dissenting. I agree with parts I, II, IV and V of the majority opinion. I disagree, however, with part III. I would conclude affirmatively that expert opinion testimony that a homicide victim was likely to have behaved violently on a particular occasion as a result of being under the influence of drugs and alcohol is admissible where the defendant properly has raised the issue of self-defense and that the exclusion of such evidence in this case constituted reversible error. I, therefore, respectfully dissent.
I begin with a brief review of the facts as the jury reasonably could have found them. At noon on May 3, 1994, the defendant approached a bank of public telephones located on the New Haven town green. As he began to use one of the telephones, the victim, Michael Stewart, approached him and said, “ ‘Get off the phone. I beeped somebody.’ ” When the defendant continued to dial the telephone, the victim slapped his face. The defendant then punched the victim and a fistfight ensued.
Monique McNeil, an eyewitness, testified that the defendant, who was smaller in stature than the victim, appeared to be “winning” the fight. At some point during the fight, McNeil observed the victim reaching into a nearby trash can. The defendant testified that he had seen an object in the can and that he thought the victim was reaching to get a gun. The defendant further testified that he stabbed the victim in order to prevent him from retrieving the gun. McNeil testified that the fight escalated once the victim reached into the garbage can, but that she could not recall whether there was anything *780in the victim’s hand when it emerged from the garbage can.
Another eyewitness, Martin Tchakirides, who was seated in his car approximately forty to sixty feet away when the sound of raised voices drew his attention to the altercation, testified that neither party appeared to be winning the fight until the defendant pulled out a knife. He further testified that the victim had tried to pull away when the defendant brandished the knife. Tchakirides also testified that he saw the defendant making slashing motions with his hand.
Detective William Badger, a thirty year veteran of the New Haven police department, was on duty and was sitting nearby in his police car at the time of the fight. When he first noticed the altercation, the defendant was “straddling [the victim] like he was riding a horse.” The victim was lying still at that point and did not appear to be fighting back. Badger saw the defendant stab the victim twice with a switchblade and Badger grabbed the defendant’s arm as he was attempting to stab the victim a third time. Badger identified himself as a police officer and seized the knife from the defendant’s hand. The defendant then stated to Badger that he was merely trying to defend himself. Badger testified that the defendant was cooperative and noncombative. The victim was then transported to Yale-New Haven Hospital, where he died approximately twelve hours later following surgery.
Malka Shah, the associate medical examiner who performed the autopsy on the victim’s body, testified that he had bled to death as a result of stab wounds to his chest and abdomen. Shah also noted two “defensive wounds” on the victim’s left arm. Shah further testified that certain needle marks on the victim’s hand and elbow were not the result of medical treatment.
*781Detective Ralph DiNello, of the New Haven police department, testified that he advised the defendant of his rights and that the defendant agreed to speak with him regarding the stabbing incident. During this interview, the defendant claimed that he had stabbed the victim in self-defense but never mentioned that the victim had reached inside the trash can. When DiNello asked him whether the victim had a weapon, the defendant specifically replied that “he never saw nor was any other weapon used.”
Hospital and autopsy reports indicated that the victim’s body contained cocaine, morphine, meperidine1 and alcohol, and that the victim had told nurses at the hospital that he was using five or six bags of heroin a day and drinking a six pack of beer daily. The defense proffered the testimony of James Merikangas, a neurologist and psychiatrist, who, based on his review of these records, stated that the victim’s body contained morphine, cocaine and alcohol at the time of the altercation. Merikangas testified that the presence of morphine constitutes a positive test for heroin and that drug users simultaneously use opiates like morphine together with cocaine, which combination is called a “speedball.” Merikangas further testified that, because of the presence of cocaine and alcohol in the victim’s body, there was a high probability that the victim was the aggressor and that he was violent during the fight with the defendant. This testimony was offered for the purposes of showing that the victim was the aggressor, that the defendant did not have a duty to retreat and that it was reasonable for the defendant to use deadly force in self-defense.
*782The trial court ruled that the expert testimony was inadmissible because: (1) it was undisputed that the victim was the initial aggressor; (2) it was not the type of opinion testimony that was admissible on this issue because it was not based on personal knowledge of the victim’s character; and (3) the defendant had not established a sufficient foundation for the evidence. This ruling encompasses two distinct issues: first, whether evidence of the victim’s violent character is admissible when the state has conceded that the victim initiated the fight; and second, if character evidence generally is admissible under that circumstance, whether expert opinion testimony that the presence of drugs and alcohol in the victim’s blood made him more likely to be violent is an admissible form of character evidence. Finally, if this evidence is admissible, it must be determined whether the trial court’s decision to exclude it requires reversal of this case.
I
I will first address the question of whether evidence of the victim’s violent character is admissible for the purpose of proving self-defense when the state has conceded that the victim initiated the fight when he slapped the defendant. I would conclude that it is.
In order to be admissible, evidence must be both probative and material to the case. “Relevant evidence is evidence that has a logical tendency to aid the trier in the determination of an issue. . . . One fact is relevant to another if in the common course of events the existence of one, alone or with other facts, renders the existence of the other either more certain or more probable.” (Citation omitted; internal quotation marks omitted.) State v. Prioleau, 235 Conn. 274, 305, 664 A.2d 743 (1995). In addition, “[e]vidence is admissible only to prove material facts, viz., those facts directly in issue or those probative of matters in issue; evidence offered *783to prove other facts is ‘immaterial.’ ” C. Tait & J. LaPlante, Connecticut Evidence (2d Ed. 1988) § 8.1.2, p. 226.
“[T]he trial court has broad discretion in ruling on the admissibility [and relevancy] of evidence. . . . The trial court’s ruling on evidentiary matters will be overturned only upon a showing of a clear abuse of the court’s discretion. . . . We will make every reasonable presumption in favor of upholding the trial court’s ruling, and only upset it for a manifest abuse of discretion. ” (Citations omitted; internal quotation marks omitted.) State v. Coleman, 241 Conn. 784, 789, 699 A.2d 91 (1997). “The discretion accorded the trial court . . . though necessarily broad, is not unlimited. ‘The trial judge’s discretion, which is a legal discretion, should be exercised in conformity with the spirit of the law and in a manner to subserve and not to impede or defeat the ends of substantial justice.’ ” State v. Carter, 228 Conn. 412, 421, 636 A.2d 821 (1994). The application of an incorrect legal standard by the trial court constitutes a clear abuse of discretion. Unkelbach v. McNary, 244 Conn. 350, 367, 710 A.2d 717 (1998) (trial court applied incorrect legal standard); see also State v. Miranda, 176 Conn. 107, 115, 405 A.2d 622 (1978) (although trial court applied legal standard in effect at time of trial, change of law on appeal necessitates reversal).
Balanced against the trial court’s discretionary authority to determine the relevancy of evidence is the defendant’s constitutional right to present a defense. “Consistent with the defendant’s [constitutional] right fairly to inform the jury of facts material to the defense, [i]t has long been the law in this state that, in a homicide prosecution, an accused may introduce evidence of the violent, dangerous or turbulent character of the victim to show that the accused had reason to fear serious harm, after laying a proper foundation by adducing evidence that he acted in self-defense and that he was *784aware of the victim’s violent character. State v. Miranda, [supra, 176 Conn. 109], citing State v. Padula, 106 Conn. 454, 456-57, 138 A. 456 (1927). More recently, we joined a majority of courts when we expanded this rule to allow the accused to introduce evidence of the victim’s violent character to prove that the victim was the aggressor, regardless of whether such character evidence had been communicated to the accused prior to the homicide. State v. Miranda, supra [109]. In Miranda, we determined that the victim’s violent character could be proved by reputation testimony, by opinion testimony, or by evidence of the deceased’s convictions for crimes of violence, irrespective of whether the accused knew of the deceased’s violent character or of the particular evidence adduced at the time of the death-dealing encounter. Id., 114. . . . State v. Smith, 222 Conn. 1, 17-18, 608 A.2d 63, cert. denied, 506 U.S. 942, 113 S. Ct. 383, 121 L. Ed. 2d 293 (1992).” (Internal quotation marks omitted.) State v. Carter, supra, 228 Conn. 422-23.
When a defendant has established that, by disallowing evidence, the trial court improperly interfered with his constitutional right to present a defense, the state must demonstrate beyond a reasonable doubt that the violation was harmless. See id., 427-28. Where a nonconstitutional evidentiary ruling is concerned, however, “the defendant must show not only that the trial court improperly barred him from using [the proffered evidence], but also that the court’s ruling caused him substantial prejudice. State v. Hines, [243 Conn. 796, 801, 709 A.2d 522 (1998)]; State v. Small, [242 Conn. 93, 109-10, 700 A.2d 617 (1997)]; State v. Beliveau, [237 Conn. 576, 592, 678 A.2d 924 (1996)].” State v. Askew, 245 Conn. 351, 371, 716 A.2d 36 (1998). Because I would conclude that the trial court abused its discretion in excluding the proffered testimony and that exclusion of this evidence resulted in substantial prejudice and *785injustice, I do not reach the question of whether that decision deprived the defendant of his constitutional right to present a defense. See State v. Genotti, 220 Conn. 796, 804, 601 A.2d 1013 (1992) (court need not reach constitutional issue when issue disposed of on evidentiary and procedural grounds); State v. Williams, 200 Conn. 310, 322, 511 A.2d 1000 (1986) (constitutional issues not considered unless absolutely necessary to decision of case).
The relevancy of the proffered evidence in this case must be measured against the purpose for which it was offered; namely, to prove that the defendant acted in self-defense and that he justifiably used deadly force against the victim. “Pursuant to § 53a-19 (a) ... a person may justifiably use deadly physical force in self-defense only if he reasonably believes both that (1) his attacker is using or about to use deadly physical force against him, or is inflicting or about to inflict great bodily harm, and (2) that deadly physical force is necessary to repel such attack. State v. DeJesus, 194 Conn. 376, 389 n.13, 481 A.2d 1277 (1984); see also State v. Carter, 232 Conn. 537, 546, 656 A.2d 657 (1995).” (Emphasis in original.) State v. Prioleau, supra, 235 Conn. 285-86. Section 53a-19 (b) (1) provides further that the use of deadly force is not justified if the defendant “knows that he can avoid the necessity of using such force with complete safety ... by retreating ??
The state argues that this court’s decision in Miranda permits the defendant to introduce evidence of the victim’s violent character, without showing the defendant’s awareness of that trait, only for the purpose of proving that the victim was the initial aggressor. Because it has conceded that the victim started the altercation, the state argues that the proffered evidence is inadmissible. I disagree.
I begin by noting that the state’s concession that the victim started the altercation does not, by itself, render *786the proffered evidence irrelevant to that specific issue. The mere fact that one party has conceded an issue does not bar the opposing party from introducing corroborating evidence on that point. C. Tait & J. LaPlante, supra, § 6.8.1, pp. 136-37; see also Wagner v. Clark Equipment Co., 243 Conn. 168, 195, 700 A.2d 38 (1997) (evidence admissible to corroborate other direct evidence in case); State v. Campbell, 93 Conn. 3, 8-9, 104 A. 653 (1918) (opposing party not required to dispense with own proof). The defendant may, therefore, introduce evidence tending to corroborate the fact that the victim, without provocation, initiated the fight when he slapped the defendant.
Additionally, it is significant that the state conceded only that the victim had initiated the fight, not that he remained the aggressor when the defendant pulled out a knife and stabbed the victim. The fact that the victim was the first to use physical force may, arguably, be enough to raise the issue of self-defense, but it is not sufficient to absolve the defendant of criminal liability. “It is not the law . . . that the person who first uses physical force is necessarily the initial aggressor under § 53a-19 (c) (2).” (Emphasis added.) State v. Jimenez, 228 Conn. 335, 340, 636 A.2d 782 (1994). The issue is, therefore, not who was the first to use physical force but whether, at the relevant time during the altercation, the defendant was justified in using deadly force. The state’s concession that the victim started the fight is, therefore, unresponsive to the purpose for which the defendant sought to introduce the proffered evidence.
Furthermore, it is significant that the defendant in the present case seeks to excuse, not merely the use of force, but the use of deadly force as provided in § 53a-19. Under the terms of that statute, the defendant will not be excused from criminal liability unless he did not have a duty to retreat and unless his use of deadly force was authorized under the circumstances. *787In Miranda, this court characterized the question of who was the aggressor as “vital” because it was contested and because the evidence was offered to prove that specific point. State v. Miranda, supra, 176 Conn. 110. There is nothing in the language of that opinion, however, indicating that the court’s reasoning may not apply equally to other contested issues that are “vital” to the defendant’s claim of self-defense. The question of whether the defendant had a duty to retreat under the circumstances and whether he was justified in using deadly force are necessary factors in a claim of self-defense and, therefore, are “vital” to that defense. A concession by the state as to the question of who initiated the altercation does not negate the importance of these other “vital” issues.
Evidence of the degree of the victim’s aggressiveness would be probative as to the issue of the amount of force necessary for the defendant to repel the attack and as to the likelihood that he could retreat with complete safety. Although a component of the issue of whether the defendant had a duty to retreat and whether it was appropriate to use deadly force necessarily hinges on what the defendant believed to be necessary, his ability to prove that his actions were justified depends, to a large degree, on what the jury believed the victim actually did. As this court stated in Miranda, in connection with that portion of the opinion holding that the defendant need not know of the victim’s violent character, “the question is what the deceased probably did, not what the defendant probably thought the deceased was going to do. The inquiiy is one of objective occurrence, not of subjective belief.” (Internal quotation marks omitted.) Id. If the jury believed the victim was violent, then it would have been more likely to credit the defendant’s testimony that he reasonably feared for his life when the victim deliberately reached into a garbage can to retrieve what the defendant thought was a *788weapon, and that the defendant did not believe that he could retreat with complete safety. I would conclude that, despite the state’s concession that the victim had been the first to use physical force, evidence of the victim’s violent character was relevant to show that he was the aggressor and that he was likely to behave violently.
II
1 next address the question of whether the expert opinion testimony offered in this case is an admissible form of character evidence. The state argues that, although opinion testimony is generally admissible to prove the victim’s character for violence, expert opinion testimony of the kind offered by this defendant is not admissible because it is not based on personal knowledge of the victim’s character.
Where character evidence is otherwise admissible, it may take the form of opinion testimony. “In theory, a trait of character may be proved in three ways: (1) by testimony concerning the individual’s reputation in the community as to the trait; (2) by testimony of those who have had an opportunity to form, and have formed, an opinion as to whether the individual possessed the trait; and (3) by evidence of specific acts of the individual under similar circumstances, from which the existence of the character trait may be inferred. . . .2 Although the second method is not allowed in a number of states, Connecticut adopted it in Richmond v. Norwich, 96 Conn. 582, 594, 115 A. 11 [1921], on the rationale that ‘[p]ersonal observation and personal knowledge are a more trustworthy reliance than general reputation.’ ” (Citations omitted; emphasis added.) State v. Miranda, supra, 176 Conn. 111-12. The state *789argues that, although opinion testimony is generally admissible, expert opinion testimony of the type offered is not admissible because it is not based on personal knowledge of the victim’s character. I disagree.
This court has stated explicitly that the “degree of inebriation” of the victim is relevant where the defendant has claimed self-defense. State v. Gooch, 186 Conn. 17, 23, 438 A.2d 867 (1982). Evidence of the victim’s drug use has been held not to support a claim of self-defense, however, when the amounts found in the victim’s bloodstream were insufficient to show that the victim was under the influence of drugs or alcohol at the time of death; State v. Walton, 41 Conn. App. 831, 838-39, 678 A.2d 986 (1996) (excluding evidence of cocaine breakdown products in victim’s blood showing that victim had ingested cocaine “ ‘many hours’ ” before death); and when no evidence was offered as to whether trace amounts of drugs found in the victim’s body bore on the issue of who was the aggressor. State v. Bember, 183 Conn. 394, 406, 439 A.2d 387 (1981);3 cf. State v. Smith, supra, 222 Conn. 18 (although evidence of victim’s intoxication admitted at trial, evidence of violent act occurring hours before shooting properly excluded). In the present case, however, there was evidence offered suggesting that the victim was under the influence of drugs at the time of the incident and not merely that he had ingested some drugs many hours before death. Merikangas testified, based on the toxological evidence, that the victim’s body contained sufficient quantities of cocaine, morphine and alcohol at the *790time of the altercation to cause him to behave violently. I would, therefore, conclude that evidence of the presence of drugs and alcohol in the victim’s body at the time of the altercation is admissible for the purpose of supporting the defendant’s self-defense claim.
I now turn to the question of whether expert opinion testimony linking the presence of drugs and alcohol in the victim’s body to his violent behavior is an admissible form of character evidence when offered to show that the defendant justifiably used deadly force in self-defense. The state argues that Merikangas’ statements are inadmissible because his opinion was not based on personal knowledge of the victim’s character. I disagree.
“Generally, expert testimony is admissible if (1) the witness has special skill or knowledge directly applicable to a matter in issue, (2) that skill or knowledge is not common to the average person, and (3) the testimony would be helpful to the court or jury in considering the issues.” State v. Kemp, 199 Conn. 473, 476, 507 A.2d 1387 (1986). The question of how particular quantities of various chemical substances affect a person’s propensity for violence requires specialized medical knowledge that would not be within the jury’s understanding. Expert testimony is ordinarily not just permitted, but required under such circumstances. See, e.g., Collette v. Collette, 177 Conn. 465, 471, 418 A.2d 891 (1979). Such testimony need not be based on a direct examination of the patient. Shelnitz v. Greenberg, 200 Conn. 58, 67, 509 A.2d 1023 (1986). Furthermore, “[a]n expert opinion or conclusion need not be based on absolute certainty so long as the opinion is stated in terms of the probable and not merely the possible.” C. Tait & J. LaPlante, supra, § 7.16.1, p. 172.
*791In the present case, Merikangas spoke from knowledge that this particular defendant had significant amounts of a variety of illegal substances in his bloodstream at the time of the altercation and that he habitually used five bags of heroin and drank a six pack of beer daily. Additionally, Merikangas testified as to the synergistic effect such drugs would have on violent behavior. On the basis of specific information about the victim, he testified that “alcohol reduces one’s impulse control and chronic alcohol reduces it more, and cocaine, particularly, acutely causes violence . . . .” Merikangas specifically stated that there was a high probability that the victim was the aggressor and that he was violent. This evidence is probative of whether and to what extent the victim was likely to behave in an aggressive and violent manner and tends to corroborate the defendant’s own testimony that he felt it was necessary to stab the victim in order to save himself.
Merikangas’ knowledge of the victim was sufficient to allow him to render an expert opinion on the effect that specific quantities of drugs and alcohol would likely have on the victim. In State v. Miranda, supra, 176 Conn. 112, this court noted that opinion testimony offered to prove the victim’s character for violence is particularly reliable because it is based on personal knowledge of the defendant rather than on general reputation. Furthermore, it is permissible to base expert opinion testimony on knowledge gleaned from hospital records rather than from a physical examination of the subject. Shelnitz v. Greenberg, supra, 200 Conn. 67. I would conclude, therefore, that expert opinion testimony of the type proffered by this defendant was admissible to prove that a homicide victim was likely to have been violent.
For the foregoing reasons, I also would conclude that the defendant has laid a proper scientific foundation *792for this expert opinion testimony.4 The state has not claimed that Merikangas was not qualified to render an opinion on the relationship between drugs and violence. The only remaining issue, therefore, is whether Merikangas based his opinion on appropriate facts under the circumstances of this case. As I have already stated, Merikangas based his opinion on specific knowledge that this particular victim was, by his own admission and as corroborated by forensic evidence, under the influence of alcohol, cocaine and heroin at the time of the altercation. On the basis of his professional training and experience, Merikangas concluded that the combination of drugs and alcohol in the victim’s system at the time of the altercation made it highly probable that he had been violent at the time in issue.
Furthermore, Merikangas did not reach the ultimate question of whether the victim’s behavior was such that the defendant’s use of deadly force was justified under the circumstances. Contrast State v. Walton, 227 Conn. 32, 59-61, 630 A.2d 990 (1993). Nor did he reach the question of whether the victim was likely to have attempted to use a weapon during the fight. He stated his opinion more generally in terms of the probability that the victim would behave in a violent manner. Although such testimony allows the fact finder to draw *793an inference that the victim was actually violent on a particular occasion based on testimony that merely indicates that violent behavior was likely, this argument is no more applicable to expert opinion testimony than it is to other forms of character evidence, which we have already concluded are properly admissible. State v. Miranda, supra, 176 Conn. 111-12. Furthermore, the evidence in this case of drug and alcohol ingestion and Merikangas’s expertise regarding its effects on human behavior, all of which served as the basis for his opinions, were more accurate and precise than the factors generally relied upon for other forms of admissible character evidence. See id.
Finally, even when an expert renders an opinion based on a hypothetical question that does not include all relevant evidence, his testimony is permitted so long as the question does not misrepresent the facts or mislead the fact finder. Shelnitz v. Greenberg, supra, 200 Conn. 77-78. This is so because it is the function of cross-examination to reveal any weaknesses in such testimony by changing or adding to the variables relied upon by the expert witness. Pischitto v. Waldron, 147 Conn. 171, 177, 158 A.2d 168 (1960) (“[c]ross-examination of an expert . . . often brings out the effect . . . which would or should result if one or more of the facts assumed were eliminated or if additional facts were assumed”). The state has not claimed that the facts relied upon by Merikangas were untrue or misleading. As such, any assertion that Merikangas’ testimony was based on improper or unstated assumptions would more properly have been addressed through cross-examination or through the presentation of opposing expert testimony than through exclusion of the evidence. I, therefore, would conclude that the expert opinion testimony offered by the defendant was admissible.
*794III
I turn finally to the question of whether the trial court’s exclusion of the evidence, even if improper, entitles the defendant to a new trial. It has long been required that, in order to be entitled to a new trial, the defendant must demonstrate that the trial court’s ruling was “both wrong and harmful.” DeCarufel v. Colonial Trust Co., 143 Conn. 18, 21, 118 A.2d 798 (1955). The state argues that, even if the proffered testimony is admissible, the defendant cannot demonstrate that he was harmed by the trial court’s ruling. I disagree.5
As I previously have stated, the defendant’s right to present evidence of the victim’s violent character is *795rooted in his constitutional right to present a defense; State v. Carter, supra, 228 Conn. 422-23; and evidence of the victim’s propensity for violence is “vital” to the defendant’s claim of self-defense. State v. Miranda, supra, 176 Conn. 110. Although other witnesses testified that the victim initiated the fight, this argument is unresponsive to the defendant’s claim that his right to use deadly force arose, not when the fight started, but when the victim apparently tried to retrieve a weapon from a garbage can during the fight. McNeil corroborated the defendant’s testimony that the victim deliberately reached into the garbage can and testified that the fight escalated at that point. Tchakirides was forty to sixty feet away when he noticed the fight and Badger was *796thirty feet away. None of the witnesses was able to testify as to what the victim did as he was reaching into the can. The proffered testimony provides a critical link between the victim’s physical condition at the time of the altercation and the likelihood that he would behave in a violent manner. I, therefore, believe that we cannot say that the evidence was unrelated to “evidence which reasonably could have affected the verdict . . . .” Costello v. Costello, 134 Conn. 536, 544, 59 A.2d 520 (1948).
The state argues that, because the defendant introduced the victim’s 1990 conviction of one count of threatening pursuant to General Statutes § 53a-62, the proffered evidence was cumulative.6 I disagree. Evidence of the victim’s drug and alcohol use at the precise time of the altercation is not sufficiently comparable to a four year old misdemeanor conviction to render the former cumulative. Unlike other forms of character evidence, such as prior criminal convictions, the proffered testimony relates specifically to the victim’s physical state at the time of the altercation and is therefore particularly probative of whether and to what extent the victim was likely to have engaged in violent behavior at the relevant time. See State v. Delgado, 8 Conn. App. 273, 283, 513 A.2d 701 (1986) (noting that character evidence relating to events close in time to altercation is “of a distinct and different order” than evidence relating to remote past).
Moreover, the case against the defendant was not overwhelming. The victim, who was larger than the *797defendant, initiated the fight, continued the altercation, and, as one state’s witness acknowledged, apparently deliberately reached into a garbage can during the fight. The state’s witnesses, Tchakarides and Badger, contradicted one another as to whether the defendant was “winning” the fight before the victim reached into the garbage can. Evidence that the victim was under the influence of illegal drugs at the time of the altercation and that this condition made it highly probable that he would behave violently, would likely have influenced the jury’s decision as to whether the defendant was correct in his assessment that the victim was attempting to retrieve a weapon. Had the jury heard the expert testimony in question and found that the victim was the aggressor and was violent during the altercation, it reasonably could have concluded that the defendant was justified in his use of deadly force and, therefore, was not guilty of murder.
Even if the jury had decided that the defendant properly acted in self-defense but that he had used excessive force, the jury, nevertheless, could have concluded that the defendant was guilty of first degree manslaughter, either based upon an intent to cause serious physical injury or based upon recklessness. See State v. Maselli, 182 Conn. 66, 73, 437 A.2d 836 (1980), cert. denied, 449 U.S. 1083, 101 S. Ct. 868, 66 L. Ed. 2d 807 (1981).7 The *798evidence was clearly sufficient to support a finding that the defendant intended to cause serious physical injury. Similarly, the defendant’s avowedly intentional stabbing of the victim constituted reckless conduct creating a grave risk of death to the victim under circumstances evincing an extreme indifference to human life, as proscribed by General Statutes § 53a-55 (a) (3). Because “[i]t is the facts as they reasonably appear to the defendant at the time he acted which measure the existence of the right of self-defense . . . [t]he jury might well have concluded that for the defendant to have believed under the circumstances revealed by the evidence that the victim was about to use deadly force upon him so that it was necessary to [repeatedly stab the victim] was such a gross deviation from the standard of conduct that a reasonable person would observe in the situation as to constitute recklessness. General Statutes § 53a-3 (13).” (Internal quotation marks omitted.) State v. Maselli, supra, 73 (firing eight shots at point blank range constituted recklessness). I would note that both the defendant and the state requested a jury instruction on the lesser included offenses of first degree intentional and reckless manslaughter, and that the trial court charged in accordance with these requests. The precluded evidence bore directly on the jury’s ability to consider these options in a meaningful way. I would conclude, therefore, that the trial court’s decision in this case caused the defendant substantial prejudice.
Finally, the state should not have been surprised and would not have been prejudiced by the introduction of the proffered testimony. This court has, in the past, noted the relevancy of evidence of the victim’s intoxication where the defendant has claimed self-defense. Additionally, the state had equal access to the victim’s *799medical records and could have cross-examined Merikangas or presented testimony refuting his conclusions. The trial court, therefore, abused its discretion when it excluded expert testimony that the level of drugs and alcohol present in the victim’s body would have made him more likely to be violent at the time of the altercation. See Unkelbach v. McNary, supra, 244 Conn. 367 (erroneous application of law constitutes abuse of discretion).
By way of summation, I would conclude that the trial court incorrectly applied the law in holding: (1) that the proffered evidence was inadmissible because the state had conceded that the victim initiated the fight; (2) that expert opinion testimony linking the victim’s drug and alcohol use at the time of the altercation was inadmissible; and (3) that the defendant had not laid a proper foundation for the proffered testimony. I, therefore, respectfully dissent from part III of the majority opinion.
Cyril Wecht, a certified anatomical, clinical and forensic pathologist, testified that meperidine is the generic name for Demerol, which was administered to the victim at the hospital. Wecht characterized the remainder of the drags as "substances of abuse that [the victim] had with him when he came to the hospital.”
In State v. Smith, supra, 222 Conn. 18, we held that “the deceased’s violent character may not be established by evidence of specific violent acts, other than convictions.”
In State v. Johnson, 139 Conn. 89, 91, 90 A.2d 905 (1952), we noted the presence of alcohol in the victim’s blood and the defendant’s testimony that the victim was quarrelsome when under the influence of alcohol. This decision antedates State v. Miranda, supra, 176 Conn. 114, wherein we first articulated our adoption of the rule that the victim’s violent character may be proved by reputation evidence regardless of whether the defendant was aware of that reputation. We, therefore, determined in Johnson that the victim’s reputation was irrelevant to the issue of the defendant’s guilt *790because the evidence did not indicate that the defendant was aware of that reputation. State v. Johnson, supra, 93.
I note that the defendant claimed affirmatively that a proper foundation had been laid for Merikangas’ testimony but did not set forth the arguments supporting that claim separately from those relating to the general question of the admissibility of expert opinion testimony linking the victim’s intoxication to his propensity for violence. Because these issues overlap to a substantial degree and because the appellant is not required to be redundant, I treat the argument as having been briefed adequately. See Potter v. Chicago Pneumatic Tool Co., 241 Conn. 199, 255 n.38, 694 A.2d 1319 (1997) (“argument, albeit brief, adequately [set] forth both the relevant factual basis and legal authority”). Although I recognize that it is the defendant’s burden to address the question of whether the trial court properly concluded that he had not laid a proper foundation for the proffered testimony, I note that the state has not defended the trial court’s exclusion of the evidence based upon an improper foundation.
I would conclude that the defendant was prejudiced substantially by the trial court ruling, regardless of how that is measured. It appears that the case law in Connecticut is inconsistent with respect to the proper standard of review to apply to erroneous evidentiary rulings. Although the cases uniformly hold that, where the error is not of constitutional magnitude, the appellant bears the burden of proving that the trial court’s ruling warrants reversal and a new trial, it is unclear what the appellant must show in order to meet that burden. Earlier cases required the defendant to show that the ruling was “both wrong and harmful.” See, e.g., DeCarufel v. Colonial Trust Co., supra, 143 Conn. 21. Several of those cases noted that “[i]f the ruling could not reasonably have affected the verdict, it is not harmful.” Id.; see, e.g., State v. Tropiano, 158 Conn. 412, 427, 262 A.2d 147 (1969), cert. denied, 398 U.S. 949, 90 S. Ct. 1866, 26 L. Ed. 2d 288 (1970); State v. Beaudet, 53 Conn. 536, 539, 4 A. 237 (1886). Subsequently, that language was recast in terms of an affirmative obligation on the part of the defendant to prove that the claimed erroneous action of the court would have been likely to affect the result. See, e.g., Higgins v. Karp, 243 Conn. 495, 506, 706 A.2d 1 (1998); State v. Esposito, 235 Conn. 802, 825, 670 A.2d 301 (1996); State v. Paulino, 223 Conn. 461, 478, 613 A.2d 720 (1992); State v. Vilalastra, 207 Conn. 35, 47, 540 A.2d 42 (1988); State v. McClain, 171 Conn. 293, 300, 370 A.2d 928 (1976). A number of cases addressing nonconstitutional evidentiary errors, however, have not stated that the defendant must prove that the result would have been different but have stated merely that the defendant must prove “ ‘substantial prejudice or injustice.’ ” State v. Beliveau, 237 Conn. 576, 592, 678 A.2d 924 (1996); see, e.g., State v. Small, 242 Conn. 93, 110, 700 A.2d 617 (1996); State v. Colton, 227 Conn. 231, 260, 630 A.2d 577 (1993), on appeal after remand, 234 Conn. 683, 663 A.2d 339 (1995), cert. denied, 516 U.S. 1140, 116 S. Ct. 972, 133 L. Ed. 2d 892 (1996); State v. Castonguay, 218 Conn. 486, 497, 590 A.2d 901 (1991). More recently, the standard has *795been refined to require that the defendant prove that the trial court’s ruling resulted in substantial prejudice and that it undermined confidence in the outcome of the trial. State v. Askew, supra, 245 Conn. 371-72.
Although it is not clear what the defendant must prove in order to satisfy each of these standards, this court has stated explicitly that the applicable standard of review is not equivalent to that by which we evaluate the sufficiency of the evidence. In Swenson v. Sawoska, 215 Conn. 148, 153, 575 A.2d 206 (1990), this court stated that denying reversal where other, properly admitted evidence was sufficient to sustain the verdict, a standard was “too restrictive in that it does not encompass situations where the erroneously admitted evidence, while not necessary itself to sustain the jury’s verdict, may nonetheless have affected the jury’s perception of the remaining evidence. Such a standard is inappropriate because it would require that we treat as harmless error any evidentiary ruling, regardless of its effect upon the verdict, so long as the evidence not implicated by the ruling was sufficient as a matter of law to sustain the verdict.” Because this court’s “review of evidential rulings, whether resulting in the admission or exclusion of evidence” is the same; Casalo v. Claro, 147 Conn. 625, 630, 165 A.2d 153 (1960); Swenson is equally applicable to the present case, where the trial court’s ruling resulted in the exclusion of evidence. This language shows quite clearly that it is not enough merely to show that other, properly admitted evidence, was sufficient to sustain the verdict. Swenson recognized that individual pieces of evidence do not exist in a vacuum and acknowledged that a piece of evidence, seemingly unimportant in its own right, may affect the fact finder’s interpretation of other highly significant evidence. Swenson v. Sawoska, supra, 153; cf. Costello v. Costello, 134 Conn. 536, 544, 59 A.2d 520 (1948) (where erroneous evidentiary rulings do not relate to evidence that “reasonably could have affected the verdict,” reversal not warranted).
Whether proffered evidence is cumulative is relevant to the issue of whether the prejudicial nature of that evidence substantially outweighs its probative value as well as to the question of whether the exclusion of the evidence will be held to be harmless. Because the trial court did not analyze the probative value or prejudice resulting from the admission of the evidence, however, I address the question of whether the evidence was cumulative only in the latter context.
The majority states that this case is comparable to State v. Munoz, supra, 233 Conn. 114-15 n.5, and State v. Person, 236 Conn. 342, 359, 673 A.2d 463 (1996) (Borden, J., concurring), in which it was noted (hat the presentation of inconsistent defenses might reasonably cause a jury to become skeptical about the defendant’s entire case. In Person, the trial court erroneously refused to charge the jury on the defense of extreme emotional disturbance where the defendant’s own testimony contradicted the claim that he was emotionally upset at the relevant time. State v. Person, supra, 356. In Munoz, the defendant claimed that he had stabbed the victim in self-defense, though not fatally, that he had an alibi for the time of death and that someone else had delivered the fatal blow. State v. Munoz, supra, 114-15. By contrast with those cases, the defenses in the present case are not necessarily inconsistent. The defendant is not claiming both that he stabbed the victim in self-defense and that someone else actually committed the crime. He is, *798instead, making two related but not inconsistent claims regarding his mental state at the time that he, admittedly, stabbed the victim.