OPINION OF THE COURT

The question presented by this appeal is whether plaintiffs reckless behavior was of such a nature as to constitute the sole legal cause of his injuries, vitiating the duty of care of a train operator. We conclude under the circumstances of this case that it was not, and that the evidence was sufficient to support the verdict. We further conclude that plaintiffs estimate of his own running speed at the time of the accident was admissible and sufficient to lay a proper foundation for plaintiff’s accident reconstruction expert to use in forming his opinion.

Plaintiff—then 18 years old—and three of his friends spent the evening of January 24, 1997 in Manhattan and, over the course of about five hours, each consumed a quantity of alcohol. In the early hours of January 25, plaintiff and his friends attempted to return home to Queens. They waited some time for the number 7 train at Queensboro Plaza—an elevated subway station—where red tape indicating track work extended across the platform. The group eventually decided that the train was not running and walked along the “catwalk”—a three-foot wide path abutting the track—to the next station (33rd Street). They made it safely to 33rd Street and, after again waiting for a while, *490decided to walk along the catwalk to the next station (40th Street). As they proceeded along the northbound route, the tracks were to the left of the catwalk and to the right was a short railing.

Before they reached the 40th Street station, the group became aware of a train approaching from behind and began running along the catwalk in an attempt to reach the station before the train. Their intent was to board the train to continue their trip home. The four teens ran in single file, with plaintiff third in line. The train struck plaintiff in the vicinity of a signal box located on the catwalk, resulting in bilateral below-the-knee amputation.1 Plaintiff was the only one of the four hit by the train.

At trial, plaintiff was permitted to testify to an estimate of his running speed at the time of the accident.2 Plaintiff established that he was familiar with his running speed because he had been a member of a gym for about two years prior to the accident and would frequently run on a treadmill calibrated in terms of miles per hour. He would usually run at seven or eight miles per hour and, based on that experience, estimated that he was running about seven or eight miles per hour at the time of the accident.

Plaintiffs expert then used that estimate in making his calculations. Computing the train’s stopping distance assuming the train operator perceived the boys on the catwalk from 151.5 feet away—the distance allegedly illuminated by the train’s headlights—and factoring in reaction time, the expert determined that the train could have stopped 51 feet before it reached plaintiff if he had been running eight miles per hour and 37 feet before it reached plaintiff if he had been running seven miles per hour.

*491The train operator offered several inconsistent versions of his conduct at the time of the accident. The police report taken at the scene indicates that the train operator stated he did not see the boys “until he was passing them because they were hidden from his view by a signal device” and that the train stopped as a result of its own emergency devices before he was able to apply the brake. However, later that night, the train operator also filled out a report for NYCTA, this time stating that he saw an object behind the signal device and then placed the train into emergency mode. At trial he gave a third account, testifying that his first reports were inaccurate in that he saw a person on the catwalk in front of the signal box from about a train car’s length away. He also testified that he placed the train into emergency stopping mode when he saw the person. The road car inspector for NYCTA who examined the train at the scene, however, only found blood on the train’s “trip cock.” The trip cock hangs in front of the train’s wheels as an emergency device that automatically stops the train if it comes into contact with an obstruction.

At the close of plaintiffs case, NYCTA moved to dismiss the complaint for plaintiffs failure to establish a prima facie case of negligence. NYCTA also moved for a directed verdict on the issue of liability at the close of proof. The trial judge reserved decision on both motions. The jury returned a verdict finding NYCTA negligent, determining that its negligence was a substantial factor in causing the accident and assigning it 25% of the fault.3 The jury allocated the remaining 75% of liability to plaintiff. NYCTA moved to set aside the verdict and Supreme Court denied all motions.

A majority of the Appellate Division affirmed, finding legally sufficient evidence to support the jury verdict as against NYCTA. Two Justices dissented in part and voted to reverse. They would have dismissed the complaint for failure to establish a prima facie case because they believed “plaintiffs own reckless conduct constituted an intervening and unforeseeable act which broke any causal connection between his injury and any alleged negligence” by NYCTA (19 AD3d at 581-582). The dissent also stated that plaintiff should not have been permitted to testify as to an estimate of his running speed, nor should the *492accident reconstruction expert have been allowed to use plaintiffs testimony as a basis for his opinion, since no foundation was properly laid and thus the testimony was “pure speculation and conjecture” (19 AD3d at 582). NYCTA appeals as of right, and we affirm.

A jury verdict rests on legally insufficient evidence where there is “simply no valid line of reasoning and permissible inferences which could possibly lead rational [individuals] to the conclusion reached by the jury on the basis of the evidence presented at trial” (Cohen v Hallmark Cards, 45 NY2d 493, 499 [1978]). We have held that this inquiry is similar to that of a trial judge determining whether to direct a verdict (see Cohen, 45 NY2d at 499). If there is a question of fact and “it would not be utterly irrational for a jury to reach the result it has determined upon . . . the court may not conclude that the verdict is as a matter of law not supported by the evidence” (Cohen, 45 NY2d at 499; Campbell v City of Elmira, 84 NY2d 505, 510 [1994]).

NYCTA argues that plaintiffs own reckless conduct—entering the catwalk after consuming several alcoholic beverages and attempting to catch the train—was the sole proximate cause of his injuries. Significantly, NYCTA did not request any instruction that the jury consider whether plaintiffs conduct was the sole proximate cause of his injuries. Neither did defendant request a jury charge on superseding or intervening cause. Rather, the jury was instructed that plaintiffs “negligence was a substantial factor in causing the accident,” and was asked to determine further whether NYCTA was also negligent, whether that negligence was a contributing factor to plaintiffs accident and the percentage of fault attributable to each party.

Plaintiffs conduct was undeniably reckless, but the jury appropriately considered plaintiffs actions and determined that he bore a far greater share of the fault. This is in keeping with the doctrine of comparative negligence {see CPLR 1411). Contrary to NYCTA’s argument and the dissent, plaintiffs conduct, although a substantial factor in causing the accident, was not so egregious or unforeseeable that it must be deemed a superseding cause of the accident absolving defendant of liability.

This case does not involve the type of dangerous, illegal conduct presented in Barker v Kallash (63 NY2d 19 [1984]) where the plaintiff was severely injured while constructing a pipe bomb. There, the Court denied plaintiff recovery for *493injuries sustained as a direct result of a serious violation of law, “not because the plaintiff contributed to his injury, but because the public policy of this State generally denies judicial relief to those injured in the course of committing a serious criminal act” (Barker, 63 NY2d at 24). Here, plaintiff surely and very substantially contributed to his injury, but did not engage in the type of dangerous criminal conduct that prohibited recovery in Barker. Nor can it be said that his conduct was such as to vitiate the duty of reasonable care owed to him by the train operator to see what there was to be seen through the proper use of his senses and to bring the train to a safe stop.

We have held that a train operator may be found negligent if he or she sees a person on the tracks “from such a distance and under such other circumstances as to permit him [or her], in the exercise of reasonable care, to stop before striking the person” (Coleman v New York City Tr. Auth., 37 NY2d 137, 140 [1975]; see Noseworthy v City of New York, 298 NY 76, 79 [1948]). The train operator’s duty certainly is not vitiated because plaintiff was voluntarily walking or running along the tracks or because of any reckless conduct on plaintiffs part.

Thus, it was not irrational for the jury to find NYCTA negligent. There is a reasonable view of the evidence that the train operator failed to see the teenagers from a distance from which he should have seen them, and that he failed to employ emergency braking measures. The jury’s determination that the operator could have avoided this accident is an affirmed finding of fact with support in the record and is beyond our further review. Plaintiffs conduct did not constitute such an unforeseeable or superseding event as to break the causal connection between his injury and defendant’s negligence.

The trial court also properly permitted plaintiff to testify regarding an estimate of his running speed. Plaintiff established a sufficient foundation demonstrating the basis of his knowledge—two years’ experience running on a treadmill calibrated to measure miles per hour. In comparable situations, both police and civilian witnesses with an appropriate basis for knowledge have been permitted to give testimony estimating the speed of moving motor vehicles (see People v Olsen, 22 NY2d 230, 231-232 [1968]; Senecal v Drollette, 304 NY 446, 448 [1952]). The reliability of plaintiffs testimony and the weight it should have been accorded were issues for the finders of fact. The jury, which was best able to observe plaintiffs testimony and evaluate his credibility, resolved the question of reliability in plaintiffs favor. *494It cannot be said that the jury’s conclusion was utterly irrational.

Additionally, the jury was entitled to credit the testimony of plaintiffs expert who used the estimated running speed in making his calculations. The expert did not express an opinion as to how fast plaintiff was running, but used plaintiffs own estimate to determine where the train could have come to rest if plaintiff was running at the speeds he asserted. As a result, it was not “pure speculation and conjecture,” but admissible and reliable evidence from which the jury properly concluded that the train could have stopped before striking plaintiff.

Accordingly, the order of the Appellate Division should be affirmed, with costs.

Since the enactment of CPLR 1411 in 1975, it has been the general rule that plaintiffs own culpability will not bar his claim, but will only be grounds for apportioning fault. In several cases, however, we have recognized that a plaintiffs fault may be so egregious in comparison to the defendant’s that it “supersede^] defendants’ conduct” (Egan v A.J. Constr. Corp., 94 NY2d 839, 841 [1999]) and becomes the “sole legal cause” (Olsen v Town of Richfield, 81 NY2d 1024, 1026 [1993]) or “sole proximate cause” (Howard v Poseidon Pools, 72 NY2d 972, 975 [1988]) of the plaintiff’s injuries. I believe the facts of this case bring it within the superseding cause doctrine.

Plaintiff here was reckless twice. It was, of course, very foolish for him and his companions to choose a narrow catwalk next to the subway tracks for pedestrian travel. But plaintiff would have escaped unharmed if he and the others, when they heard a train coming, had done what minimal common sense would require—stand still, as far as they could get from the tracks, and let the train pass by. Instead, they chose to race the train to the next station. On these facts, I think it is fair to say that plaintiffs injuries were entirely his own fault, even if a nonnegligent motorman might have been able to stop the train in time to avoid the accident.

While no two cases are identical, precedents support the view that this plaintiffs recklessness was serious enough to be a “superseding cause.” In Merino v New York City Tr. Auth. (89 NY2d 824, 825 [1996]) we refused recovery to a plaintiff who had fallen onto the subway tracks as a result of his own intoxication, holding, as an alternative ground for dismissal, that *495plaintiff “failed to show that the defendant’s alleged negligence was a substantial factor in causing his injury.” Several Appellate Division cases also deny recovery on superseding cause grounds to plaintiffs who put themselves in the paths of trains (e.g., Wadhwa v Long Is. R.R., 13 AD3d 615 [2d Dept 2004]; Lassalle v New York City Tr. Auth., 11 AD3d 661 [2d Dept 2004]; Snyder v New York City Tr. Auth., 2 AD3d 162 [1st Dept 2003], lv denied 2 NY3d 707 [2004]; Gao Yi Feng v Metropolitan Transp. Auth., 285 AD2d 447 [2d Dept 2001]). Gao Yi Feng is perhaps the closest to this case on its facts: the plaintiff there was struck while standing “directly adjacent to the train tracks with his back to oncoming trains” (285 AD2d at 447).

Cases denying recovery, even under a comparative negligence regime, to highly reckless plaintiffs speak of causation in defending their results, but in each case it is clear that the extraordinary degree of the plaintiffs fault—not just the foreseeability of the plaintiffs conduct—has been decisive. A pure foreseeability analysis does not explain the cases. For example, I would not hesitate to uphold an award if this plaintiff had been pushed by someone else into the path of the train, rather than recklessly placing himself there, though there is not much difference in the foreseeability of the two events.

The principle we have applied in previous cases, and should apply here, is that people whose failure to take care of themselves is extreme may not shift any of the consequences to others. Anyone of normal human compassion will sympathize with plaintiff; he is not the only 18 year old who ever acted recklessly, and he has paid a much higher price for it than most. But I do not think it consistent with law or wise policy to hold, as the majority does, that the New York City Transit Authority must compensate him in part for his loss. The Transit Authority moved, before the case was submitted to the jury, for a directed verdict, asserting that plaintiffs reckless conduct was the sole legal cause of the accident. I think that motion should have been granted.

Chief Judge Kaye and Judges G.B. Smith and Read concur with Judge Ciparick; Judge R.S. Smith dissents and votes to reverse in a separate opinion in which Judges Rosenblatt and Graffeo concur.

Order affirmed, with costs.