On April 2, 2004, a Superior Court jury found the defendant guilty of home invasion, assault with intent to commit rape, indecent assault and battery on a person over fourteen years old, and assault by means of a dangerous weapon (G. L. c. 265, §§ 18C, 24, 13H, and 15B[b], respectively), and breaking and entering with intent to commit a felony (G. L. c. 266, § 17).1 The defendant alleges numerous errors.
1. Background. On the evidence, the jury could properly find that at approximately 1:00 p.m., on April 26, 2002, the thirty-six year old victim, who was legally blind, returned to her apartment, where she lived alone. It was located in a multiple-unit, low-income, elderly housing complex in Andover. As she entered the front foyer and walked down the stairs to the basement where her apartment was located, she observed a man she did not recognize mopping the basement floor. The man in question, the defendant, was working at the apartment complex on a work-release program from the Correctional Alternative Center under the supervision of three Andover housing authority (AHA) employees. One-half hour earlier, Earl Bell, one of the AHA employees, had assigned the defendant to clean the hallways in the victim’s apartment building.
After a brief exchange of words with the defendant, the victim entered the door to her apartment, closed the door behind her, *512locked it with a “push-button” mechanism, and prepared to take a nap by removing her skirt and blouse. While she was changing, she heard a knock. Through the closed door, she spoke with a person who she thought was the man she had just seen mopping the floor (the defendant acknowledged to the police it was he who had knocked) informing her that she should be careful if she left the apartment because the floors were wet and slippery.
A minute or two later, an intruder opened the door to her apartment and entered. The intruder grabbed her from behind and told her, “Be quiet. I just want to make love to you.” Given his position, she was not able to observe the intruder’s face, but she was able to discern that he was a white male (possibly Hispanic), taller than she, with a medium-heavy build, and wearing a white or light gray sweatshirt and sneakers.
The intruder grabbed the victim around the neck and applied pressure on her neck with the inside of his elbow. He placed his hand on her mouth tightly, up against her nostrils, to the point that she complained of having difficulty breathing. She pulled at his hand in an effort to release the pressure, scratching his hand in the process.
When the intruder loosened his grasp slightly, the victim managed to get away and ran to the telephone. He quickly chased after her, knocked the telephone from her hand, and grabbed her from behind again by placing his hand over her mouth and applying pressure on her neck with the inside of his elbow. She made repeated efforts to struggle free, once calling out to a nearby neighbor, who later reported having heard only muffled sounds. The intruder shuck her in the face more than once, knocking off her thick, “Coke-bottle” like glasses onto the floor and causing her to bleed from the nose.2
Fatigued by the effort to escape, the victim gave up resisting, and the intruder forced her on the bed, stomach-side down, hands spread in front of her. Although pinned by his weight, she pulled her underpants down on demand. During the assault, the intruder used a piece of duct tape roughly seven inches long to cover the victim’s mouth. She managed to work the tape off *513from her mouth, however, and his repeated efforts to seal her mouth with another piece of duct tape proved unsuccessful.
The intruder attempted anal rape, “thrusting, and trying to get in.” His attempts continued for what “seemed like ... a long while.” The intruder then “just stopped,” put his pants back on, and left the apartment, closing the door behind him. The victim estimated that the entire assault, from the time the intruder entered the apartment to the time he left, took ten to fifteen minutes.
After several minutes of groping, without glasses, for the telephone, which had been knocked to the floor by the intruder, the victim reached a 911 operator. The call took place no more than five minutes after the intruder left the apartment. At 1:23 p.m., Officer Michael Connor arrived and observed the victim in her sports bra and underwear, still on the phone with the 911 dispatcher. He noted further that she was shaking nervously, had blood on her face, a black eye, bruises, and blood on her hands. The victim complained of a “black spot” on her eye, obscuring her already limited vision, which was eventually revealed to be the result of bleeding that occurred behind the retina of her right eye. After telling the officer that a man wearing a white sweatshirt had broken in, beaten her, and raped her, she was taken to Lawrence General Hospital to receive medical assistance.
2. The tape recording of the 911 call. The defendant objected to the Commonwealth’s putting in evidence, and playing for the jury, the recording of the victim’s 911 call to the police, on the basis that (1) the recording was “testimonial” under Crawford v. Washington, 541 U.S. 36, 51-52 (2004), and Davis v. Washington, 547 U.S. 813 (2006), and (2) the victim, because of her inability to remember what she had said to the police 911 dispatcher during that call, was “unavailable” for cross-examination. For purposes of decision, we can assume, without deciding, that the conversation on the 911 recording was testimonial. The victim was not, however, unavailable for cross-examination. She in fact testified at the trial.
The defendant bases his claim that the victim was “unavailable for cross-examination” on an assertion that she could not remember what she had said in her 911 call to the police. The *514victim did so state. “Unavailability as a witness,” the defendant argues, quoting from Rule 804(a)(1) of the Uniform Rules of Evidence,3 “includes situations in which the declarant . . . (C) testifies to a lack of memory of the subject matter of [her] statement.” The victim’s testimony, however, belies her statement that she could not remember, which seems to be true only in the limited sense that she could not then, at trial, recall everything she had said in the 911 call two years earlier. Much of what she said she did remember. Moreover, she had a remarkably detailed present memory of the events preceding her 911 call and could be effectively cross-examined on statements made to the dispatcher during the call.
A subsidiary contention is that the judge, who was faced with ruling on the Crawford motion the first day of the trial, withheld a ruling until the fifth day with the result that the jury did not hear the 911 recording until the sixth day, as the last evidence •in the trial. Thus, the defendant was put, as his counsel ruefully explained, “in [the] untenable position of putting, as the last thing the jury hears, the complaining witness back in front of the jury to recount that initial moment.” Although courts “have frequently expressed [their] preference for early rulings on motions in limine,” Commonwealth v. Vaidulas, 433 Mass. 247, 250 (2001), here the judge was faced with a question (what makes a report “testimonial”?) under the Crawford decision that was utterly novel (Crawford was decided March 8, 2004, and the trial began March 24, 2004) and remains a gray area four years later. The conscientious judge had not yet read the case and had to ponder with little assistance the extent to which the spontaneous utterance exception to the hearsay rule survived it. The delay was understandable and unavoidable, however unfortunate the effect on the defendant’s preferred order for presentation of the Commonwealth’s case.
3. The inconclusive DNA evidence. In the investigation following the victim’s removal by ambulance to the hospital, the police located a crumpled, bloodstained, white sweatshirt in a trash can in the laundry room across the hall from the victim’s apartment. The sweatshirt was linked to the defendant by two AHA employees, one of whom had given it to the defendant the *515previous week, and another who claimed to have seen the defendant wearing it that morning. Shortly after the crime, that employee saw the defendant no longer wearing the sweatshirt and in a different building (which struck the employee as odd because he knew the defendant’s work assignment that day was at the victim’s building). Through DNA testing, the State police crime laboratory was able to make positive identification of both the victim’s and the defendant’s DNA on the sweatshirt. That match, including the gathering, testing, and analysis of the results, came in without objection and with the usual astronomical improbabilities of others being the source of the DNA. Given the victim’s impairments in identifying the defendant as the intruder, the sweatshirt DNA results were doubtless the most incriminating evidence against the defendant.
The defendant objected when the Commonwealth put in evidence the police chemist’s testimony that DNA tests of stains on the defendant’s sweatpants and of a swabbing from the doorknob on the inside of the victim’s apartment door were, because of a mix of DNA from different persons, inconclusive, such that the defendant could not be identified as a source of the DNA on the doorknob but also could not be excluded as a source,4 and the victim could not be identified as a source of the DNA on the defendant’s sweatpants but also could not be excluded as a source.
On appeal, going somewhat beyond the objection at trial, the defendant argues that inconclusive DNA testing results, like positive matches, should not come in evidence without a statistical analysis enabling comparison to the statistical likelihood of a random match. See Commonwealth v. Curnin, 409 Mass. 218, 222 n.7 (1991). See also Commonwealth v. Mathews, 450 Mass. 858, 871 (2008). The defendant relies on Commonwealth v. Lanigan, 419 Mass. 15, 20 (1994) (Lanigan II) (“Evidence of a match based on currently used testing processes is meaningless without evidence indicating the significance of the match”). The chemist, Stacey Edward, who had done the DNA testing and analysis at the State police crime laboratory, explained that *516certain of the samples she had to work with showed DNA of more than one person, such that at various sites on the DNA band she found the alleles that one with the defendant’s DNA profile would leave but other alleles as well, with the result that the defendant could not be excluded as a source of the DNA, but also could not be identified as a source.5 Beyond this, Edward did not go; she had not attempted a statistical analysis.6
It was not error to let the jury learn that the mixed DNA samples on the interior doorknob and on the defendant’s sweatpants could not be excluded as having come from the defendant or the victim respectively. “That a scientific test provides descriptive information short of a definitive identification does not make it unreliable, irrelevant, or otherwise inadmissible. Very general descriptive information about a perpetrator (e.g., height, weight, hair color) is routinely admissible, even though the traits described are not unique to a defendant.” Commonwealth v. McNickles, 434 Mass. 839, 854 (2001). There, as here, the limited results possible from DNA testing of mixed samples were deemed admissible, so long as they were not presented to the jury as a match.7
This case presents a situation similar to that envisioned in Commonwealth v. Mathews, 450 Mass, at 872, in which the defense is urging that the investigation was deficient for failing *517to utilize DNA testing more extensively (failure to test exemplars from AHA employees, who had access to master keys, against the samples of crime scene bloodstains), and the Commonwealth, in response, is trying to demonstrate the extent of the testing it did employ, even if the testing was inconclusive. The Mathews decision approved the “chemist simply informing] the jury that the significance of an inconclusive result is ‘that there was just not enough DNA information in the DNA profile we obtained to either include or exclude’ [the defendant in that case] (or [the victim]) as a potential match.” Ibid. The key to admissibility is relevance. “A trial judge’s determination of the relevance of DNA test results is no different from other evidentiary decisions, and we will afford that decision substantial deference.” Id. at 872 n.15. See Commonwealth v. Nesbitt, 452 Mass. 436, 253 (2008). Similarly, here we find no fault with the judge’s determination as to the relevance of the DNA testing on the doorknob and the sweatpants.
4. Limiting cross-examination. The defendant’s strategy at trial was to show that the police had conducted a flawed investigation, settling too early on the defendant as the sole suspect upon learning that he was a convict on work release, and hence neglecting to do DNA testing and fingerprinting on the three AHA employees. Two in fact had significant criminal records: Francisco Roman, a conviction in Puerto Rico of second-degree murder in 1995; and Robert Dion, a charge of aggravated rape in 1986 in Essex County, on which he had ultimately pleaded guilty to the lesser-included offense of assault and battery. The judge allowed the defendant’s counsel to elicit from the investigating officers that they had not asked the three AHA employees whether they had prior criminal records, but she did not permit counsel to ask the officers whether the knowledge that two of the AHA employees had prior criminal records would have been important to the investigation.
Bearing in mind that the defendant’s purpose was to show a flawed investigation,8 the judge allowed the defendant substantial latitude in constructing that defense. Specifically, she allowed it to be brought out (1) that two of the employees had *518prior records; (2) that the police had failed to ask them about their records and failed to check their criminal histories; (3) that the police did not make fingerprint comparisons to the prints of Dion or Roman; (4) that two of the employees had access to the master key for the victim’s building; and (5) by stipulation, that Roman had been convicted in 1995 of second-degree murder.
“While a criminal defendant is entitled to a reasonable cross-examination of the Commonwealth’s witnesses, the scope of the cross-examination rests largely within the discretion of the trial judge, and . . . will not [be] disturbed] . . . unless the defendant demonstrates that [he] was prejudiced . . . .” Commonwealth v. Farley, 443 Mass. 740, 748 n.12 (2005), and cases cited. The judge was not obliged to let the defendant divert the trial into collateral matters; the important aspect of the witnesses’ criminal records was that the police had failed to investigate so as to learn whether those employees had records and should be investigated more extensively, as the defendant was. The judge exercised her discretion reasonably and properly.
5. Roll of duct tape. The defendant argues that his trial counsel provided ineffective assistance by failing to move for dismissal of the charges, or other remediation, based on the investigating police officers’ failure to take control of a roll of duct tape that was in the victim’s apartment when the officers arrived. The victim did not have duct tape in her apartment; inferentially it was brought in by the intruder, who used two strips from it to try to seal the victim’s mouth. Those two strips remained, but fingerprint traces were too smeared for print comparisons. A witness testified that duct tape can be a good medium for fingerprints; but the remaining roll (it showed in the police photographs) had not been tested and was lost by the time of trial.
The situation was not one calling for sanctions. The judicial remedy for destruction or loss of potentially exculpatory evidence turns on three factors: the culpability of the Commonwealth, the materiality of the evidence, and the potential prejudice to the defendant. Commonwealth v. Willie, 400 Mass. 427, 432 (1987). As to culpability, this case is unlike Commonwealth v. Sasville, 35 Mass. App. Ct. 15, 18, 22-24 (1993) (where the Commonwealth intentionally authorized the destruction of an aborted fetus, eliminating the defendant’s chance to prove con*519clusively through blood tests that he was not the father), and also unlike Commonwealth v. Willie, supra at 429-430 (where the Commonwealth either used up or failed to preserve secretion samples from a sheet which the defendant sought access to for his own testing). In this case, so far as the evidence shows, the police investigators simply neglected to take possession of the roll of duct tape, not knowing at that time that it had almost certainly been brought into the apartment by the intruder. The defendant’s speculation that bad faith may have been involved is the product not of reasonable possibility based on “concrete evidence” but rather on “fertile imagination.” Commonwealth v. Neal, 392 Mass. 1, 12 (1984). As to materiality, while the roll of tape might have shown fingerprints of one other than the defendant, such a showing would not have definitively excluded the defendant as the intruder, because it came, most likely, from a supply of such tape in a housing authority storeroom and might have been touched by others. As for prejudice to the defendant, the testimony that duct tape was a good medium for fingerprints referred to the sticky side of the tape, which, the testimony went, preserves a good print unless it is restuck, destroying the image. The slippery side of the tape is prone to smearing. The two strips with which the intruder tried to seal the victim’s mouth showed the practical limitations of duct tape as a fingerprint-preserving medium.
Finally, the defendant’s contention that there was insufficient evidence to permit a finding by the jury that the roll of duct tape, used by the defendant to facilitate the rape by taping the victim’s mouth, constituted a “dangerous weapon” under G. L. c. 265, §§ 15B and 18C, is untenable in light of Commonwealth v. Cruz, 430 Mass. 182, 194-195 (1999).
Judgments affirmed.