Lenk, J.
The defendants, Michael Cowels and Michael Mims, were convicted by a Superior Court jury in December, 1994, of murder in the first degree in the stabbing death of Belinda Miscioscia. Among the evidence presented against them at trial were two “bloody” towels. The Commonwealth offered testi-
many suggesting that the defendants had used the towels to clean themselves after stabbing the victim. Testing performed on the towels at the time of the trial indicated the presence of human blood. The testing, however, was inconclusive. Further testing was performed on just one of the towels; the sample on the other was too small to be tested. An expert testified that the further testing neither identified nor excluded the defendants or the victim as the sources of the blood. In June, 1997, this court affirmed the defendants’ convictions. See
Commonwealth
v.
Cowels,
425 Mass. 279, 285-293 (1997).
In 2008, the defendants filed separate motions for a new trial, based in part on deoxyribonucleic acid (DNA) testing performed on the previously tested towel by an independent laboratory after their convictions. That testing revealed that the blood contained on the towel did not belong to either of the defendants or the victim, but instead to an unidentified male. The defendants also argued in their motions that they had been deprived of the effective assistance of counsel.
After a nonevidentiary hearing, the motion judge, who was also the trial judge, denied the motions. In 2011, each defendant filed a “gatekeeper” petition before a single justice of the county court, pursuant to G. L. c. 278, § 33E, seeking leave to appeal from the denial of his motion for a new trial. In January, 2014, the single justice allowed both petitions, and the matters were entered in this court the following month.
We conclude that, given the towels’ role as one of the few pieces of physical evidence that corroborated the testimony of a key prosecution witness whose credibility was sharply challenged, the towels likely were a real factor in the jury’s deliberations. Consequently, we believe that there is a substantial risk that, had the newly available DNA testing been available at the time of the trial and resulted in the inadmissibility of the towels in the Commonwealth’s case, the outcome of the trial would have been different. The defendants, therefore, must receive a new trial.
1.
Background,
a.
Evidence at trial.
Miscioscia’s body was found in a yard behind an industrial building in Chelsea on the morning of Monday, June 28, 1993. An autopsy revealed that she had been stabbed six times. The fatal stab wound perforated her heart. She also had been slashed several times, and suffered numerous bruises and defensive wounds. Her body and clothing
were “blood soaked.” A pair of eyeglasses was found a few feet from the body, and a bag of marijuana was stuffed inside her bra.
The police investigation quickly turned towards the defendants, as they were among the last people to have seen the victim alive on the preceding Saturday night. The trial presented the jury with two conflicting timelines of the defendants’ activities on that night.
The Commonwealth presented — largely through the testimony of Robert Salie, a friend of Cowels and Mims — the following timeline. On the night of Saturday, June 26, 1993, the victim was in her brother’s apartment in Chelsea with her brother, her brother’s girl friend, and Peter Rowe, whom the victim was dating. The victim and Rowe had plans to go to the Wonderland Ballroom that evening. According to the witnesses present in the apartment, the victim left by herself shortly before 9 p.m. to purchase marijuana for the group from Cowels. She had come to know Cowels while attending a course that she was required to take in conjunction with a conviction of operating a motor vehicle while under the influence of alcohol.
The victim met Cowels a few blocks away from the apartment, where Cowels was attending a party along with his friend, Mims. After purchasing the marijuana, the victim did not return to the apartment. Instead, Salie testified, Cowels, Mims, and the victim arrived at Salie’s home, located approximately one and one-half miles from the location of the party, at approximately 9:30 p.m. There the four smoked a marijuana cigarette, and Salie witnessed the victim and the defendants prepare to have sex in his bedroom. Salie was invited to participate, but declined. The defendants and the victim then emerged from Salie’s bedroom and left his apartment shortly afterwards.
Salie testified that the defendants returned to his apartment, unaccompanied by the victim, at approximately 11:30 p.m. After he allowed them to enter, they immediately went to his bathroom. They remained in the bathroom for approximately twenty minutes, during which time he could hear the water running in the sink. When the defendants emerged from the bathroom, Cowels was in his underwear and carrying a plastic bag containing the clothes he had been wearing that evening. Heading towards Salie’s bedroom, Cowels asked Salie if he could borrow some clothes. Cowels also held out a sneaker that he had been wearing, on which Salie saw a spot that he believed to be blood. “[I]f you fuck with me,” Cowels told Salie, “this is what happens.” Cowels
remarked that Salie would read in the newspapers the following day about “[h]ow we killed her,” adding, “she was a fucking pig and she got what she deserved.” He told Salie that “if [he] said anything [he]’d get hurt.”
Salie gave clothes to both defendants. After dressing, Mims told Salie to “keep [his] mouth shut,” and Cowels put a finger to Salie’s head and reiterated that if Salie said anything, he would “get fucking hurt.” The defendants left carrying bags of their clothing.
The defendants returned at approximately 1 a.m., without their bags, and again told Salie that he had “better keep [his] mouth shut.” Cowels departed, while Mims spent the night on a couch in Salie’s apartment. On his way out, Cowels gave Salie a ride to the store to purchase a pack of cigarettes.
Although Salie’s testimony was cmcial to the Commonwealth’s case, his credibility was extensively impeached during cross-examination. In his first two interviews with police, Salie had not offered the account that he ultimately offered at trial. Instead, he indicated that the defendants had come to his apartment only once on the evening of the victim’s death, when Mims arrived to sleep over, and that the victim was never present in his apartment. Salie offered the narrative to which he later testified at trial after he learned that he could be charged as an accessory after the fact for providing the defendants with clothing. Salie also entered into a cooperation agreement with the Commonwealth under which he avoided a mandatory term of incarceration for several unrelated motor vehicle offenses, for which Cowels had been called to serve as a prosecution witness against Salie. Salie had a long criminal history, including numerous drug-related offenses, and admitted to being a recovering heroin addict. Finally, the defense challenged the inconsistency between Salie’s testimony that he was afraid of the defendants and his decision to go for a ride with Cowels to pick up cigarettes immediately after Cowels supposedly threatened him, and to watch a softball game in which the defendants were playing on the following day.
The Commonwealth offered evidence that, it contended, corroborated Salie’s testimony. Approximately one month after the stabbing, police recovered two towels, one from inside and the other from behind a hamper in Salie’s bathroom. The towels were visibly stained; one was a large bath towel containing a large, rust-colored stain, covering the majority of the towel’s surface. Testing confirmed the presence of human blood on both. Salie
testified that he last saw the towels on June 26, 1993, when they were on the rack in his bathroom and were clean. According to the Commonwealth’s serologist, the bloodstains on one towel — a small hand towel •— were too small to be tested without exhausting the sample. Further testing performed on the other towel revealed a “weak reaction” to type A blood. Both of the defendants and the victim have type O blood. The Commonwealth’s serologist, however, testified that the reaction was too weak to draw any conclusions from it, stating that the blood on the towel could belong to anybody.
The Commonwealth also introduced evidence that both of the defendants got rid of their old shoes and acquired new ones at approximately the same time shortly after the murder. An officer who interviewed Cowels soon after the stabbing testified that he appeared at the police station wearing “a brand new pair of white high-top sneakers.” Richard Polovick, a friend of Cowels, testified that, sometime after the victim’s death, Cowels stopped and asked Polovick to repair a tire on his Chevy Nova automobile. When Cowels opened the trunk of the vehicle, Polovick saw a pair of sneakers. Polovick asked Cowels whether the shoes were the “notorious sneakers that people were talking about.” Cowels responded, “Yes, that’s them,” and Polovick watched as Cowels threw the sneakers into some bushes. Polovick later directed police to the location, where the sneakers were recovered several weeks after they had been left. Testing revealed trace amounts of nonvisible, “occult” blood in a recessed portion of the sole of one of the sneakers. The forensic examiner, however, testified that the sample was too small to determine the blood type, or even whether it came from a human or an animal.
Similarly, the Commonwealth offered testimony that Mims appeared to play softball with friends on Sunday, June 27, wearing “green high-tech[] sneakers,” which Salie testified he had lent to Mims the previous night. Asked by a friend what had happened to his old sneakers, which were expensive Reebok high-tops that were in good condition the last time the friend saw them, Mims responded that he had thrown them out because they had gotten “wet and squeaky” in the rain the previous night. However, it had not rained that night. Unlike Cowels’s shoes, Mims’s shoes were never recovered.
Finally, the prosecution offered evidence of a vaginal smear swab taken from the victim. Testing of the swab revealed the presence of semen. The Commonwealth’s expert, however, tes-
tilled that she could not identify, based on her analysis of the swab, whether the seminal fluid came from either of the defendants or both, or from someone else altogether.
In addition to the evidence that was introduced to corroborate Salie’s account of the evening, the Commonwealth offered evidence that the defendants sought to construct a false alibi in the days following the murder. Larry Bavis testified that the defendants and a third person named Victor Grimaldi urged Bavis to lie to “cover” for them by saying that he was playing pool with them at Triple O’s, a bar in the South Boston section of Boston, on the night of the stabbing. In response to that request, Bavis became angry, and called the police to report the conversation.
Additionally, a police officer who interviewed Cowels shortly after the victim’s death testified to several incriminating statements Cowels made. Cowels indicated that he had met the victim at “drunk driving school,” and that he had had sex with her on several previous occasions, although he had not done so for “several months.” According to the officer’s testimony, Cowels described the victim as a “dirty pig,” and said that she was “such a dirty pig that one time he and Mims had had sex with her at the same time.” When the officer informed Cowels that the victim was dead, Cowels “began sobbing,” and stated, “I’m only twenty-three. I don’t want to go to jail.”
The defendants offered an alternative timeline of their activities that evening, which was introduced at trial primarily through the same police officer’s account of what the defendants had said to him during an interview. According to this timeline, the victim arrived at the party in Chelsea to purchase marijuana at approximately 7:30 p.m., rather than 9 p.m. as in the Commonwealth’s version. The defendants, the victim, and Grimaldi then traveled to Triple O’s, where they played pool and drank until approximately 11 p.m. At 11 p.m. they left; Grimaldi, who lived upstairs from Triple O’s, went home; meanwhile, the victim accompanied the defendants to the residence in Revere that Cowels shared with his sister, Barbara Cowels. About, twenty minutes later, Barbara Cowels arrived at her home, where she encountered the defendants and the victim. She chastised her brother for drinking, and kicked him out of the house. The defendants then dropped the victim off near the party where they had picked her up at the beginning of the evening, and continued on to a bar in Revere, where they remained until approximately 2 a.m. Once the bar closed, the defendants went to Salie’s house, purportedly for the first time
that evening. Mims spent the night there, while Cowels went to his girl friend’s house.
The defense introduced testimony by four witnesses. Two of those witnesses corroborated this alternative timeline. Barbara Cowels testified that when she arrived at her home at approximately 11:20 p.m. that Saturday night, she encountered the defendants along with the victim. She stated that, after a brief argument with Cowels concerning his drinking, the defendants and the victim left the house. Similarly, John Heald, a friend and neighbor of the Cowelses, testified that on the evening of the stabbing, sometime between 10 and 11:30 p.m., he observed the defendants leaving the apartment with a young woman. Because the testimony of both of these witnesses in conjunction suggested that the victim was still alive and in the company of the defendants at approximately 11:30 p.m., it undermined Salie’s statement that the defendants arrived at his apartment to clean up after the stabbing at that time.
The defense also introduced an alternative account of the incident when Cowels asked Polovick to repair a tire. A coworker at the print shop where Cowels was employed at the time of the stabbing testified that he accompanied Cowels on a trip to purchase new sneakers. On the way, he and Cowels stopped to have Polovick repair a tire. According to the co worker’s testimony, after Polovick repaired the tire, he and Cowels continued on to a shoe store. There Cowels purchased new sneakers and put the old sneakers on the back seat of his vehicle. The coworker also testified that workers at the print shop got ink and chemicals all over any clothes that they wore to work, and that Cowels wore his sneakers to work.
Finally, the defense called a chemist at the State police crime laboratory. The chemist, who was not assigned to the case,
testified that she believed that the laboratory could have performed a comparison test, using blood from the victim and the defendants, that could have determined whether any of them was the source of the blood found on the towel. No comparison test, however, was ever performed. Both defendants were found guilty of murder in the first degree on a theory of extreme atrocity or cruelty.
b.
Posttrial proceedings.
After this court affirmed the defendants’ convictions,
Commonwealth
v. Cowels, 425 Mass. 279, 285-293 (1997), Mims filed a petition for a writ of habeas corpus in the United States District Court for the District of Massachusetts. Mims ví. DiPaolo, 98-CV-11203-MEL (D. Mass. Apr. 1, 1999). The petition was denied, and, in an unpublished opinion, the United States Court of Appeals for the First Circuit affirmed the denial.
Mims
v.
DiPaolo,
201 F.3d 428 (1st Cir. 2000).
Cowels, initially proceeding pro se, moved for a new trial on July 3, 1998. Cowels later filed an assented-to motion to stay proceedings on his motion for a new trial while his subsequently obtained counsel sought permission to perform DNA testing on certain items of evidence. A Superior Court judge, who also had been the trial judge, authorized Orchid Cellmark (Cellmark) to perform the requested testing. DNA testing on the towel that had been subjected to further testing in the original trial revealed that the blood found on it did not come from either of the defendants or the victim, but instead from an unidentified male. DNA testing of the vaginal swab taken from the victim identified Mims as a contributor of sperm recovered from the victim, and excluded Cowels as a contributor. Cellmark also replicated the presumptive test for blood which the Commonwealth had performed on the sneakers before the first trial. Cellmark’s test established that there was no blood on the parts of the sneakers that were tested.
Based on the results of these tests, in February, 2008, Cowels filed an amended motion for a new trial. Shortly thereafter, Mims also sought a new trial, pressing the same arguments as Cowels. The defendants argued that they were entitled to a new trial based on “newly discovered” evidence, in the form of the DNA test results on the towel and the vaginal swab. They further contended that they were deprived of effective assistance of counsel because trial counsel (1) failed to conduct independent forensic testing on the sneakers; (2) failed to call certain witnesses whose testimony would have corroborated the defendants’ account of their activities that evening; and (3) failed adequately to develop the defense
that Peter Rowe, who had been dating the victim and had been “stood up” by the victim on the night of the stabbing, was the actual killer.
The judge
rejected both arguments and denied the defendants’ motions for a new trial. The judge concluded that the towels would not have been admissible in light of the DNA testing establishing that neither the defendants nor the victim were the source of the blood found on one of the towels. The judge determined, however, that “there is no substantial risk that the jury would have reached a different conclusion if the ‘bloody’ towels were not in evidence.” The judge also rejected the ineffective assistance of counsel arguments.
2.
Discussion.
The defendants argue, as they did in their original motions, that they are entitled to a new trial based on (1) newly discovered evidence, in the form of the DNA testing on one of the towels and the vaginal swab; and (2) ineffective assistance of counsel. In reviewing a judge’s decision on a motion for a new trial, we “examine the motion judge’s conclusion only to determine whether there has been a significant error of law or other abuse of discretion.”
Commonwealth
v.
Grace,
397 Mass. 303, 307 (1986). While we “extend[] special deference to the action of a motion judge who[, as here,] was also the trial judge,”
id.,
we nonetheless conclude that, in the circumstances here, the defendants are entitled to a new trial based on the DNA testing performed on the towel.
Accordingly, we need not reach the
defendants’ claims based on the vaginal swab testing or their ineffective assistance of counsel claims.
Rule 30 (b) of the Massachusetts Rules of Criminal Procedure, as appearing in 435 Mass. 1501 (2001), allows a trial judge to “grant a new trial at any time if it appears that justice may not have been done.” To prevail on a motion for a new trial on the basis of newly discovered evidence, a defendant must meet both prongs of a two-part test. First, a defendant “must establish that the evidence was unknown to the defendant or trial counsel and not reasonably discoverable at the time of trial.”
Commonwealth
v.
Shuman,
445 Mass. 268, 271 (2005). Second, a defendant must show that the evidence “casts real doubt on the justice of the conviction.”
Commonwealth
v.
Grace,
397 Mass. at 305. The Commonwealth contends that the defendants have failed to satisfy either of the elements necessary to prevail on a motion for a new trial.
a. In its brief, the Commonwealth maintains that the DNA evidence identified on the towels recovered from Salie’s bathroom does not constitute “newly discovered” evidence. Even if the argument were not waived, we would find it unpersuasive, given that this court did not determine the admissibility of DNA testing of the type performed here until 1997,
Commonwealth
v.
Vao Sok,
425 Mass. 787, 789 (1997), and that the very article that the Commonwealth cites as establishing that DNA testing was available in 1993 indicates that it was then “still at the experimental stage.” Thompson, Evaluating the Admissibility of New Genetic Identification Tests: Lessons from the “DNA War,” 84 J.
Crim. L. & Criminology 22, 30 n.36 (1993). We need not reach the issue of whether the evidence is “newly discovered,” however, since it plainly was waived. Generally, “[a]n issue not raised or argued below may not be argued for the first time on appeal.”
Carey
v.
New England Organ Bank,
446 Mass. 270, 285 (2006), quoting
Century Fire & Marine Ins. Corp.
v.
Bank of New England—Bristol County, N.A.,
405 Mass. 420, 421 n.2 (1989). See
Commonwealth
v.
LaBriola,
430 Mass 569, 570 n.l (2000). As the judge observed, in the proceedings below “[t]he Commonwealth d[id] not dispute that the DNA results are ‘newly discovered.’ ”
b. To satisfy the second element of the test for a motion for a new trial, a defendant must establish not only that the new evidence is
“material and credible[,] but also [that it] . . . carries] a measure of strength in support of the defendant’s position. . . . Thus newly discovered evidence that is cumulative of evidence admitted at the trial tends to carry less weight than new evidence that is different in kind.... Moreover, the judge must find there is a substantial risk that the jury would have reached a different conclusion had the evidence been admitted at trial. . . . The strength of the case against a criminal defendant, therefore, may weaken the effect of evidence which is admittedly newly discovered. . . . The motion judge decides not whether the verdict would have been different, but rather whether the new evidence would probably have been a real factor in the jury’s deliberations.” (Citations omitted.)
Commonwealth
v.
Grace,
397 Mass. at 306.
This case differs from many cases involving motions for a new trial on the basis of newly discovered evidence in one respect. In such cases, a defendant generally offers newly discovered evidence that was not presented in the original trial, but that, the defendant argues, probably would have been a real factor in the jury’s deliberations if it had been offered. See, e.g.,
Commonwealth
v.
Wright,
469 Mass. 447, 447-448 (2014) (defendant argued “that newly discovered evidence in the form of third-party culprit evidence warranted a new trial”);
Commonwealth
v.
Raymond,
450 Mass. 729, 729-730 (2008) (“The defendant alleges that the prosecution withheld the fact of an agreement it purportedly made with a key witness . . . ”). The defendants’ motions here, however, are based less on newly discovered evidence that
could have been admitted in evidence at the trial, than on newly available analysis that would remove from the jury’s consideration evidence admitted at trial in the Commonwealth’s case. The judge observed that “the ‘bloody’ towels would not have been admitted in evidence had DNA testing shown that the blood on the towels was not that of Cowels, Mims, or Miscioscia.” Although the defendants suggest that they might offer the DNA testing performed on one of the towels in evidence in a new trial, ostensibly to support an argument that Salie intentionally “frame[dj” the defendants, the primary value of the DNA testing for the defendants plainly derives from the way in which it eliminates the towels as evidence against the defendants.
This distinction, however, does not raise significant conceptual problems for our analysis. In the typical case, where a defendant argues on the basis of newly discovered exculpatory evidence that was not presented at the original trial, we ask “whether the new evidence
would probably have been
a real factor in the jury’s deliberations” had it been presented (emphasis supplied).
Commonwealth
v.
Grace,
397 Mass. at 306. In this case, where the defendants argue on the basis of a newly available analysis that likely would have rendered inculpatory evidence presented at the original trial inadmissible, we ask whether that inculpatory evidence “likely
was
a ‘real factor’ in the jury’s deliberations such that its elimination would cast real doubt on the justice of the defendant’s conviction” (emphasis supplied).
Commonwealth
v.
Sullivan,
469 Mass. 340, 350 (2014). See
Commonwealth
v.
DiBenedetto,
458 Mass. 657, 667 (2011), quoting
Commonwealth
v.
Grace, supra
(in case where newly available analysis undermines test results presented in initial trial, analyzing whether “evidence concerning the . .. test results probably was not ‘a real factor in the jury’s deliberations,’ and not likely to create ‘a substantial risk that the jury would have reached a different conclusion’ if it had not been admitted at trial”). If we conclude that the subsequently eliminated inculpatory evidence likely did play an important role in the jury’s deliberations, then we must conclude that there is a “ ‘a substantial risk that the jury would have reached a different conclusion’ if it had not been admitted at trial.”
After a detailed review of the trial record, we determine that the towels likely were a “real factor” in the jury’s deliberations, and that there is consequently a substantial risk that the outcome of the trial would have been different had the towels been excluded
altogether or neutralized through expert testimony indicating that blood found on one of the towels matched neither the defendants nor the victim. We reach this conclusion based on the paucity of physical evidence in the case, the vital importance of Salie’s testimony, and the substantial challenges to his credibility.
The case against the defendants was entirely circumstantial. There were no eyewitnesses to the crime. The Commonwealth never found the weapon. There was no forensic evidence at the crime scene tying either defendant to the crime. Although the victim’s body showed defensive wounds, a police investigator who interviewed Cowels and Mims within days of the stabbing observed no cuts or scratches on either defendant. The only physical evidence that the Commonwealth offered linking the defendants to the crime was the towels, Cowels’s sneaker, and the vaginal swab.
Due to the dearth of physical evidence, the case against the defendants hinged, to a significant extent, on the testimony of Salie. The problems with Salie’s credibility, moreover, were numerous and significant. In the prosecutor’s own words, Salie was “a junkie” with a “checkered background” and a “long criminal record.” Indeed, he was impeached at trial with at least nineteen prior convictions of crimes ranging from drug possession to arson. Salie originally had given an account that was consistent with Cowels’s and Mims’s claim that they only visited his home once on that evening, before changing his account to the narrative presented at trial. Salie, moreover, had a motive to change his story, and to point his finger at Cowels specifically: not only did Salie’s testimony enable him to avoid prosecution as an accessory after the fact, but it also allowed him to avoid jail time on an unrelated motor vehicle offense, for which Cowels was to serve as a witness for the prosecution. It is, in short, difficult to imagine a witness with more credibility problems than Salie.
To counter these issues with Salie’s credibility, the prosecutor repeatedly emphasized the ways in which Salie’s testimony had been “confirmed” and “verified.” He observed that Salie’s testimony had been “validated by so many different people and so many different sources.” The prosecutor referred to the towels in the context of this discussion, as evidence that “substantiated” Salie’s account. Noting Salie’s testimony that the defendants ran into his bathroom to clean and change, the prosecutor stated: “And when the police do a search, what do they find? They find
bloody towels. Those towels are in evidence, ladies and gentlemen. Another piece confirming Mr. Salie.”
This court has observed that evidence likely functions as a real factor in a jury’s deliberations where the evidence “is more credible than any other evidence on the same factual issue and bears directly on a crucial issue before the jury, such as the credibility of an important prosecution witness.”
Commonwealth
v.
Tucceri,
412 Mass. 401, 414 (1992). Here, the towels were the only physical evidence corroborating a key element of an important prosecution witness’s testimony, and functioned to reinforce Salie’s severely challenged credibility. Accordingly, we conclude that the towels likely were a real factor in the jury’s deliberations.
The Superior Court judge offered several reasons for reaching the contrary conclusion. First, the judge observed that the Commonwealth’s expert testified that the testing performed on the towels was “inconclusive,” and that defense counsel’s closing arguments “highlighted] the limited evidentiary value of the ‘bloody’ towels.” The Commonwealth echoes that argument on appeal. It contends that, because the testing performed on the towels was inconclusive, “[r]emoving the towel [tested by Cellmark] from the calculus of the trial, or adding to it the fact that neither defendants] nor the victim’s blood was found on the towel,. . . does not ‘carry a measure of strength in support of the defendant’s position.’ ”
There is, however, a significant difference between a test that fails to say whether the blood came from the defendants or the victim and a test that definitively establishes that the blood did
not
come from either the defendants or the victim. The inconclusive test results allowed the prosecutor to cite the “bloody towels” as physical evidence that corroborated Salie’s testimony. As the defendants argue, “the use of the towel at trial was a ... powerful visual for the jury.” A test that definitively excluded the defendants and the victim, by contrast, would not merely have reduced the weight that the jury might have given the towels in substantiating Salie’s testimony. Rather, it would have meant that the towels would not have been admissible at all, at least when offered by the prosecution.
Second, the judge concluded that, although the towels functioned to corroborate Salie’s testimony, they likely were not a real factor in the jury’s deliberations because “Salle’s credibility was brought into question numerous times throughout his testimony,” and there were “numerous reasons put forth at trial for finding
Salie’s testimony unreliable and incredible.” But the repeated challenges to Salie’s credibility increase, rather than decrease, the importance that the towels likely had in the jury’s deliberations. It is difficult to see how the jury could have voted to convict if the jurors had not believed Salie’s testimony. Much of the other evidence against the defendants was valuable primarily because it confirmed details in Salie’s account: the original testing of the vaginal swab smear confirmed Salie’s statement that he witnessed the defendants and the victim preparing to have sex; Polovick’s testimony regarding Cowels’s disposal of his sneaker and the original testing performed on the shoe were valuable because they confirmed Salie’s testimony that Cowels brandished a bloody sneaker at him after the killing. Furthermore, Salie’s testimony that the defendants came into his apartment, cleaned up, and discarded their clothing on the evening of the killing explained why the prosecution could not present any other physical evidence linking the defendants with an especially bloody stabbing. Because the jury’s verdict indicates that the jury very likely did credit Salie’s testimony, despite the challenges it faced, and because the towels were among the very few pieces of physical evidence that buttressed Salie’s credibility, there is a real question whether the jury would have credited his testimony had the towels not been presented.
Third, the judge cited case law holding that “evidence of a type merely tending to impeach or to corroborate credibility of a witness ordinarily will not be the basis for ordering a new trial.”
Commonwealth
v. Shuman, 17 Mass. App. Ct. 441, 448 (1984), overruled in part on other grounds by
Commonwealth
v.
Jones,
59 Mass. App. Ct. 157 (2003). However, we have never adopted an inflexible rule that newly discovered evidence that merely corroborates or impeaches a witness’s testimony is an insufficient basis for a motion for a new trial. In fact, we have found that, in rare cases, a new trial may be warranted “[wjhere the Commonwealth’s case depends so heavily on the testimony of a witness” and where the newly discovered evidence “seriously undermines the credibility of that witness.”
Commonwealth
v.
Liebman,
388 Mass. 483, 489 (1983). Furthermore, the DNA testing on the towel does not “merely impeach” Salie’s testimony.
Commonwealth
v.
Sullivan,
469 Mass. 340, 352 (2014). Rather, it “negates a key piece of physical evidence that the prosecution relied on in arguing that the jury should credit [Salle’s] testimony.”
Id.
In
Commonwealth
v.
Sullivan, supra
at 353, we affirmed a Superior
Court judge’s decision granting a defendant’s motion for a new trial based on newly discovered evidence. There, at the original murder trial, the Commonwealth had offered expert testimony indicating that a jacket, allegedly worn by the defendant during the commission of the crime, had blood on its cuffs that was “ ‘consistent’ with that of the victim.”
Id.
at 345. There, as here, the prosecutor argued in closing that the jacket corroborated the testimony of a witness whose credibility was significantly challenged.
Id.
at 349. There, as here, subsequent DNA testing excluded the victim as the source for the genetic material identified on the jacket.
Id.
at 349-350. We concluded that “the purported blood on the defendant’s cuffs and the hair in [the] defendant’s pocket were not merely cumulative of other physical evidence presented at trial.”
Id.
at 352. Rather, “[t]hey were different in kind because they served as the sole pieces of physical evidence indicating the defendant had been in the presence of the victim during the killing.”
Id.
Here, similarly, Salie testified that the defendants spent twenty minutes in his bathroom cleaning up after allegedly participating in a bloody stabbing. Police investigators acquired a warrant to search Salie’s apartment for “any bloody clothing, hairs, any trace of blood that may be found in the apartment.” They searched the apartment for over one hour, during which time they confiscated hair from the couch and also performed tests for blood on the couch. Despite this search, police recovered no evidence indicating that the victim had ever been present in Salie’s apartment, and the towels were the only evidence seized from Salie’s apartment that corroborated his testimony. We think that the absence of any physical evidence supporting Salie’s testimony likely would have carried real weight in the jury’s deliberations.
Finally, the judge observed that “there was other evidence implicating Cowels and Mims.” The Commonwealth takes this argument a step further. It contends there is no substantial risk that the outcome of the trial would be different in the absence of the towels because “the Commonwealth’s case against the defendants was . . . strong.”
To address this argument, we must clarify the proper approach to assessing a motion for a new trial on the basis of newly discovered evidence. We have observed that “[t]he strength of the case against a criminal defendant. . . may weaken the effect of evidence which is admittedly newly discovered.”
Commonwealth
v.
Grace,
397 Mass. 303, 306 (1986). See
Commonwealth
v.
Moore,
408 Mass. 117, 127 (1990). In considering the over-all strength or weakness of the prosecution’s case, however, a reviewing court must ensure that its focus remains on whether, in light of “a full and reasonable assessment of the trial record,” the evidence at issue “would have played an important role in the jury’s deliberations and conclusions.”
Commonwealth
v.
Tucceri,
412 Mass. at 414. The over-all strength or weakness of the evidence presented against a defendant is significant, therefore, because it provides the context within which to assess whether the newly discovered evidence would have been a real factor in the jury’s deliberations. Where a case is “a weak one for conviction,” for instance, a jury is more likely “to pay attention to collateral factors and even to make them decisive,” and thus a court is more likely to conclude that evidence relating to one of these factors was or would have been a real factor in the jury’s deliberations.
Commonwealth
v.
Bennett,
43 Mass. App. Ct. 154, 162 (1997).
The analysis remains focused, however, on “what effect the omission might have had on the jury.”
Commonwealth
v.
Tucceri,
412 Mass. at 411. “[Ojur inquiry is not whether the verdict may have been different, but whether the evidence in question probably served as a real factor in the jury’s deliberations.”
Commonwealth
v.
Sullivan,
469 Mass. at 353. Where we determine that newly discovered evidence likely would have functioned as a real factor in the jury’s deliberations, or (as in this case) that subsequently discredited evidence likely did function as a real factor, we may not then assess whether the jury still would have reached the same conclusion. Instead, the determination that the evidence likely was a real factor in the jury’s deliberations demands a new trial. We have justified this approach as “preserving], as well as it can in the circumstances, the defendant’s right to the judgment of his peers,” since it ensures that the court’s analysis turns on “what effect the omission might have had on the jury,” rather than on “what... impact the late disclosed evidence has on the judge’s personal assessment of the trial record.”
Commonwealth
v.
Tucceri, supra.
Here, although the Commonwealth asserts that its over-all case against the defendants was strong, it does not contest that Salie’s testimony was the linchpin. Without Salie’s testimony, the case against the defendants would not have been strong. In light of the unique facts presented here — given Salie’s importance to the
prosecution’s case and the towels’ status as one of only a few items of physical evidence that bolstered his severely beleaguered credibility — we determine that the towels likely were a real factor in the jury’s deliberations. Consequently, there is a substantial risk that the newly available testing excluding the victim and the defendants as possible sources of the blood on one of the towels would have altered the outcome.
3.
Conclusion.
The judgments of conviction are vacated and set aside, and the matters are remanded to the Superior Court for a new trial.
So ordered.