(dissenting, with whom Spina, J., joins). I respectfully dissent from the court’s opinion because expert opinion based on blood alcohol level alone cannot give rise to triable issues of fact absent some evidence of the patron’s behavior at, or close in time to, the serving of the last drink. See Vickowski v. Polish Am. Citizens Club of Deerfield, Inc., 412 Mass. 606, 611-612 & n.5 (1996). Mindful that “[cjases of this character must be decided one by one, applying common law principles,” permitting this suit to proceed constitutes, in my opinion, a *169departure from precedent. Id. at 612, quoting McGuiggan v. New England Tel. & Tel. Co., 398 Mass. 152, 161 (1986). See Hopping v. Whirlaway, Inc., 37 Mass. App. Ct. 121, 126-127 (1994) (on remand, “the plaintiffs ought not to be permitted to put to their toxicologist witness hypothetical questions which ask him whether [the patron] would have been recognizably intoxicated ... if he had shown difficulty walking unless evidence is received that [the patron], in fact, had demonstrated difficulty in walking”); Kirby v. Le Disco, Inc., 34 Mass. App. Ct. 630, 632 (1993) (patron’s acknowledgment that he was “ ‘probably legally’ drunk,” without evidence of condition at tavern, not enough to defeat summary judgment). But see McGuiggan v. New England Tel. & Tel. Co., supra at 162 (doctor’s affidavit might be thought to raise a factual dispute as to what patron’s condition was just before leaving premises).
Until today, circumstantial evidence has been admitted only to bolster direct evidence of a patron’s appearance. Vickowski v. Polish Am. Citizens Club of Deerfield, Inc., supra at 612 n.5. See Gottlin v. Graves, 40 Mass. App. Ct. 155, 158-159 (1996) (evidence of patron’s intoxicated condition at accident scene, twenty minutes after he left tavern, admitted to corroborate direct evidence that patron appeared intoxicated at some point prior to serving of last drink). This well-grounded rule is an effective safeguard against unwarranted speculation. Vickowski v. Polish Am. Citizens Club of Deerfield, Inc., supra at 610 (evidence from accident scene, “unsupported by additional probative evidence . . . bearing on [patron’s] demeanor” at premises, would invite “unacceptable speculation on the part of a jury”).
First, proof derived from blood or urine tests that a patron was intoxicated at a specific time is not particularly helpful in cases such as this where the tavern’s duty is triggered only on the patron’s appearance. See Fandozzi v. Kelly Hotel, Inc., 711 A.2d 524, 528 (Pa. Super. Ct. 1998) (visibly intoxicated standard “mandates that tavern owners be governed by physical appearances rather than medical diagnoses”). Second, while the objectivity of scientific proof often assists the trier of fact, it could create confusion where jurors are called on to evaluate an inherently subjective issue such as whether someone appeared intoxicated. As the majority points out, ante at 166, the effects of alcohol consumption differ greatly from person to person. See E.F. Fitzgerald, Intoxication Test Evidence § 22:4, at *17022-10 n.2 (2d ed. 2000) (“Opining after the fact that obvious symptoms of intoxication ‘would’ have been present in a given individual, to a reasonable scientific certainty, is indeed a perilous course given the variety of observable symptoms whenever large groups of people are studied”). This fact takes on even greater significance in cases such as this, where liability hinges on an individual’s appearance.
Here, the “bolstering” rule would simply require the plaintiff to produce an evidentiary “link” between the expert’s testimony and the patron’s demeanor at, or close in time to, his leaving the tavern. Whereas the majority views the expert’s opinion, itself, as the “necessary fink between an elevated blood alcohol level some two hours after Gagne’s last drink and signs of intoxication at the time he was served that last drink,” ante at 166, I am concerned that this “link” does not shed any more light on the only factually (and legally) relevant issue, namely, how the patron actually looked. In short, the expert’s testimony constitutes additional evidence, but does not rise to the level of bolstering evidence given the absence of any “link” to the tavern.
Moreover, our “reluctance” to rely on speculative evidence as a surrogate for direct evidence of “obvious intoxication” should be particularly acute here, where the record is replete with extensive direct evidence of the patron’s behavior at the tavern, all of which indicated he appeared sober. Vickowski v. Polish Am. Citizens Club of Deerfield, Inc., supra at 612. For these reasons I do not believe that expert testimony regarding how a particular individual “would have looked” should, without some shred of bolstering evidence related to his appearance or behavior at the time of, or close in time to, his last drink, constitutes an acceptable source of triable issues of fact. Because the factual issues disputed here (i.e., patron’s appearance and quantity of drinks) all emerge from the expert’s opinion, I would affirm the judge’s grant of summary judgment.1
Regardless of whether the argument was waived, the expert’s opinion that the patron consumed at least nine drinks similarly does not suffice as a triable issue of fact. See ante at 168 n.3. Again, where a disputed statement emerges from an unbolstered expert opinion it does not, in my opinion, warrant a trial.