*785 **106 On the night of January 31, 2005, members of the Boston fire department in the Dorchester section of Boston responded to the sound of banging on their fire house door. Outside the station was the defendant, Manolo Salazar, covered in blood. After examining the defendant, firefighters found only a minor cut on his right hand. Police investigation revealed the source of the rest of the blood on the defendant's body: Carlos Cruz, the defendant's roommate, whom police found dead in their shared apartment.

After a jury trial, the defendant was convicted of murder in the first degree on the theory of deliberate premeditation. 1 He raises several arguments on appeal: (1) the judge erred in denying his motions for a required finding of not guilty on the murder charge because the evidence was insufficient to establish deliberate premeditation; (2) he should be afforded a new trial because trial counsel was ineffective in failing to introduce evidence in support of a defense based on voluntary intoxication; and (3) improper statements in the prosecutor's closing argument created a substantial likelihood of a miscarriage of justice. 2 The defendant also asks us to exercise our authority under G. L. c. 278, § 33E, to either reduce **107 his conviction to murder in the second degree or order a new trial.

Having thoroughly reviewed the defendant's asserted errors and the record as a whole, we discern no reversible error. However, given the unique circumstances of this case, we exercise our authority under G. L. c. 278, § 33E, to reduce the defendant's verdict to murder in the second degree.

1. Background . a. Trial . Because the defendant challenges the sufficiency of the evidence, we recite the facts that the jury could have found in the light most favorable *786 to the Commonwealth and reserve additional facts for later discussion.

At approximately 8:40 P.M. on January 31, 2005, members of the Boston fire department heard banging on the door of a station in Dorchester. They found the defendant at the door. He fell down. His clothes, including the socks on his shoeless feet, were covered in blood. The lone injury firefighters discovered was a minor laceration between the defendant's right thumb and forefinger that was not bleeding heavily enough to account for all of the blood on his clothes. After an ambulance arrived, the defendant became combative and was handcuffed to a backboard by paramedics before being transported to a Boston hospital.

Initially, the defendant told police that his last name began with a "Z" instead of an "S" and provided them with an inaccurate home address. Subsequent investigation led police to the defendant's home in Dorchester, a three-unit apartment building. On entry into the defendant's third-floor apartment, police discovered the victim lying dead on the floor between a hallway and a bedroom. A Boston Police Department detective observed a large pool of blood below the victim's body. Inside that pool of blood was a kitchen knife with a wooden handle and a serrated edge, which was "drastically" bent.

The Commonwealth's chief medical examiner at the time of trial testified to the victim's autopsy report, which had been prepared by another medical examiner. He detailed the victim's injuries, noting that the "most lethal" wound was a cut beginning at the victim's left ear and continuing to his neck. He described the wound as a deep stab wound that injured both the victim's **108 carotid artery and his jugular vein, causing him to bleed to death. He described eleven additional injuries; five were abrasions and six were clearly caused by a sharp object. The additional sharp object injuries included three stab wounds (one to the left arm, one to the left shoulder, and one near the left armpit that was four inches deep) and three, more shallow, incised wounds (one to the outside of the left arm, one to the left side of the chest, and one that extended from the victim's left index finger to his ring finger). The medical examiner described the injury across the victim's fingers as consistent with the victim "trying to deflect" the knife with that hand. He further opined that the victim's wounds were consistent with the knife that was found underneath him, and that the cut on the defendant's hand was consistent with the murder weapon slipping in the defendant's hand.

Police viewed reddish-brown stains throughout the apartment that created a trail from the area of the victim's body down the hallway, through the kitchen, onto the back porch, over the third-floor railing and down to the railings on the second and first floors, through the backyard, and over a chain-link fence. That trail then led through a vacant lot and onto nearby streets, eventually leading to the fire station. Deoxyribonucleic acid testing identified the victim's blood as a possible source of many of these "reddish-brown" stains, including those on the defendant's pants and socks, one on the porch railing on the third floor, and one on a snow pile in the vacant lot.

The defendant testified in his own defense and denied killing the victim. He testified that the victim was "almost like my brother" and that the two had spent the day of January 31 together in the apartment cooking and drinking beer. Although the defendant did not recall how many beers he consumed, his testimony suggested that he was drinking beer from at least 11:30 A.M. until approximately 4 P.M. , at which point he fell asleep on the *787 couch. The defendant testified that he was awakened by loud voices arguing in the apartment and saw two strange men inside the apartment arguing with the victim. One of the men had a knife and began stabbing the victim, causing the defendant to intervene. The defendant said that he was then beaten by the two men and fled the apartment, going directly to the fire station to find help.

The judge denied the defendant's motions for a required finding of not guilty. At the charge conference, trial counsel requested that the judge instruct the jury on intoxication as relevant to both **109 intent and deliberate premeditation. The judge so instructed the jury, but noted to counsel that "there is a paucity of evidence on [intoxication] and there's certainly no scientific evidence that I've seen."

b. Motion for new trial . The defendant filed a motion for a new trial after his conviction, asserting ineffective assistance of counsel based on trial counsel's failure to introduce medical records showing his levels of intoxication. 3 The jury had heard evidence suggesting that the defendant was intoxicated on the night of the murder, including testimony from a woman who had translated for the defendant at the hospital. The woman testified that the defendant's speech "was kind of sluggish," his breath smelled of alcohol, and his eyes were bloodshot. Trial counsel did not mention intoxication in her opening statement, and she referenced the evidence of the defendant's intoxication in her closing argument only in an attempt to explain disparities in the defendant's recollection of events.

Not before the jury, however, were the defendant's medical records, which were discussed before trial at an evidentiary hearing on the defendant's motion to suppress. His breathalyzer results indicated either a 0.298 or a 0.289 blood alcohol content (BAC), his blood alcohol readings were 0.226 and 0.288, and he had a serum alcohol level of 307. The defendant was diagnosed with alcohol intoxication, and a medical technician testified at the suppression hearing that the defendant's breathalyzer results indicated a very high level of intoxication. 4

Trial counsel submitted an affidavit stating that her failure to introduce medical records indicating the defendant's high level of intoxication was not a strategic decision, but rather an oversight on her part. In denying the defendant's motion without an evidentiary hearing, the motion judge implicitly discredited trial counsel's affidavit and determined that the record indicated "that counsel consciously chose not to offer the medical records." Emphasizing how a vigorous intoxication defense would have undermined the defendant's testimony that he was innocent, the motion judge found that trial counsel made a tactical decision to **110 focus on a third-party culprit theory rather than introduce significant evidence of intoxication.

The defendant filed a motion for reconsideration, which was allowed, and an evidentiary hearing on the motion for a new trial followed. Trial counsel testified at the hearing. In an amended decision, the motion judge noted that the defendant's medical records showed "very significant intoxication" and credited trial counsel's *788 testimony that she had intended to present and rely on those medical records as part of her intoxication strategy. He further concluded that trial counsel never intended to call as a witness at trial the medical technician who had administered the defendant's breathalyzer test at the hospital and who had testified at the suppression hearing that the defendant's BAC was "very high."

The judge issued an amended decision on the defendant's motion for a new trial. The judge concluded that trial counsel pursued a dual-defense strategy and had intended to introduce the medical records as a part of a voluntary intoxication defense. The judge held that, because trial counsel inadvertently failed to introduce those records because she believed that they were in evidence, trial counsel's performance fell measurably below that of an ordinary fallible lawyer. 5 See Commonwealth v. Saferian , 366 Mass. 89 , 96, 315 N.E.2d 878 (1974). The judge determined that this error did not warrant a new trial, however, because the jury heard "substantial evidence of intoxication ... and received full instructions to consider the defendant's voluntary intoxication on the questions of intent and deliberate premeditation." The motion judge noted that trial counsel, despite her mistake, skillfully sought to "soft-pedal" the intoxication defense. In other words, she attempted to present evidence of the defendant's intoxication without undermining his claim to innocence.

2. Discussion . a. Sufficiency of the evidence . The defendant maintains that the evidence presented at trial was insufficient to establish deliberate premeditation, and that the judge's denial of his motions for a required finding of not guilty was therefore error. We review the denial of a motion for a required finding of not guilty to determine "whether the evidence offered by the Commonwealth, together with reasonable inferences therefrom, **111 when viewed in its light most favorable to the Commonwealth, was sufficient to persuade a rational jury beyond a reasonable doubt of the existence of every element of the crime charged." Commonwealth v. Whitaker , 460 Mass. 409 , 416, 951 N.E.2d 873 (2011), quoting Commonwealth v. Lao , 443 Mass. 770 , 779, 824 N.E.2d 821 (2005), S . C ., 450 Mass. 215 , 877 N.E.2d 557 (2007) and 460 Mass. 12 , 948 N.E.2d 1209 (2011). See Commonwealth v. Latimore , 378 Mass. 671 , 677-678, 393 N.E.2d 370 (1979).

To prove murder in the first degree on the theory of deliberate premeditation, the Commonwealth must prove "not only that the defendant intended to kill, but that the defendant decided to kill after a period of reflection." Whitaker , 460 Mass. at 418 , 951 N.E.2d 873 . There is "[n]o particular length of time of reflection ... required to find deliberate premeditation, and the decision may be made in only a few seconds." Commonwealth v. Rakes , 478 Mass. 22 , 34, 82 N.E.3d 403 (2017). Circumstantial evidence alone may be sufficient to prove deliberate premeditation. Commonwealth v. Veiovis , 477 Mass. 472 , 480, 78 N.E.3d 757 (2017).

We have recognized as proper considerations in a deliberate premeditation analysis the number and severity of the injuries, including defensive wounds, the procuring of a murder weapon in one room and carrying it to use in another, and the location of a victim's wounds. See Whitaker , 460 Mass. at 419 , 951 N.E.2d 873 *789 ("Deliberate premeditation may be inferred from the nature and extent of a victim's injuries, the duration of the attack, the number of blows, and the use of various weapons"); Commonwealth v. Nolin , 448 Mass. 207 , 216 n.7, 859 N.E.2d 843 (2007) (deliberate premeditation established by "the number and severity of the injuries" to victim's face and head). See also Commonwealth v. Townsend , 453 Mass. 413 , 429, 902 N.E.2d 388 (2009) (victim's fifty-eight stab wounds, including several defensive wounds, sufficient for deliberate premeditation); Commonwealth v. Farley , 432 Mass. 153 , 157-158, 732 N.E.2d 893 (2000), S . C ., 443 Mass. 740 , 824 N.E.2d 797 , cert. denied, 546 U.S. 1035 , 126 S.Ct. 733 , 163 L.Ed.2d 577 (2005) (evidence sufficient to establish deliberate premeditation where "the Commonwealth relied on the fact that the victim had been stabbed multiple times; the location of the wounds ; the duration of the attack; and the inference that the defendant carried a knife from the kitchen to the victim's upstairs bedroom"); Commonwealth v. Andrews , 427 Mass. 434 , 440, 694 N.E.2d 329 (1998) (shooting unarmed victim four times from close range sufficient for deliberate premeditation); Commonwealth v. Watkins , 373 Mass. 849 , 852, 370 N.E.2d 701 (1977) ("that the defendant, after a quarrel, went to the kitchen, picked up a knife, and returned to stab the victim is sufficient" for premeditation). **112 We conclude that, here, the evidence and the reasonable inferences that stem from it, considered in the light most favorable to the Commonwealth, were sufficient to prove that the murder was deliberately premeditated. The knife used to kill the victim was a kitchen knife, and the location of the victim's body between the hallway and a bedroom allows for a reasonable inference that the defendant retrieved the weapon from the kitchen before the killing. The fatal wound was a deep wound to the victim's neck. The victim had at least six other stab or incised wounds, including defensive wounds on his left hand and injuries to his left side, arm, and shoulder, that left the knife "drastically" bent. The judge's denial of the defendant's motions for a required finding of not guilty was proper.

b. Ineffective assistance of counsel . The defendant maintains that he was deprived of effective assistance of counsel at trial because his counsel failed to introduce the defendant's medical records indicating his high level of intoxication on the night of the murder. Because the defendant was convicted of murder in the first degree, we review the claim under the more favorable standard articulated in G. L. c. 278, § 33E, under which we must determine whether there was a substantial likelihood of a miscarriage of justice. Commonwealth v. Alicea , 464 Mass. 837 , 845, 985 N.E.2d 1197 (2013), quoting Commonwealth v. Gonzalez , 443 Mass. 799 , 808, 824 N.E.2d 843 (2005). To make this determination, we ask "whether there was an error in the course of the trial (by defense counsel, the prosecutor, or the judge) and, if there was, whether that error was likely to have influenced the jury's conclusion." Alicea , supra , quoting Commonwealth v. Wright , 411 Mass. 678 , 682, 584 N.E.2d 621 (1992), S . C ., 469 Mass. 447 , 14 N.E.3d 294 (2014). "[W]e consider a defendant's claim even if the action by trial counsel does not constitute conduct 'falling measurably below that ... of an ordinary fallible lawyer.' " Gonzalez , supra at 808-809, 824 N.E.2d 843 , quoting Commonwealth v. MacKenzie , 413 Mass. 498 , 517, 597 N.E.2d 1037 (1992). 6

*790 We agree with the motion judge that trial counsel's oversight in failing to introduce the defendant's medical records was error. We **113 differ from the motion judge's ruling that it was a reasonable strategic decision by trial counsel to not supplement those records with expert testimony regarding the relationship between the BAC reading and intoxication. See Commonwealth v. Colturi , 448 Mass. 809 , 817-818, 864 N.E.2d 498 (2007) ("the jury would be left to guess" at meaning of defendant's BAC absent expert testimony).

We recognize that trial counsel was faced with a difficult strategic predicament at trial. The defendant, as is his constitutional right, testified in his own defense. See Commonwealth v. Brown , 479 Mass. 163 , 171, 92 N.E.3d 1189 (2018). He unequivocally denied killing the victim. Therefore, the motion judge found that trial counsel decided to pursue the voluntary intoxication defense ever mindful of its potential to undercut the primary defense of innocence. See Commonwealth v. Morales , 453 Mass. 40 , 43-45, 899 N.E.2d 96 (2009). The motion judge ultimately concluded that trial counsel "soft-pedaled" the intoxication defense so as to leave the question in the jury's hands without overemphasizing the defendant's intoxication. There are situations where the appropriate course of action would be to pursue alternative defenses -- "rid[e] 'two horses' " into battle. Commonwealth v. Kolenovic , 471 Mass. 664 , 676, 32 N.E.3d 302 (2015), S . C ., 478 Mass. 189 , 84 N.E.3d 781 (2017). Trial counsel chose such a strategy in the present case. However, having made that decision, trial counsel was required to pursue each of those defenses effectively. Although the choice to soft-pedal the intoxication defense was reasonable, her failure to introduce the most compelling evidence of the defendant's intoxication was error.

Nevertheless, the jury were presented with evidence that allowed them to consider the question of intoxication. Most notably, the defendant testified that he had spent at least four and one-half hours drinking beer with the victim on the day of the murder, although he did not know how many beers he had consumed. The woman who interacted with the defendant at the hospital noted that his breath smelled of alcohol, his eyes were bloodshot, and his speech was sluggish. Although there is no dispute that the defendant drank and showed signs of intoxication, there also was ample evidence before the jury to show that the defendant was not intoxicated to the point of debilitation. He was coherent in his interactions with medical personnel and was physically able to climb to the ground from the deck of his third-floor apartment.

The judge instructed the jury that they could consider the mitigating effect of voluntary intoxication, stating, "[A]ny evidence **114 that you find credible of the defendant's consumption of alcohol, you may consider that evidence in determining whether the defendant specifically intended to commit an offense or whether the defendant deliberately premeditated the killing of [the victim]."

Because the defendant's intoxication may have been probative of a lack of capacity to formulate the intent necessary to prove murder by deliberate premeditation, the jury could have found the defendant guilty of a lesser offense or acquitted *791 him had they determined his level of intoxication to be "debilitating." Commonwealth v. Carter , 475 Mass. 512 , 524, 58 N.E.3d 318 (2016). See Commonwealth v. Murphy , 442 Mass. 485 , 504, 813 N.E.2d 820 (2004). That debilitation must be to the point that it "could support a reasonable doubt as to the defendant's ability to form the requisite criminal intent" or to deliberately premeditate. Commonwealth v. Lennon , 463 Mass. 520 , 523, 977 N.E.2d 33 (2012). As evidenced by the request for an intoxication instruction, trial counsel pursued this defense. Trial counsel also pointed to the defendant's intoxication in her closing argument as a possible explanation for disparities in the defendant's account of the night in question. It was an error by counsel, therefore, to not introduce the defendant's medical records and accompanying expert testimony to explain the significance of his BAC, particularly because expert testimony likely would have shown the BAC to demonstrate a high level of intoxication. See Commonwealth v. Wall , 469 Mass. 652 , 671, 15 N.E.3d 708 (2014) (testimony that BAC of 0.21 per cent "very high").

Therefore, the question is whether this error created a substantial likelihood of a miscarriage of justice. Despite trial counsel's errors, her approach to the trial was consistent with the motion judge's determination that she "soft-pedaled" the intoxication defense. At the time of the closing argument, trial counsel believed that the defendant's medical records had been entered in evidence. Despite this belief, she opted not to argue that the defendant's intoxication diminished his capacity to form intent or deliberately premeditate. This suggests that although trial counsel chose to ride two horses into battle, she focused primarily on only one of them in closing argument. It "is a well-known and time-honored approach" to avoid emphasizing a defense that would undermine a primary defense theory. Commonwealth v. Walker , 443 Mass. 213 , 228, 820 N.E.2d 195 (2005). See Commonwealth v. Spray , 467 Mass. 456 , 473-475, 5 N.E.3d 891 (2014). When considering a defense attorney's strategy at trial, "we conduct our review with some deference **115 to avoid characterizing as unreasonable a defense that was merely unsuccessful." Kolenovic , 471 Mass. at 673 , 32 N.E.3d 302 , quoting Commonwealth v. Valentin , 470 Mass. 186 , 190, 23 N.E.3d 61 (2014). "This measure of deference is as it must be because, ultimately, counsel alone has the benefit of the full factual picture that dictates the choice of those matters to be revealed to the fact finder and those that are better left unexposed to court room scrutiny. From that vantage point, counsel 'knows best how to defend a client.' " Kolenovic , supra , quoting Commonwealth v. Glover , 459 Mass. 836 , 843, 948 N.E.2d 415 (2011). 7 *792 We recently considered a similar question in Commonwealth v. Montrond , 477 Mass. 127 , 134-136, 75 N.E.3d 9 (2017), where the defendant's trial counsel argued that a shooting was accidental while choosing not to introduce evidence of the defendant's possible intoxication. Although the evidence of intoxication in the Montrond case, which we termed to be "tepid at best," 8 may have been "somewhat helpful to the defendant[ ]," we concluded that it "hardly would have given rise to a compelling inference that the defendant was so intoxicated that he could not appreciate the need to check the safety lock before pointing a loaded gun at someone's head and pulling the trigger." Id . at 136 , 75 N.E.3d 9 . The evidence of intoxication here rises above the tepid level seen in the Montrond case. It is not, however, so strong that we are concerned that **116 the failure to admit medical evidence of the defendant's intoxication would have been likely to influence the jury's decision.

The defendant's BAC, although high, would have been just one of many factors the jury could have considered in determining whether he was intoxicated to the point of debilitation, such that there was "reasonable doubt as to [his] ability to form the requisite criminal intent." 9 Lennon , 463 Mass. at 523 , 977 N.E.2d 33 . Although evidence of the defendant's BAC would have been the strongest evidence to suggest he was debilitated by alcohol, the entirety of the evidence before the jury coupled with trial counsel's reasonable decision to "soft-pedal" the intoxication defense leaves us unpersuaded that the defendant's medical records would have influenced the jury's verdict. We conclude that defense counsel's error in failing to admit them in evidence did not create a substantial likelihood of a miscarriage of justice.

c. Closing argument . The defendant asserts that the prosecutor made improper remarks in his closing argument. Because trial counsel did not object to the closing argument at trial, we review to determine whether there was an error and, if so, whether that error created a substantial likelihood of a miscarriage of justice. See Veiovis , 477 Mass. at 488 , 78 N.E.3d 757 .

"To be sure, 'prosecutors are held to a stricter standard than are errant defense counsel and their clients.' " Commonwealth v. Goitia , 480 Mass. 763 , 775, 108 N.E.3d 993 (2018), quoting *793 Commonwealth v. Arroyo , 442 Mass. 135 , 147, 810 N.E.2d 1201 (2004). The defendant contends that three portions of the prosecutor's closing argument were improper: (1) a reference to the twenty-month period between the murder and trial, which accompanied the prosecutor's suggestion that the defendant now had no more time to produce a different explanation for the victim's death, such as a third party; (2) the statement that the jury should use their "moral compass" and "gut" in evaluating the evidence and testimony; (3) the statement that the defendant's intoxication was "not an excuse" for the murder, "[w]hether he had two beers, four beers, or [forty-four] beers." **117 We discern no error in the prosecutor's reference to the twenty-month period between the murder and trial. The prosecutor discussed that period of time while highlighting the disparities in the defendant's story throughout that time period, including the detailed nature of the defendant's testimony at trial as compared to the defendant's statements at the time of the killing. This portion of the argument was grounded in the evidence presented at trial and was a reasonable commentary on the defendant's credibility. See Commonwealth v. Tu Trinh , 458 Mass. 776 , 788, 940 N.E.2d 871 (2011).

The prosecutor's call for jurors to follow their moral compass is troublesome, but we conclude that even if it was error, it did not create a substantial likelihood of a miscarriage of justice. The relevant portion of the prosecutor's argument is as follows:

"You ... impartial jurors also have something inside you besides your commonsense. And it's what I refer to as a moral compass. That little moral compass, when you know based on your gut that something's wrong, that something is askew with a certain situation. And I suggest to you, ladies and gentlemen, based on the evidence in this case, that your moral compass goes haywire when you consider what was in that apartment and his actions and his statements in the aftermath of 8:40 P.M. approximately on January 31st of 2005. You know, based on your everyday life experience that something was horribly amiss with his behavior, his statements and his actions that night."

It is well established that it is proper to ask a jury to rely on their common sense and life experience in assessing evidence and credibility. See Lao , 460 Mass. at 22 , 948 N.E.2d 1209 . The prosecutor's invocation of the jury's "moral compass" however, was a step beyond an ordinary call for the jury to rely on their life experience and common sense, and approached an improper appeal to the jury's emotions. It was thus better left unsaid. We need not determine whether it was improper, however, because even if it was, we are unpersuaded that this sole misstatement would undermine the jury's verdict, particularly where the jury were properly instructed that closing arguments are not considered evidence. See Commonwealth v. Kozec , 399 Mass. 514 , 517, 505 N.E.2d 519 (1987). Considering the statement in the context of the rest of the closing argument and the trial as a whole, any prejudice was minor and does not warrant reversal. See Commonwealth v. Braley , 449 Mass. 316 , 329, 867 N.E.2d 743 (2007).

**118 Finally, although the prosecutor's statement that the defendant's possible intoxication did not "excuse" his actions was technically consistent with the law, we are concerned that the prosecutor crossed the line into a misstatement of the law. See Commonwealth v. Rollins , 470 Mass. 66 , 81, 18 N.E.3d 670 (2014), quoting Commonwealth v. Bins , 465 Mass. 348 , 367, 989 N.E.2d 404 (2013) ("We have repeatedly warned that, in 'closing argument,' '[l]awyers shall not and must not misstate principles *794 of law' "). It is true that intoxication does not serve as an excuse in the way that, for instance, self-defense does. See Commonwealth v. Walczak , 463 Mass. 808 , 817, 979 N.E.2d 732 (2012) (Lenk, J., concurring), quoting Model Jury Instructions on Homicide 8 (1999) (self-defense can serve as legal justification for a killing, rather than mitigating factor). In the context of the trial as a whole and the rest of the prosecutor's closing argument, however, the prosecutor's statement regarding excuse appears to be a colloquial suggestion that the defendant's intoxication could neither excuse his actions nor diminish his culpability. Thus, the prosecutor's pronouncement that the defendant's intoxication was no excuse for his actions adversely had an impact on the effectiveness of the judge's intoxication instruction.

That error, however, was not significant enough that our confidence in the jury's decision is shaken. The defendant's trial counsel referred to the defendant's intoxication in her closing argument only as a means to explain discrepancies between his multiple versions of the night of the killing. The prosecutor's statement, while erroneous, was a brief, isolated statement in his closing argument and was not egregious enough to infect the whole of the trial. Finally, the judge properly instructed the jury that closing arguments are not evidence, and it is well-established that "[t]he jury are presumed to follow instructions." Commonwealth v. Hernandez , 473 Mass. 379 , 392, 42 N.E.3d 1064 (2015), citing Commonwealth v. Andrade , 468 Mass. 543 , 549, 11 N.E.3d 597 (2014). See Kozec , 399 Mass. at 517 , 505 N.E.2d 519 . With all that in mind, we conclude that the prosecutor's error was unlikely to have influenced the jury's ultimate decision and therefore did not create a substantial likelihood of a miscarriage of justice.

d. Review pursuant to G. L. c. 278, § 33E . The defendant asks us, in the alternative to his grounds for appeal addressed above, to exercise our authority under G. L. c. 278, § 33E, to either order a new trial or reduce his conviction to murder in the second degree. It is our statutory duty "to consider broadly the whole case on the law and the facts to determine whether the verdict is **119 consonant with justice" (quotations and citation omitted). Commonwealth v. Vargas , 475 Mass. 338 , 363-364, 57 N.E.3d 920 (2016). "[F]or any ... reason that justice may require, [we may] (a) order a new trial or (b) direct the entry of a verdict of a lesser degree of guilt, and remand the case to the superior court for the imposition of sentence." G. L. c. 278, § 33E. Although we recognize that our power under § 33E"is to be exercised sparingly," Commonwealth v. Seit , 373 Mass. 83 , 95, 364 N.E.2d 1243 (1977), we are persuaded that, here, a verdict of murder in the second degree would be more "consonant with justice." Vargas , supra .

Our duty under § 33E"does not ... convert this court into a second jury." Commonwealth v. Penn , 472 Mass. 610 , 621, 36 N.E.3d 552 (2015), cert. denied, --- U.S. ----, 136 S.Ct. 1656 , 194 L.Ed.2d 773 (2016), quoting Commonwealth v. Franklin , 465 Mass. 895 , 916, 992 N.E.2d 319 (2013). But this case presents one of the rare situations in which we conclude that the jury's verdict of murder in the first degree was supported by the evidence, but was not consonant with justice. In Commonwealth v. Lanoue , 392 Mass. 583 , 591-592, 467 N.E.2d 159 (1984), S . C ., 400 Mass. 1007 , 513 N.E.2d 683 (1987) and 409 Mass. 1 , 563 N.E.2d 1367 (1990), we exercised our authority under § 33E where the evidence was sufficient to convict the defendant of murder in the first degree on a theory of deliberate premeditation, but the evidence was entirely circumstantial and *795 "thin." There was evidence that the victim in the Lanoue case had sustained multiple injuries and that the defendant was intoxicated at the time of the killing, and although the judge instructed the jury on the relationship between intoxication and deliberate premeditation, the adequacy of the instruction was questionable. Id . We reduced the defendant's verdict to murder in the second degree. Id . at 592, 467 N.E.2d 159 .

Due to the unusual nature of this case, we reach the same conclusion as we did in the Lanoue case. 10 There was sufficient evidence before the jury to support their conclusion that the **120 defendant had killed the victim after deliberate premeditation, but it was far from compelling. In addition to there being no definitive evidence as to what happened in the apartment on the night of the killing other than the victim being stabbed multiple times, there is nothing to suggest that there was any ill will between the defendant and the victim, or to suggest that there was any motive for the killing. See Seit , 373 Mass. at 94 , 364 N.E.2d 1243 (whether defendant and victim had good relationship prior to killing relevant to mitigation analysis under § 33E ). The defendant's intoxication is another factor that we consider, particularly where it was incompletely presented as a defense and where the prosecutor made an inappropriate statement about it in his closing argument. See Commonwealth v. Ransom , 358 Mass. 580 , 583, 266 N.E.2d 304 (1971) (alcohol having "probably played a part" in murder proper consideration in § 33E analysis). In sum, we conclude that a conviction of murder in the second degree is more consonant with justice and we reduce the verdict accordingly.

The case is remanded to the Superior Court, where the verdict of murder in the first degree and sentence imposed shall be vacated. A verdict of guilty of murder in the second degree shall be entered, and a sentence imposed.

So ordered .