S.Ct. 1285. But the Court later ruled that a Miranda warning does have constitutional stature. Dickerson v. United States, 530 U.S. 428, 438, 120 S.Ct. 2326, 147 L.Ed.2d 405 (2000). The Court has not considered the impact of Dickerson on the Elstad rule.5 In the absence of a definitive ruling to the contrary, we will continue to follow the Elstad rule in comparable circumstances.6 We note that El-stad limited its ruling to its own peculiar facts, stating “the admissibility of any subsequent statement should turn in these circumstances solely on whether it is knowingly and voluntarily made.” 470 U.S. at 309, 105 S.Ct. 1285. Throughout the Elstad opinion, the Court emphasized that “these circumstances” involved a “voluntary but unwarned admission from the defendant” (470 U.S. at 303, 105 S.Ct. 1285); given in a situation where the officers failed to give a Miranda warning because of an error in determining that the interrogation was “custodial”; and the unwarned statement was “unaccompanied by any actual coercion or other circumstances calculated to undermine the suspect’s ability to exercise his free will.” 470 U.S. at 309, 105 S.Ct. 1285. Indeed, the Court recognized that in different circumstances, “[w]hen a prior statement is actually coerced, the time that passes between confessions, the change in place of interrogations, and the change in identity of the interrogators all bear on whether that coercion has carried over into the second confession.” 470 U.S. at 310, 105 S.Ct. 1285 (citations omitted).

The district court’s order denying Bailey’s motion to suppress his Mirandized statement concluded: “The totality of the circumstances reveal that Defendant was not coerced, was properly Mirandized, and freely, knowingly, and voluntarily waived his rights. Defendant’s third statement is therefore admissible.” This conclusion *390combines elements of fact and law and must be viewed in the light of the purposes of the Miranda rule and caselaw applying Miranda under similar facts.

In that context, Bailey argues that El-stad is factually distinguishable and that this case is controlled by the Eighth Circuit decision in United States v. Carter, 884 F.2d 368, 373-74 (8th Cir. 1989). The state argues that Elstad does apply and further relies upon our interpretation of Elstad in State v. Scott, 584 N.W.2d 412, 419-20 (Minn. 1998).

Bailey distinguishes Elstad because (a) the circumstances of Bailey’s interrogation were more coercive; (b) Elstad’s non-Mir-andized statement was brief, only indicating his presence at the crime scene; and (c) Elstad’s Mirandized statements occurred after a significant pause in the interrogation.

Bailey argues that Carter is more analogous. Carter was interviewed for 55 minutes without a Miranda warning. Carter, 884 F.2d at 369. After he made some incriminating statements, he was given a Miranda warning, was further interviewed, signed a waiver form and wrote a handwritten confession. Id. The Eighth Circuit held that the warned confession was inadmissible, distinguishing Elstad as follows:

In this case, there was no passage of time to speak of between the unwarned confession and the subsequent warnings and confession, all of which occurred as part and parcel of a continuous process. Thus, the second confession came almost directly on the heels of the first. Although Elstad precludes the formulation of a “rigid rule” in determining the admissibility of the second confession, our review of “the surrounding circumstances and the entire course of police conduct with respect to the suspect,” convinces us that the second confession cannot be allowed into evidence.

Carter, 884 F.2d at 373 (citations and footnote omitted).7

We read Carter as applying a bright-line rule to these circumstances, which avoids the need to determine case-by-case the issue of voluntariness. The court emphasized the ease and clarity of the application of Miranda and said:

If the police are permitted * * * to ignore Miranda until after they obtain a confession, the courts will once again be embroiled in the endless case-by-case voluntariness inquiries Miranda was designed to prevent, and the case-of-appli-eation rationale enunciated by the Supreme Court will be largely nullified.

Id. at 374.

The state argues that this case is controlled instead by Scott. Police took Scott into custody at gunpoint, handcuffed him and took him to the police station. Scott, 584 N.W.2d at 419. Police then interrogated Scott for three hours, with video- and audiotapes made of the entire interrogation. Id. at 415. Although police did not give Scott a Miranda warning until 15 minutes into the interrogation, Scott did not make any incriminating statements *391prior to the warning. Id. After the warning, Scott confessed to committing two drive-by shootings. Id. We held that the two confessions were admissible under Elstad. Scott, 584 N.W.2d at 419-20.

We agree with Bailey that Elstad is distinguishable. Elstad involved the situation where the unwarned statement was “clearly voluntary” because of the absence of any coercive circumstances. Elstad, 470 U.S. at 310-11, 105 S.Ct. 1285. In fact, Elstad focused on the circumstances where it was not clear that police questioning involved the custodial interrogation necessary to trigger a Miranda warning. The Court supported its conclusion with this rationale:

Because Miranda warnings may inhibit persons from giving information, this Court has determined that they need be administered only after the person is taken into “custody” or his freedom has otherwise been significantly restrained. Unfortunately, the task of defining “custody” is a slippery one, and “policemen investigating serious crimes [cannot realistically be expected to] make no errors whatsoever.” If errors are made by law enforcement officers in administering the prophylactic Miranda procedures, they should not breed the same irremediable consequences as police infringement of the Fifth Amendment itself. It is an unwarranted extension of Miranda to hold that a simple failure to administer the warnings, unaccompanied by any actual coercion or other circumstances calculated to undermine the suspect’s ability to exercise his free will, so taints the investigatory process that a subsequent voluntary and informed waiver is ineffective for some indeterminate period. Though Miranda requires that the unwarned admission must be suppressed, the admissibility of any subsequent statement should turn in these circumstances solely on whether it is knowingly and voluntarily made.

470 U.S. at 309, 105 S.Ct. 1285 (citations omitted).

The facts in Elstad show that the unwarned statement was made without actual coercion. Police went to the Elstad home with a warrant for his arrest in connection with a recent burglary in the neighborhood. After defendant’s mother let the officers in, one of them met with Elstad in the living room and said he felt that Elstad was involved in the robbery, to which Elstad responded “Yes, I was there.” 470 U.S. at 300-01,105 S.Ct. 1285. This was Elstad’s only unwarned statement and the circumstances did not involve actual coercion before the interrogation continued at the station. Moreover, one hour had passed and Elstad had been confronted by his father, who spoke angrily to him. Id. at 301, 105 S.Ct. 1285. Thus, even if the prior circumstances would have been coercive, the passage of time would have diminished any police coercion and any coercion by Elstad’s father would have superceded any police coercion.

The circumstances of the current case are quite different. Bailey’s unwarned statements were made during interrogation in the detective’s car immediately after he was arrested at gunpoint, placed against the squad car, patted down for weapons, handcuffed, and placed in the back seat. These facts make it clear that Bailey was in “custody” and that the first interrogation was accompanied by actual coercion.

Further, we conclude that Scott is distinguishable because Scott’s non-Mirandized interrogation lasted only 15 minutes and because Scott made no incriminating statements during that time. Scott, 584 N.W.2d at 415. Bailey’s non-Mirandized *392interrogations took longer,8 and Bailey made substantially the same inculpatory statements before the warning as after.

Following Carter, we hold that where a suspect is apprehended under coercive circumstances, is subjected to lengthy custodial interrogation before being given a Miranda warning, does not have the benefit of a significant pause in the interrogation after the Miranda warning is given, and essentially repeats the same inculpatory statements after the Miranda warning as before, the statements made after the Miranda warning are inadmissible.

Our decision is consistent with the purposes of Miranda. If police are permitted to cure the illegality of a coercive unwarned custodial interrogation by merely providing the warning after they have already obtained inculpatory evidence, they would have little incentive to give the warning at the beginning of their custodial interrogation.9

This conclusion necessitates a harmless error analysis. Based on Dickerson, 530 U.S. at 438, 120 S.Ct. 2326, we conclude that this was constitutional error. When a constitutional error results in evidence being improperly admitted at trial, we ask whether the verdict rendered was “surely unattributable” to the constitutional error. State v. Chomnarith, 654 N.W.2d 660, 665 (Minn. 2003); see also Sullivan v. Louisiana, 508 U.S. 275, 278-82, 113 S.Ct. 2078, 124 L.Ed.2d 182 (1993).

None of Bailey’s statements constituted a confession, but all were dubious statements of fact that brought his credibility strongly into question. In his closing argument, the prosecutor capitalized on these statements by arguing that “[p]eople lie because they want to hide something.” In light of the significant impact of this evidence, we cannot say that Bailey’s conviction was “surely unattributable” to the admission of his statements. Accordingly, we reverse Bailey’s conviction and order a new trial, at which Bailey’s statements to police shall not be admitted in evidence. We will also address Bailey’s other claims of error because they concern issues that likely will arise in the new trial.

II.

In February 1992, police lieutenant Brad Johnson authorized the destruction of several pieces of physical evidence.10 *393On Johnson’s authorization, other items were released to Bailey’s wife in 1992 and were not available at trial.11 Bailey moved to preclude any reference to this evidence at trial. He now characterizes this as a motion under his Sixth Amendment right to confront witnesses and under Minn. R.Crim. P. 9.01, subd. I.12 The state argues that this issue is more properly framed as a due process argument. The district court denied Bailey’s motion, though it allowed Bailey to inform the jury about the destruction and release of the evidence.

In Arizona v. Youngblood, the United States Supreme Court held that “unless a criminal defendant can show bad faith on the part of the police, failure to preserve potentially useful evidence does not constitute a denial of due process of law.” 488 U.S. 51, 58, 109 S.Ct. 333, 102 L.Ed.2d 281 (1988). In State v. Koehler, 312 N.W.2d 108, 109 (Minn. 1981), we refused to overturn a conviction on grounds that physical evidence was “inadvertently lost before appellant had an opportunity to examine” it, concluding that (1) “there is no suggestion that the State intentionally lost the [evidence] to avoid discovery of evidence beneficial to the defense” and (2) “it is doubtful whether the [evidence was] exculpatory.” (Internal quotation marks and citation omitted.)

The district court in the instant case found that

the exculpatory value of the evidence, if any, would not have been apparent at the time it was destroyed. And also, that the destruction of the evidence was accidental and/or negligent, rather than intentional. While Lieutenant Johnson intentionally signed the property inventory sheets ordering the destruction of the evidence, he made it clear that this was a mistake on his part. Based upon the analysis of these factors there will be no “blanket sanction” prohibiting the State from introducing any testimony regarding the destroyed evidence.

We agree with Bailey that the destruction of the evidence was “intentional.” But, as in Koehler, there is no suggestion that the State destroyed or released items “to avoid discovery of evidence beneficial to the defense.” See Koehler, 312 N.W.2d at 109 (quoting Lee v. State, 511 P.2d 1076, 1077 (Alaska 1973)). Moreover, there is no indication that any of the destroyed or released items had exculpatory value. See id. We therefore hold that that the district court did not err in allowing testimony regarding the evidence that was destroyed and released, and such testimony will likewise be admissible in a new trial.

III.

Bailey challenges the state’s DNA evidence on six alternative grounds. Before discussing these grounds, some general description of the DNA evidence will provide a helpful context.

At trial, Bailey raised several concerns about the quality of the DNA samples. As the district court noted, two of the 1984 slides were reexamined in 2000: a vaginal *394smear slide and an oral smear slide. The oral smear slide was inadequate to yield any results and the vaginal slide also had significant deficiencies. Both slides were prepared after the victim had been, deceased for an estimated three days and thus likely contained bacteria detrimental to the preservation of the DNA. By 2000, the sample on the vaginal smear slide was so small that only one test, with a Profiler Plus kit, could be conducted. The sample was sixteen years old, the conditions of its storage were unknown and, as will be discussed in section E below, the slide was heated with a Bunsen burner to remove the cover slip. As a result of some or all of these factors, the profile obtained from the sample was incomplete: reliable results were obtained for only six of the ten loci tested. While the results at those six loci matched Bailey, all experts agreed that a person is excluded if he or she fails to match at any one locus.

Thus, Bailey argues that the conditions of the sample, whether by natural causes or from improper handling, deprived him of the opportunity to be excluded at one of the four uninterpretable loci tested by the Profiler kit, as well as the three additional loci that normally would be tested by a Cofiler kit. This argument heightens Bailey’s criticism of the state’s handling of the DNA testing and the use of Bunsen burner heat to remove the cover slip, the destruction of the sample in testing without notice to Bailey, and Bailey’s complaints about allowing the state’s expert to give interpretative testimony about the four uninterpretable loci.

A. Chain of Custody

Minnesota Rules of Evidence 901(a) (2002) provides: “General provision. The requirement of authentication or identification as a condition precedent to admissibility is satisfied by evidence sufficient to support a finding that the matter in question is what its proponent claims.” One year prior to the adoption of Rule 901(a), we held that

[tjhere can be no rigid formulation of what showing is necessary in order for a particular item of evidence to be admissible. Rather, admissibility must be left to the sound discretion of the trial judge. He must be satisfied that, in all reasonable probability, the item offered is the same as the item seized and is substantially unchanged in condition.

Admissibility should not depend on the prosecution negativing all possibility of tampering or substitution, but rather only that it is reasonably probable that tampering or substitution did not occur. Contrary speculation may well affect the weight of the evidence accorded it by the factfinder but does not affect its admissibility.

State v. Johnson, 307 Minn. 501, 504-05, 239 N.W.2d 239, 242 (Minn. 1976) (emphasis added) (internal citations omitted); see also State v. Hager, 325 N.W.2d 43, 44 (Minn. 1982) (reaffirming Johnson after the adoption of Minn. R. Evid. 901(a)).

Bailey identifies two problems with the chain of custody presented by the state. First, he argues that the state’s evidence concerning the initial filing of the DNA evidence was insufficient because the state failed to call the pathologist, Dr. Robert Ackerson, who performed the autopsy and made that filing. Second, he argues that the state “failed to establish a chain of custody for the biological sample after it was placed in [storage in] HCMC’s basement” but before it was retrieved when the investigation was restarted in 2000.

The state, however, provided evidence of Dr. Ackerson’s actions through the testimony of other witnesses and through business records prepared as part of the autopsy process. Moreover, BCA forensic *395scientist Catherine Knutson’s testimony— that the cover slip on the sample slide was “adhered” and “intact” and that the label was still affixed — prompted the district court to find that the state had shown an adequate foundation for the DNA evidence.

Viewing the chain of custody issue in isolation, we cannot say that the district court abused its discretion in finding that the state had met its foundational burden. See Johnson, 307 Minn, at 504-05, 239 N.W.2d at 242. However, we will consider the age and condition of the sample further when we analyze Bailey’s due process arguments below.

B. Loci Whose Peaks Have Low RFU Values

Knutson’s tests yielded interpretable DNA results at six of the ten loci that the Profiler Plus kit tests. These were represented by peaks on a graph called an electropherogram. Pursuant to BCA protocol, eleetropherogram peaks must be higher than 150 relative fluorescent units (RFUs) in order to be reliable enough to be interpretable. According to Knutson’s pretrial testimony, the BCA considers peaks of less than 150 RFUs to be unreliable.

Four of the loci did not yield interpretable DNA results because their peaks were less than 150 RFUs in height. During the trial cross-examination of Knutson, Bailey established that if. a person does not match at any one locus, he is excluded. Bailey’s counsel then asked Knutson:

Q And in this cáse you were able to answer the question for five locations on the gene; is that correct?

A Yes.

Q But for the other eight locations on the gene you were not able to answer the question because either the results were below your threshold or you didn’t do the Cofiler; is that right?

A Yes.13

Although Knutson had testified pretrial that peaks that were less than 150 RFUs were not interpretable, in the sense that they could not provide a match, the state claimed that Bailey’s cross-examination opened the door for some interpretation by implying “that, had Knutson done the test properly, Bailey could have been excluded at one of the eight loci not included in the DNA profile generated.” On redirect, the state asked Knutson a series of questions to show that while a locus with peaks under the 150 RFU threshold were not reliable enough to support a match, they can sometimes provide enough reliable information to exclude a person.

Bailey objected to this testimony and the use of the accompanying charts, arguing that the prosecution had failed to establish foundation for opinion evidence concerning any locus with a peak under 150 RFUs. The district court overruled the objection, holding that the redirect questions were within the scope of the cross-examination. Knutson testified that the relative placement of the peaks at these four loci was consistent with Bailey’s profile and thus did not exclude him.

Evidentiary rulings are subject to an abuse of discretion standard. State v. Robledo-Kinney, 615 N.W.2d 25, 29 (Minn. 2000). The district court’s decision that *396Bailey’s cross-examination opened the door to some limited discussion of the sub-150 RFU peaks was likely within the court’s discretion. But the court did not respond to Bailey’s foundation objection. On that point, Knutson’s testimony on the use of sub-150 RFU peaks appears to us unclear. Should this testimony be offered at the new trial, the district court should examine the foundation for any interpretation of electropherogram peaks that are less than 150 RFUs in height.

C. Product Rule

Bailey argues that Knutson’s testimony concerning probability statistics was inadmissible because it referred to the “product rule” rather than the “interim ceiling method” approved in State v. Bloom, 516 N.W.2d 159, 160 (Minn. 1994). In Bloom, we relied significantly on the recommendation of the National Research Council (NRC) that the “interim ceiling method” should be used for probability statistics. 516 N.W.2d at 160.

We have addressed this question in two other cases recently before this court. See State v. Miller, 666 N.W.2d 703 (Minn. 2003); State v. Roman Nose, 667 N.W.2d 386 (Minn. 2003). Essentially, we have concluded that, subsequent to Bloom, the NRC re-examined the issue and determined that the interim ceiling method is not appropriate to PCR-based systems. The NRC now recommends that the product rule be used. Accordingly, we conclude that the district court did not abuse its discretion in allowing probability statistics testimony based on the product rule.

D. Sample Exhaustion and Lack of Notice

On August 24, 2000, approximately 6 weeks before she conducted her tests, Knutson sent the prosecutor a letter regarding “a homicide case involving Bill Daymond Bailey and Agnes Mary Fafrow-icz.” The letter notified the prosecutor that Knutson’s laboratory would be conducting a test on the DNA sample in the Fafrowicz murder on or about October 2, 2000, and that the testing “may preclude any further tests or experiments.” The prosecutor did not notify Bailey or anyone associated with him about the testing. Knutson proceeded with her DNA tests. The testing exhausted the entire DNA sample.

Minn. R.Crim. P. 9.01, subd. 1(4), provides in pertinent part:

Reports of Examinations and Tests. * * * . If a scientific test or experiment of any matter * * * may preclude any further tests or experiments, the prosecuting attorney shall give the defendant reasonable notice and an opportunity to have a qualified expert observe the test or experiment.

Bailey argues that the spirit of the rule makes it applicable here, and that the appropriate sanction for the failure to notify him of destructive testing is to exclude all DNA evidence. The state argues that Rule 9.01, subd. 1(4), only applies to a “defendant” and that, at the time of the testing, Bailey was not a defendant because he was not then subject to a charge. The district court found that

the testing was clearly conducted during the investigation phase [rather than the prosecution phase] of the case. The Court finds this distinction significant and declines to burden the State with discovery obligations prior to the time a case is charged (even, as here, where the State is focused on a single suspect). The Court also rejects Defendant’s argument that the State failed to give notice in bad-faith, [sic] or for strategic purposes. There is simply no such evidence on the record. If there were, the *397Court would likely view this issue much differently.

In the alternative, the district court also found that Bailey was not prejudiced by the prosecution’s failure to notify him because he “would not have been entitled to a portion of the sample to conduct an independent test” and because “there is no evidence that,” even if Knutson had been accompanied by a defense expert, “the extraction or testing procedure followed by the BCA would have been any different.”

The construction of a procedural rule is a question of law subject to de novo review. Kastner v. Star Trails Ass’n, 646 N.W.2d 235, 238 (Minn. 2002). However, once a discovery violation has occurred, a district court “is particularly suited to determine the appropriate remedy and has wide discretion in deciding whether to impose sanctions.” State v. Freeman, 531 N.W.2d 190, 197-98 (Minn. 1995). “Absent a clear abuse of discretion, a reviewing court will not overturn the trial court’s decision. * * * [T]he preclusion of evidence is a severe sanction that should not be lightly invoked.” Id. at 198.

We recognize that there are policy arguments that would make it desirable to apply Rule 9.01 to these unusual facts where, although Bailey was not technically a defendant because the previous indictment had been dismissed, - he had once been charged, he remained the sole suspect, and the purpose of the testing was to compare.the sample taken from the victim to his DNA.14 But the precise words used in Rule 9.01 cannot be extended to apply to these facts. In fact, the history of the Rule suggests that the restriction to a presently charged “defendant” was deliberate. When the notice requirement of Rule 9.01 was adopted in 1989, the language was drawn in part from Rule 421 of the Uniform Rules of Criminal Procedure. Although Rule 421 proposed that notice be given to the defendant “and any person known or believed to have an interest in the matter,” which clearly would have included Bailey, Minnesota Rule 9.01 did not include these words and, instead, requires notice only to “the defendant.”

Thus, we conclude that the district court did not err in ruling that Rule 9.01 did not apply to Bailey. But we hasten to caution prosecutors that there may be due process implications from a decision to not give notice where there is a well-identified suspect who is the primary object of the DNA testing and who will surely become a defendant if that testing confirms a match to his DNA. In that connection, we disagree with the district court’s conclusion that Bailey was not prejudiced by the lack of notice. To the contrary, it would have been extremely valuable for Bailey to have an expert present to observe that testing, particularly the application of Bunsen burner heat, as discussed next.15

E. Bunsen Burner

The proponent of scientific evidence has the burden to establish the proper foundation for the admissibility of a scientific test by showing that the method*398ology used is generally reliable and that it produced reliable results in the specific case. Goeb v. Tharaldson, 615 N.W.2d 800, 816 (Minn. 2000). In Minnesota, the two-pronged Frye-Mack test must be satisfied before such evidence may be admitted. State v. Traylor, 656 N.W.2d 885, 891 (Minn. 2003). That test proceeds as follows:

First, a novel scientific technique that produces evidence to be admitted at trial must be shown to be generally accepted within the relevant scientific community, and second, the particular evidence derived from the technique and used in an individual case must have a foundation that is scientifically reliable. Put another way, the Frye-Mack standard asks first whether experts in the field widely share the view that the results of scientific testing are scientifically reliable, and second whether the laboratory conducting the tests in the individual case complied with appropriate standards and controls.

State v. Roman Nose, 649 N.W.2d 815, 818 (Minn. 2002) (citations omitted). The standard of review in Frye-Mack analyses is also two-pronged:

Whether a particular principle or technique satisfies the first prong, general acceptance in the relevant scientific field, is a question of law that we review de novo. District court determinations under the second prong, foundational reliability, are reviewed under an abuse of discretion standard, as are determinations of expert witness qualifications and helpfulness.

Goeb, 615 N.W.2d at 815 (internal citations omitted).

The DNA samples in this case were stored on two slides. The samples were protected by cover slips that were adhered to the slides by a mounting medium. BCA forensic scientist Knutson removed the cover slips by heating the slides “over a low heat” for 30 seconds, using a Bunsen burner, until the mounting medium boiled and loosened.

At a pretrial hearing, Knutson conceded that this method of removing a cover slip had “not been studied” and that “[i]t can lead to degradation” of the DNA sample. Defense expert witness Dan E. Krane testified that the Bunsen burner method has not been validated and that, under the standards of the DNA Advisory Board (DAB), the use of such a technique without a validation study would violate guidelines and could cause a laboratory to lose its accreditation.16 In Traylor, we held that the DAB standards are the appropriate ones to govern DNA testing. 656 N.W.2d at 900.

The district court concluded that “[t]he record is very clear that the BCA has not done a validation study on the use of Bunsen burners to heat slides for purposes of removing cover slips to extract DNA. The reason why such a study has not been done remains a mystery.” The court found the BCA’s failure to conduct a validation test “troubling,” but it admitted the DNA evidence based upon testimony from state experts Knutson and James Iverson that (1) the Bunsen burner method “had been successfully used in 10-15 previous extractions conducted by the BCA”; (2) the method “was still in use in the BCA’s laboratory”; and (3) Iverson “feels the procedure is appropriate and reliable.”

*399The court found it significant that all expert witnesses agreed that “there are no studies which have found that it is possible to alter one complex DNA profile to another complex DNA profile.” And the court stated that “it would be pure speculation for the Court to conclude that the Bunsen burner was primarily responsible for the partial DNA profile obtained.” “In conclusion,” the court said, “the State has narrowly met its burden of demonstrating reliability under Goeb.” Because the court’s decision relied heavily on Iverson’s claim that the Bunsen burner technique had been used successfully in 10-15 previous cases, the court granted Bailey’s motion for discovery on those cases.

By the time of trial, it appeared that two of the key factors relied upon by the district court in its pretrial ruling were, or turned out to be, doubtful. First, the court’s emphasis on the expert agreement that a DNA profile cannot be altered to fit a different profile missed the main point of Bailey’s argument. Bailey’s concern was not that someone else’s DNA sample may have been altered to match his. Instead, his concern was that someone else’s DNA sample may have been degraded, possibly by the burner technique, to the point that it was indistinguishable from his because reliable interpretation could not be made at certain loci that might have excluded him.

Second, the state failed to substantiate Iverson’s claim of 10-15 successful uses of the Bunsen burner technique by the BCA. The state was able to produce in discovery only five “Burner cases.”17 Bailey argued at trial that those cases evidenced sample degradation similar to that of the sample in this case.

Also, after the pretrial ruling, Iverson ordered a validation study on the Bunsen burner method. At trial, he testified that the results indicated that “it took anywhere from 60 to 90 seconds of direct holding of that slide in the flame to create the point where you couldn’t get a full DNA profile * * *, and again it was holding it right in the flame.” In contrast, Knutson testified that, in this case, she only held the slide at the edge of the flame for 30 seconds. Bailey argues that the validation study used only pristine DNA samples, unlike the 16-year-old sample in this case, and was not reliable. Because the study had not been conducted until after the pretrial hearings, and the results had not been provided to Bailey or the court at the time of Bailey’s motion to reconsider the motion to suppress DNA evidence, it was not considered by the court when deciding Bailey’s motion to suppress the DNA evidence.

We hold that the district court’s findings are insufficient to satisfy fully the second prong of the Fry e-Mack standard for the admission of DNA results where the Bunsen burner technique has been used. The second prong requires proof of foundational reliability of the operating procedure as actually applied in the specific case. The district court did find the Bunsen burner technique was reliable, but that determination was tenuous because of the absence of validation studies and the reliance upon the state’s claim of 10-15 successful uses, a claim that remains unsubstantiated. Further, the district court did not address the factual dispute concerning whether the *400DAB standards require validation studies for such an operating procedure. Finally, the district court’s holding that “it would be pure speculation for the Court to conclude that the Bunsen burner was primarily responsible for the partial DNA profile obtained” (i.e., that it damaged the sample), improperly shifts the foundational burden away from the state.

We therefore direct that a Frye-Mack second prong hearing on the Bunsen burner procedure be held to determine whether any DNA evidence is admitted at the new trial. See Roman Nose, 649 N.W.2d at 822-23 (requiring a Frye-Mack hearing on remand).

F. PCR-STR Testing; Due Process Issues

Bailey’s final challenge to the DNA evidence is that the Polymerase Chain Reaction and Short Tandem Repeats (PCR-STR) method and the Profiler Plus kit used by Knutson do not meet the Frye-Mack standards of general acceptance in the scientific community and foundational reliability, as held by the Minnesota Court of Appeals in Traylor, 641 N.W.2d at 339-41. Bailey further adopts the argument made in Traylor, that even if the PCR-STR method meets the Frye-Mack standards, its use violates due process. Subsequent to the filing of Bailey’s brief in this appeal, we reversed the court of appeals’ decision in Traylor and held that the PCR-STR method and the Profiler Plus and Cofiler kits satisfied the Frye-Mack requirements of general acceptance in the scientific community and foundational reliability. Traylor, 656 N.W.2d at 900. We also rejected Traylor’s due process argument. Id. But there were facts in Traylor that were critical to our due process analysis that are absent here and that necessitate further review of this argument.

Traylor was convicted of second-degree assault and controlled substance possession in connection with a stabbing incident. Id. at 887. A key piece of evidence against Traylor was a DNA sample from a bloody knife found at the scene of the crime. Id. Traylor challenged the district court’s admission of the DNA evidence, arguing that (1) the BCA’s use of PCR-STR testing failed to meet the Frye-Mack standard and (2) the same testing violated his due process rights. Id. at 887, 898. Central to both of Traylor’s arguments was the fact that he was not given access to the genetic primer sequences in the Profiler and Cofi-ler kits because those sequences are guarded as proprietary information by the kits’ manufacturer. Id. at 890.

We summarized the parties’ due process arguments as follows:

Traylor relies heavily on State v. Schwartz in making this constitutional challenge. In Schwartz, we recognized that “[t]he fair trial and due process rights [under the Constitution] are implicated when data relied upon by a laboratory in performing tests are not available to the opposing party for review and cross examination.” We further noted in Schwartz that “[i]deally, a defendant should be provided with the actual DNA sample(s) in order to reproduce the tests” and that if that is not possible, “access to the data, methodology, and actual results is crucial so a defendant has at least an opportunity for independent expert review.” Our Minnesota discovery rules also echo the concerns set forth in Schwartz. Rule 9.01, subd. 1(4) of the Minnesota Rules of Criminal Procedure provides that defense counsel has a right to inspect results of scientific tests and that, if a test precludes any further testing, the defense must receive reasonable notice and an opportunity to have a qualified expert observe the test.

*401The state contends that the BCA’s policy fully addresses the access to information concerns stated in Schwartz and in the discovery rules. Specifically, the state points to the BCA’s policy, which requires that, when possible, a portion of the evidence sample be retained at the BCA laboratory. If the entire sample must be used, the BCA’s policy requires the scientist to notify the prosecuting attorney so that the defense has the opportunity to have its own expert observe the testing. In this case, as is common in PCR-STR testing situations, the DNA sample was not consumed, and a portion of the DNA sample was available for further testing by Traylor. Further, the state contends that Traylor had full access to all information in the BCA’s possession. Such available information included extensive documentation of the BCA’s work, including methodology, actual results of all testing, and compliance with standards and controls. Therefore, the state asserts, the BCA has provided Traylor with sufficient access to the laboratory’s testing data and results for review and cross-examination, allaying constitutional concerns.

Traylor, 656 N.W.2d at 898 (quoting State v. Schwartz, 447 N.W.2d 422, 427-28 (Minn. 1989)) (internal citations omitted).

Athough the author would conclude that, on this record, the admission of the DNA evidence violated Bailey’s due process rights, the majority of the court does not agree. The holding of the majority is that the concerns expressed about the age and condition of the DNA sample, the lack of notice to Bailey of destructive testing, the use of the Bunsen burner technique to remove the cover slip from the lab slide, and the unavailability of the genetic primer sequences of the Profiler kit do not, individually or collectively, rise to the level of a due process violation. Instead, Bailey’s complete access to the data, methodology and results of the BCA’s DNA tests satisfied his due process rights under Schivartz, 447 N.W.2d at 427 (Minn. 1989) and State v. Jobe, 486 N.W.2d 407, 419 (Minn. 1992). A more complete discussion of the majority’s conclusion on this issue is contained in the concurrence and dissent of Justice Russell Anderson.

IV.

On July 23, 1985, Bailey pleaded guilty to three burglaries he committed in April 1985, about one year after the death of Agnes Fafrowicz. In connection with his plea, Bailey described the facts of three burglaries. On April 1, 1985, Bailey broke into the South Minneapolis house of an 83-year-old woman, “confronted” her, “hit her,” took her money and left. Later that same day, Bailey forced his way into the home of a 69-year-old woman, pushed her to the ground, cut her with a knife, and fled without stealing anything. On April 19, 1985, Bailey broke into another house in South Minneapolis, carrying a knife and intending to steal items from the house. He was caught by a Minneapolis police officer when he exited the building. A1 three burglaries occurred within a short distance of Bailey’s South Minneapolis residence.

The state asserts, and Bailey does not dispute, that Fafrowicz “lived just around the corner from Appellant” at the time she was murdered. Over Bailey’s objection, the state was allowed to submit as Spreigl evidence Bailey’s statements made in connection with his plea of guilty to each of the April 1985 burglaries. See State v. Spreigl, 272 Minn. 488, 139 N.W.2d 167 (1965).

This court reviews a district court’s decision on whether to admit Spreigl evidence for abuse of discretion. *402State v. Kennedy, 585 N.W.2d 385, 389 (Minn. 1998). Evidence of other crimes cannot be used to show the defendant’s character for committing those crimes but can be used to show motive, intent, absence of mistake, identity, or a common scheme or plan. Id.; see also Minn. R. Evid. 404(b). Before such evidence may be admitted, the state must establish (1) by clear and convincing evidence that the defendant participated in the other crime; (2) that the evidence of the other crime is relevant and material to the state’s case; and (3) that the probative value of the evidence of the other crime outweighs its potential for unfair prejudice. Kennedy, 585 N.W.2d at 389. The questions here are whether the Spreigl evidence was relevant to the state’s proof of identity and whether the probative value of the evidence outweighed its potential for unfair prejudice.

Bailey argues that the factual nexus between the burglaries and the murder is weak because the burglaries “oe-curr[ed] nearly a year after Fafrowicz’s murder,” and because burglary and murder while committing sexual assault are crimes that are “not even of the same generic type * * *.” Bailey also argues that the Spreigl evidence had no probative value in this case because the other identity evidence offered by the state is strong, quoting State v. DeWald, 464 N.W.2d 500, 504 (Minn. 1991) (“This court has stated that Spreigl evidence is admissible only if the trial court finds the direct or circumstantial evidence of defendant’s identity is otherwise weak or inadequate, and that it is necessary to support the state’s burden of proof.” (internal quotations and emphasis omitted)).

We conclude that there is a strong factual nexus between the crimes and the instant charge. The burglaries all involved home invasions at locations near Bailey’s residence. In two of the three burglaries, Bailey attacked lone female homeowners who were over age 60. Two of the burglaries involved knives. The probative value of this evidence upon the question of identity is strong, whereas the prejudicial effect of the evidence appears no more significant than in any case of admissible Spreigl evidence.

We cannot say that the district court abused its discretion in finding the Spreigl evidence relevant on the question of identity or in finding that the evidence’s probative value outweighed its prejudicial effect.

V.

Finally, Bailey argues that there were eight instances of prosecutorial misconduct that require reversal. Because we have granted a new trial on other grounds, we need not determine whether any of these instances was sufficiently prejudicial to warrant reversal of the conviction.18 But we will comment on each of them to provide guidance for a new trial.

Bailey argues that the prosecutor misstated the state’s burden of proof on four occasions. Misstatements of the burden of proof are improper and constitute prosecutorial misconduct. State v. Hunt, 615 N.W.2d 294, 302 (Minn. 2000).

First, the prosecutor’s opening statement referred to DNA probability evidence, stating that “99.999976 [percent] of *403the population of all human beings are expected to be excluded” from the tests in the instant case, but Bailey was not excluded, and then raised the rhetorical question, “Is that proof beyond a reasonable doubt?” We have cautioned the state to avoid any attempt to equate DNA probability statistics with proof beyond a reasonable doubt. See State v. Carlson, 267 N.W.2d 170, 176 (Minn. 1978) (citing Laurence Tribe, Trial by Mathematics, 84 Harv. L.Rev. 1329 (1971)) (“Testimony expressing opinions or conclusions in terms of statistical probabilities can make the uncertain seem all but proven, and suggest, by quantification, satisfaction of the requirement that guilt be established ‘beyond a reasonable doubt.’ ”); see also Bloom, 516 N.W.2d at 169 (“Prosecutors and trial courts are cautioned that we will not hesitate to award a new trial to a defendant if our review of the trial record reveals that quantitative or qualitative DNA identification evidence was presented in a misleading or improper way.”). We conclude that the prosecutor’s statements were improper and should be avoided in a new trial.

Second, during closing argument, the prosecutor urged the jurors, “as the truth seekers, search for the truth in the evidence; don’t search for doubt in the evidence, because you are the truth seekers. Search for the truth in the evidence, but give the defendant the benefit of any reasonable doubt.” Bailey argues that a jury’s real role requires “ ‘searching for doubt’ in all the evidence.” The state maintains that similar closing arguments were deemed appropriate in State v. Ashby, 567 N.W.2d 21, 28 (Minn. 1997) (“The prosecutor did tell the jury to ‘keep its eyes on the prize’ of truth * * *.”), and State v. Atkins, 543 N.W.2d 642, 648 (Minn. 1996) (noting that the prosecutor told the jury that it “would be an ‘unspeakable injustice’ to consider the lesser-included offenses and to acquit on the charge of first-degree murder”), and that the court’s subsequent instruction removed any taint. We conclude that the prosecutor’s statements, while somewhat confusing, were not a clear misstatement of the burden of proof and, in such event, would be super-ceded by the court’s jury instructions on burden of proof.

Third, the prosecutor twice argued in closing that the only question before the jury was the identity of Fafrow-icz’s attacker. Bailey points out that he did not concede any of the statutory elements of the crime. The state agrees that this was “technically error.”

Fourth, during the direct interrogation of Detective Nelson, the prosecutor asked, “So, [Bailey’s] only alibi witness [his wife] was out of state at the time?”. During his closing argument, the prosecutor again made a reference to Bailey’s wife “not [being] home with him, not able to indicate to the officers as to his whereabouts on the night of May 16th or May 17th * * Bailey suggests that these references improperly implied that Bailey had some responsibility to provide evidence of an alibi. We conclude that the prosecutor’s reference to the defendant’s “only alibi witness” was improper and should be avoided in a new trial.

Bailey next complains that the prosecutor made three comments that were inappropriately disparaging of Bailey’s defense, his counsel and his expert witness. We have “repeatedly warned prosecutors that it is improper to disparage the defense in closing arguments * * * .” State v. Griese, 565 N.W.2d 419, 427 (Minn. 1997); see also State v. Powers, 654 N.W.2d 667, 679 (Minn. 2003).

First, the prosecutor disparaged defense DNA expert Mueller. In the *404opening statement, the prosecution said that Mueller “continues to walk around the country advocating the use of the counting method [as opposed to the Product Rule], coming in, in criminal cases like this, because he gets paid for it.” In the closing argument, the prosecutor said that “[Mueller] uses that soft language of ‘consultant’ when, in fact, all he is, is a paid witness by the Defense in criminal cases.” We conclude that it was improper for the prosecutor to go beyond the testimony of the expert witness by making these references to the witness’s character.

Second, the prosecutor elicited testimony from Detective Nelson implying that defense counsel was from the Public Defender’s office. We view the reference as inadvertent, but agree it should be avoided in a new trial.

Third, during closing argument, the prosecutor told the jury that the defense case amounted to the allegation that “somebody involved with the State surreptitiously snuck the defendant’s DNA under that cover slip,” a contention he likened to that made by O.J. Simpson in his murder trial. This kind of comparison to the Simpson case has been held improper by this court. See State v. Thompson, 578 N.W.2d 734, 743 (Minn. 1998). Closing argument must be confined to the record and this comparison obviously goes outside the record. This was misconduct and should be avoided in a new trial.

Finally, Bailey complains that the state appealed to the passion and prejudice of the jury. At the end of the prosecution’s closing argument, the prosecutor stated that “the prosecution brought this back up, to right a wrong. This family deserves resolution to this case * * * . We have held that prosecutors must avoid inflaming the jury’s passions and prejudices against the defendant. State v. Porter, 526 N.W.2d 359, 363 (Minn. 1995) (citing State v. Morgan, 235 Minn. 388, 391, 51 N.W.2d 61, 63 (1952)). This argument should be avoided in the new trial.

Reversed and remanded for further proceedings consistent with this opinion.

(concurring in part, dissenting in part).

I respectfully concur in part and dissent in part. In light of all the events and conditions surrounding Bailey’s statements, including the entirety of the police conduct and Bailey’s experience in the criminal justice system, in my view, Bailey’s statements were voluntary. As for DNA evidence, I see no error in the admission of the results of the DNA analysis. There is a difference between laboratory procedures aimed at ensuring the integrity of physical evidence and validated methods for forensic casework analyses aimed at ensuring the reliability of the DNA results of the analysis in any given case. I would affirm.

During the course of the investigation of the death of Agnes Fafrowicz, homicide investigators learned that the victim’s checkbook was missing. They also found the victim’s car with the keys still in the ignition in a parking lot about a block away from her home. On May 22, 1984, Officers Snobeck and Nelson “eyeballed” the car, a 1970 Ford Custom 500 with drum brakes, for “ideas, something missing, something hidden there.” They looked under the hood and noticed a corroded red and white Atlas battery located on the passenger’s side. The officers then walked in the tall grass around the victim’s house looking for “trace evidence, anything that might have been missed.” At some point during the morning of May 22, the officers were notified that two checks had cleared the victim’s bank account; one was dated May 13, 1984 and made payable to *405an insurance company and the other, dated May 17, 1984 and made payable to “Bill Vollmar-Bailey,” was cashed at a liquor store close to the victim’s residence and, according to the bank, appeared to be a forgery. The estimated date of the victim’s death was May 16, 1984. The officers went to the bank, obtained the two checks, returned to the police department and made photocopies of the checks. The officers also ran Bailey’s name through the local records to see if it was in the system and when that came up empty, the officers went to the liquor store where the check had been cashed.

At the liquor store, the officers talked to the owner who told them that Bill Bailey lived in the apartment building next door. The owner also said that Bailey told him that he (Bailey) had “done” 17 years in an Oklahoma prison and had been recently released. As the officers were talking to the owner, Bailey walked through the parking lot and into the apartment building and the liquor store owner pointed him out as the person who had cashed the check. The officers called the police department to check Bailey’s record in Oklahoma and as they were doing that, Bailey quickly left the apartment building and started walking up the street smoking a cigarette.

The officers got into their unmarked vehicle,19 drove down the street, turned the corner and pulled up on the “wrong side of the road” next to where Bailey was walking. The officers stopped Bailey at gunpoint, identified themselves, patted down Bailey’s outer clothing for weapons, took his Camel cigarette, handcuffed him and placed him in the vehicle.20 The officers confronted Bailey with the check made out to him and their suspicion that it was a forgery. Bailey said the check was valid and that he received it for cutting the victim’s grass and working on her car cleaning the battery and doing a brake job. He said the car had disc brakes. He said the victim wrote the check out to him on Friday, May 18, 1984, which was two days after she died. The officers transported Bailey to the police department, took off the handcuffs and seated him in the interview room where Bailey was allowed to smoke his Camel cigarettes.

At the police department, the officers again questioned Bailey as to how he came into possession of the check and Bailey gave essentially the same information. About ten minutes into the questioning, when Bailey indicated that the battery he cleaned was on the driver’s side of the car, Officer Nelson told Bailey he was under arrest for murder and advised him of his Miranda rights. Officer Nelson asked Bailey if he understood those rights. Bailey said he did. He said he was familiar with the law, having had access to a law library in prison. Bailey agreed to waive his rights and provided essentially the same information, adding that he wasn’t a mechanic and only cleaned the brakes and that he’d talked to the victim about painting her house.21

*406In the execution of a search warrant at Bailey’s apartment, Officers Snobeck and Nelson found $230 under some clothing in Bailey’s dresser drawer and a flashlight, later identified by the victim’s daughter as belonging to her mother. A comparison of the shoes worn by Bailey at the time of arrest with shoeprints found on the windowsill and siding of the victim’s house indicated that the tread patterns were “grossly similar,” meaning that Bailey’s shoes could not be eliminated; and forensic testing did not eliminate Bailey as the source of saliva found on a Camel cigarette butt found in the victim’s house and the Camel cigarette butts collected during the station house interrogation.

In June 1984, Bailey was indicted by grand jury for first-degree murder. The indictment was dismissed in November 1984. The case was reopened in 2000, DNA testing indicated a nexus between Bailey and the crime, and a second grand jury indicted Bailey for the same offense. Following an omnibus hearing, the district court denied Bailey’s motions to suppress the statement given after the Miranda warning and DNA evidence.

Post-Miranda Statement

Because of the coercion inherent in custodial interrogations, a suspect must be “warned prior to any questioning that he has the right to remain silent, that anything he says can be used against him in a court of law, that he has the right to the presence of an attorney, and that if he cannot afford an attorney one will be appointed for him prior to any questioning if he so desires.” Miranda v. Arizona, 384 U.S. 436, 479, 86 S.Ct. 1602, 16 L.Ed.2d 694 (1966).22 Statements obtained in the absence of a Miranda warning are, with certain exceptions, inadmissible. Id.

But when a suspect’s initial statement is obtained in violation of Miranda, the admissibility of a subsequent statement made after Miranda warnings is governed by Oregon v. Elstad, 470 U.S. 298, 105 S.Ct. 1285, 84 L.Ed.2d 222 (1985). The Court in Elstad concluded that in the absence of actual coercion, a Miranda violation need not bar the admission in evidence of subsequent, properly warned statements. Elstad, 470 U.S. at 309, 105 S.Ct. 1285. The Court reasoned that errors in the administration of Miranda procedures

should not breed the same irremediable consequences as police infringement of the Fifth Amendment. It is an unwarranted extension of Miranda to hold that simple failure to administer the warnings, unaccompanied by any actual coercion or other circumstances calculated to undermine the suspect’s ability to exercise his free will, so taints the investigatory process that a subsequent voluntary and informed waiver is ineffective for some indeterminate period. Though Miranda requires that the unwarned admission must be suppressed, the admissibility of any subsequent statement should turn in these circumstances solely on whether it is knowingly and voluntarily made.

Id.

The Elstad Court said that the “subsequent administration of Miranda warnings *407to a suspect who has given a voluntary but unwarned statement ordinarily should suffice to remove the conditions that precluded admission of the earlier statement.” Elstad, 470 U.S. at 314, 105 S.Ct. 1285. The Court continued, stating that “[i]n such circumstances, the finder of fact may reasonably conclude that the suspect made a rational and intelligent choice whether to waive or invoke his rights.” Id. The admissibility of the second or subsequent statement is decided by reviewing the totality of the circumstances: “[T]he finder of fact must examine the surrounding circumstances and the entire course of police conduct with respect to the suspect in evaluating the voluntariness of his statements. The fact that a suspect chooses to speak after being informed of his rights is, of course, highly probative.” Id. at 318, 105 S.Ct. 1285. We follow the Elstad rule. State v. Scott, 584 N.W.2d 412, 419 (Minn. 1998); State v. Moorman, 505 N.W.2d 593, 600 (Minn. 1993).

The voluntariness of a statement depends on the totality of the circumstances. State v. Patricelli, 357 N.W.2d 89, 92 (Minn. 1984); State v. Jungbauer, 348 N.W.2d 344, 346 (Minn. 1984). Relevant factors include the defendant’s “age, maturity, intelligence, education and experience,” as well as the defendant’s ability to comprehend. Jungbauer, 348 N.W.2d at 346 (citing State v. Linder, 268 N.W.2d 734, 735-36 (Minn. 1978)). The nature of the interrogation is also relevant, including its length and surrounding circumstances, and whether the defendant was denied any physical needs or access to friends. Id. A statement is involuntary if police actions were so coercive, manipulative and overpowering as to “ ‘deprive[ ] [a suspect] of his ability to make an unconstrained and wholly autonomous decision to speak as he did.’ ” State v. Ritt, 599 N.W.2d 802, 809 (Minn. 1999) (quoting State v. Pilcher, 472 N.W.2d 327, 334 (Minn. 1991)). The trial court resolves testimonial disputes as to the historical facts, and the appellate court independently determines, on the basis of all factual findings that are not clearly erroneous, whether or not the confession was voluntary. State v. Anderson, 396 N.W.2d 564, 565 (Minn. 1986) (citing Miller v. Fenton, 474 U.S. 104, 106 S.Ct. 445, 88 L.Ed.2d 405 (1985); State v. Linder, 268 N.W.2d 734 (Minn. 1978); Doan v. State, 306 Minn. 89, 234 N.W.2d 824 (1975)).

Here, in considering the totality of the circumstances on the facts as found by the district court, in my view, Bailey’s statements were voluntary. Bailey was 29 years old. He had significant prior experience with the criminal justice system, including Miranda warnings and waivers of those rights in connection with offenses dating back to the early 1970’s. He had recently been released from prison and he informed the police that he was familiar with the law, having had access to the prison law library. The administration of the Miranda warning before the third statement was careful and complete. Although in custody, Bailey was not subject to any physical deprivations and was allowed to smoke cigarettes during his statements at the police station. The entire transaction, from the detention on the street to the end of the third statement at the police station, lasted only about an hour and a half. In addition, “given that [Bailey] did not in fact confess, it cannot be said that [his] will was overborne by the [police] questioning.” State v. Mills, 562 N.W.2d 276, 284 (Minn. 1997) (citing Pilcher, 472 N.W.2d at 334 (“That he adhered to this woven tapestry of lies shows that Pilchers will was not overborne.”)).

The majority distinguishes Elstad, noting that unlike the suspect in Elstad, Bailey was apprehended under coercive circumstances. [A]bsent deliberately coercive or improper tactics in obtaining *408the initial statement, the mere fact that a suspect has made an unwarned admission does not warrant a presumption of compulsion. Elstad, 470 U.S. at 314, 105 S.Ct. 1285. The circumstances of Baileys arrest on probable cause for a homicide only established his custodial status and the need for a Miranda warning; furthermore, the location of the initial questioning did not rise to the level necessary to nullify the voluntariness of Baileys statements. Cf. United States v. Doe, 149 F.3d 634, 639 (7th Cir. 1998) (Miranda waiver and confession voluntary despite defendant being questioned while handcuffed in police car). Here, the district court found that the police did not use deliberately coercive or improper tactics in obtaining Baileys statements.

As for the majority’s reliance on the similarity of the warned and unwarned statements as bearing on the admissibility of the warned statement, it seems to me that kind of analysis comes very close to embracing the taint or cat-out-of-the-bag analysis rejected by Elstad, 470 U.S. at 303-04, 105 S.Ct. 1285; see also Scott, 584 N.W.2d at 419 (recognizing that the traditional taint analysis does not apply to Miranda violations). Finally, regarding the majority’s reliance on the absence of a significant pause between the unwarned and warned statements, that too comes close to the break-in-the-stream-of-events analysis also rejected by Elstad, 470 U.S. at 310, 105 S.Ct. 1285.23

Inasmuch as the district courts findings related to the voluntariness of both the unwarned and warned statements are not clearly erroneous, in considering the totality of the circumstances, the unwarned and warned statements were voluntary. In the words of the court that ruled on the same issue in connection with the original indictment in 1984, [i]t would be sheer fiction to conclude otherwise. Before the warning of his rights, [Bailey] said the check was for work on the victims yard and car. After the warning of his rights, rights which he knew, he said the check was for work on the victims yard and car. His explanation was voluntary. The decision of the district court to admit the post-Miranda statement should be affirmed.

DNA: Bunsen Burner

Bailey argues that the DNA test results should have been excluded because the state failed to establish that the BCA followed proper protocols in removing the cover slip from the forensic-sample slide. The evidence was that following visual inspection of the vaginal smear slide, the BCA analyst determined that biological material was adhered to the slide. The analyst then removed the cover slip by using a Bunsen burner to heat the bottom of the slide for approximately 30 seconds, when the mounting media had softened to *409the point that it had just begun to liquefy and the cover slip loosened. The analyst swabbed the biological material off the slide and prepared it for DNA extraction. The analyst then extracted the DNA from the sample and tested it with the Profiler Plus kit. Results were obtained for five of the nine loci tested, plus the amelogenin (the gender gene).24

The BCA had protocols for removing forensic samples from slides, including freezing and prying a cover slip off a slide or soaking the slide in Xylene for several hours; but none of the documented protocols involved heat with a Bunsen burner. Given the common usage of glass slides with cover slips, the district court found it “troubling” and even “regrettable” that no “validation study” had been done for the use of a Bunsen burner flame to remove a cover slip from a glass slide.

Nevertheless, the court allowed the DNA evidence. The court considered that scientific studies documented the use of boiling water to remove a cover slip from a glass slide. The court considered testimony from experts for the state and defense who agreed that mounting media affixing a cover slip to a slide would melt at the same temperature regardless of whether the heat came from boiling water or a Bunsen burner flame; they agreed that environmental insults could degrade or destroy a sample; and they agreed that while heat could degrade DNA to the point that no profile could be obtained, it could not alter the profile. The court also considered the likelihood that the sample had been degraded for reasons other than heat from the Bunsen burner; the victim had been dead for three days before the sample was collected; the sample was over 16 years old; and the conditions of storage were unknown. The court concluded that the evidence established the reliability of the DNA test results at the five interpretable loci; and the court further concluded that the fact the results were not interpretable at the remaining loci was an issue that went to the weight of the evidence and not to its admissibility.25 The court revisited the DNA evidentiary ruling well before trial and determined that the state had demonstrated that the method of extraction was scientifically valid.26

*410The Federal Bureau of Investigation’s Quality Assurance Standards for Forensic DNA Testing Laboratories require laboratories to have procedures “to ensure the integrity of the physical evidence.” (Standard 7.1). Accordingly, laboratories that engage in forensic DNA testing must follow “documented procedures that minimize loss, contamination, and/or deleterious change of evidence.” (Standard 7.1.3). The FBI Standards also require laboratories to “use validated methods and procedures for forensic casework analyses” (Standard 8.1), including “[developmental validation.” (Standard 8.1.1). , The FBI Standards also require laboratories to follow written analytical procedures, including “a procedure for differential extraction of stains that potentially contain semen.” (Standard 9.1.3).

It seems to me that in Bailey’s case, the method of removing the cover slip from the glass slide had more to do with ensuring the integrity of the physical evidence that was tested rather than the methods and procedures employed for the forensic analysis that require the kind of validation necessary to assure the reliability of the results of that analysis. Certainly, procedures for handling the physical evidence need to be “validated” to assure that such procedures do not harm the evidence; but I believe the district court appropriately concluded that the degradation of the evidence was an issue of weight or value to be given the evidence by the jury and not the admissibility of the analysis of the degraded DNA. Rulings on evidentiary matters rest within the sound discretion of the trial court, and we will not reverse such eviden-tiary rulings absent a clear abuse of discretion. State v. Chomnarith, 654 N.W.2d 660, 665 (Minn. 2003). I believe the admission of the DNA evidence was well within the district court’s discretion; and given that a validation study constructed along the lines suggested by the expert for the defense has been accomplished and was presented at trial, I see little point in remanding this issue for further litigation.

DNA Due Process

The dissent of Justice Meyer concludes that admission of the DNA evidence violated Bailey’s due process rights because he was not given access to the genetic primer sequences in the Profi-ler/Cofiler kits. Due process concerns are implicated “ ‘when data relied upon by a laboratory in performing tests are not available to the opposing party for review and cross examination.’ ” State v. Traylor, 656 N.W.2d 885, 898 (2003) (quoting State v. Schwartz, 447 N.W.2d 422, 427 (Minn. 1989)). In rejecting the same claim in Traylor, we noted that the BCA did not have the primer sequences when it performed DNA analysis using the kits in that case:

Instead, through the use of its own testing of the kits, the BCA validated that *411the kits produce reliable results. Tray-lor likewise could have obtained the kits and performed the same type of validation testing as the BCA laboratory. Moreover, Traylor could have perused any number of publicly available validation studies that have been performed on these kits since their inception. With the DAB standards and procedures to guide him, Traylor could have also questioned the BCA technicians about the procedures and methodology followed, their validation studies, and their interpretation of the results. Traylor did not need the primer sequences or unlimited access to Perkin-Elmer’s validation studies to do so. Finally, and importantly, there was a portion of the DNA sample at issue available for Traylor to perform his own tests, an opportunity Traylor did not pursue.

Traylor, 656 N.W.2d at 899-900 (footnote omitted).

The dissent of Justice Meyer ' distinguishes Traylor factually, concluding that due process was violated here where the sample was so small and degraded that it yielded only a partial profile, with interpretable results having been obtained at only five of the nine loci, and where Bailey had neither notice as to the consummation of the entire sample through testing nor access to a portion of the sample. But the issue in Traylor involved access to the primer sequences which, in turn, was relevant only to the reliability of the Profi-ler/Cofiler test kits. Neither the parties nor this court thought it pertinent to examine the quality and size of the DNA sample in that case as it related to the primer sequences and validation studies on those kits. True, we noted that a portion of the forensic sample was available for independent testing in Traylor, but that had more to do with the Schwartz/Jobe obligation of the BCA laboratory to make available its testing data and results. See Schwartz, 447 N.W.2d at 427; State v. Jobe, 486 N.W.2d 407, 419 (Minn. 1992). To the extent the reliability of the Profi-ler/Cofiler test kits is in dispute here, as in Traylor, Bailey had access to the data, methodology and actual results of the DNA tests and the fact that the sample was no longer available for independent testing did not violate his due process rights.

(concurring in part, dissenting in part).

I concur with the majority on all issues except the due process issue concerning the DNA evidence. On that issue, I would conclude that the admission in evidence of incomplete DNA results from the state’s *412PCR-STR testing violated due process where the state does not provide the defendant with (1) the genetic primer sequences used in the proprietary testing Mts, (2) a portion of the DNA sample for independent testing, or (3) notice that the sample will be exhausted in testing and an opportunity to observe the state’s testing procedure.

In considering the scope of our due process ruling in Traylor, it is important to recognize that in Traylor the DNA sample was less than six weeks old; there were no claims that it had been degraded by age, storage conditions or handling procedures; the DNA sample was of sufficient size and condition as to allow testing by both the Profiler and the Cofiler kits, providing interpretable results at all 13 loci tested by the combination of those two kits; the DNA sample was also of sufficient size to provide a portion to Traylor so his experts could conduct their own testing, both to verify the work of the BCA lab and to verify the reliability of the Profiler and Cofiler kits; and the DNA sample had not been subjected to the application of heat but had been extracted from a swab by a chemical process: See State v. Traylor, 656 N.W.2d 885, 900 (Minn. 2003). Under these facts, we concluded that there had been no violation of due process, even though Traylor was not provided with the genetic primer sequences in the two kits, stating:

We agree with the district court. Due process requires that the defense have the same amount of information as the prosecution on a scientific test so that the defense is able to adequately cross-examine the prosecution’s experts. In this case, the BCA did not have Perkin-Elmer’s validation studies or the primer sequences when it performed DNA analysis using the kits. Instead, through the use of its own testing of the kits, the BCA validated that the kits produce reliable results. Traylor likewise could have obtained the kits and performed the same type of validation testing as the BCA laboratory. Moreover, Traylor could have perused any number of publicly available validation studies that have been performed on these kits since their inception. With the DAB standards and procedures to guide him, Traylor could have also questioned the BCA technicians about the procedures and methodology followed, their validation studies, and their interpretation of the results. Traylor did not need the primer sequences or unlimited access to Perkin-Elmer’s validation studies to do so. Finally, and importantly, there was a portion of the DNA sample at issue available for Traylor to perform his own tests, an opportunity Traylor did not pursue. Accordingly, we conclude that the district court did not abuse its discretion in ruling that Traylor’s due process right to a fair trial was not violated.

Id. at 899-900.

Our due process analysis in Traylor was crucially founded on the facts that Traylor had “the same amount of information as the prosecution” about the tests performed and, “importantly, there was a portion of the DNA sample at issue available for Traylor to perform his own tests, an opportunity Traylor did not pursue.” Id. at 900. These statements do not apply in Bailey’s case. The evidence about the age and condition of the DNA sample, the lack of notice to Bailey of testing by the state that destroyed the sample, and the use of a not yet validated Bunsen burner technique to remove the cover slip from the lab slide, when added to the unavailability of the genetic primer sequences of the Profiler kit, make Bailey’s due process argument considerably more pressing than Traylor’s. The due process concerns raised by the use of proprietary kits in PCR-STR testing did not disappear after our decision in Traylor. Our decision in that case only establishes that due process is satisfied *413where a defendant is given the same information that the BCA possesses and an opportunity to perform his or her own tests on a given sample. See id.

I conclude that the admission of the DNA evidence violated Bailey’s right to due process because Bailey was not allowed to examine the genetic primer sequences applied in the Profiler kit that was used by the state; the DNA sample was so degraded that only a partial profile could be obtained from 6 of the 10 loci tested; the sample was not large enough to allow testing of additional loci by a Cofiler kit; the sample was destroyed by the state’s testing so no part was available for independent testing by Bailey’s experts; and Bailey was not given notice of the destructive testing or an opportunity to have an expert observe the testing procedures. Accordingly, I would hold that, on the record before us, the DNA evidence is inadmissible. Such a holding would not foreclose the state from attempting to overcome some of these deficiencies for purposes of a new trial.

(concurring in part, dissenting in part).

I join in the concurrence and dissent of Justice Meyer, except to the extent that it relies on the state’s failure to give Bailey notice regarding exhaustion of the sample and an opportunity to observe the testing procedure to conclude that Bailey’s right to due process was violated. On the facts presented, I do not believe that those failures resulted in a due process violation.