On December 5,1978, David E. Colter suffered severe arm injuries while greasing the gears of a twin screw sand classifier owned and operated by Colter’s employer, Marshfield Sand & Gravel Company (Marshfield).2 The machine was manufactured by a division of Barber-Greene Company (Barber-Greene) which sold it to Worcester Sand & Gravel Company (Worcester). Worcester traded the machine to New England Road Machinery Company (New England), which later sold it to Marshfield. Colter sued Barber-Greene, Worcester, and New England asserting causes of action for negligence, breach of implied warranty of merchantability, and breach of implied warranty of fitness for a particular use.3 The jury found Barber-Greene to have been 36% negligent, found New England to have been 15% negligent, found Worcester not to have been negligent,4 and found Colter to have been 49% contributorily negligent. On the warranty count, the jury found that Barber-Greene and New England each had breached its implied warranty of merchantability and had proximately caused Colter’s injuries. However, the jury also found that Colter’s recovery was barred because he had actual knowledge of the machine’s defective condition, but nonetheless proceeded unreasonably to use the machine.

Barber-Greene and New England moved for judgment notwithstanding the verdict or for a new trial, arguing that the finding on the warranty count was inconsistent with and negated *52the finding that Barber-Greene’s and New England’s negligence proximately caused Colter’s injuries.5 The trial judge ruled that New England and Barber-Greene were entitled to judgment in their favor. Colter subsequently filed a motion for relief from judgment on the basis of Richard v. American Mfg. Co., 21 Mass. App. Ct. 967 (1986).6 The judge allowed the motion and entered judgment for Colter. New England and Barber-Greene appealed. This court granted Barber-Greene’s application for direct appellate review. We conclude that there is sufficient evidence on the issue whether the negligent design of the sand classifier proximately caused Colter’s injuries to submit that claim to the jurors. We agree with Barber-Greene that the evidence on negligent failure to warn is insufficient. Because the jurors were not asked to apportion Colter’s negligence between the two theories of negligence, there must be a new trial on negligent design.

At the time of his accident, Colter was employed as a manager at Marshfield’s quarry in Weymouth. Marshfield operated quarrying equipment at this location including the twin screw sand classifier involved in Colter’s accident. The twin screw classifier consisted of a rectangular hopper containing two screw augers. The screws were driven by a set of bevel gears which ran at a speed of approximately twenty to thirty revolutions per minute. The gears, which were powered by an electric motor, were mounted on a steel frame approximately twenty feet above ground. A conveyor belt fed wet sand into the hopper. The screws carried the sand upward to another con*53veyor belt. The action of the screws removed the water from the sand; the water poured out the low end of the machine.

Barber-Greene’s specifications called for a guard to cover the bevel gears, and they did not sell the machine without one. When Worcester purchased the machine involved in Colter’s accident in 1952, the purchase included a bevel gear guard. By the time New England took the twin screw sand classifier from Worcester in trade, there was no guard on the machine and the owner’s manual which showed the guard in place was missing as well. New England sold the machine to Marshfield in 1969 without the guard or the owner’s manual. New England did not inform Marshfield that the machine’s bevel gears should be covered. Although Barber-Greene manufactured replacement guards, Marshfield never obtained a guard for the gears.

Colter began his employment with Marshfield in 1971 as a concrete salesman assigned to Marshfield’s Cohasset office.7 In 1975, Colter was promoted to the position of safety director at the Weymouth plant. In the summer of 1977, Marshfield moved the sand classifier to Weymouth and mounted it on the twenty-foot high steel structure. Although Colter was generally aware of the dangers of exposed gears, he was not aware that the machine required a guard. Personnel from the United States Bureau of Mine Safety and Health Administration frequently inspected the Weymouth facility, but never instructed Colter to obtain a bevel gear guard for the sand classifier because the gear box was mounted above ground.

On the day of the accident, Colter arrived at work sometime between 10:30 and 11 a.m. The plant’s operation had started late that day because the cold weather had frozen some pipes. Shortly after lunch, Colter passed the “wet end of the plant” and heard a loud screaming “steel-on-steel” noise emanating from the gears of the sand classifier. Colter had never heard the machine make this kind of noise and, because the noise was so loud, he feared that the machine would come apart. Colter knew that the company’s operations at that time were *54critical, and that, if the sand classifier broke, it would put Marshfield’s client out of business. Responding to what he believed to be an emergency situation,8 Colter instructed another employee to get him a grease gun. After obtaining the grease gun, Colter drove a front-end loader to the steel structure supporting the sand classifier. Colter climbed onto the structure, and stood in the middle of the conveyor belt on two angle irons. He injected grease into the gears on one side and the noise stopped. Before Colter descended from the machine, his jacket caught in the gears, pulling in his arms. Colter suffered severe injuries requiring the amputation of his right arm below the elbow, amputation of his left index finger and causing substantial loss of function in his left arm.

1. Sufficiency of the evidence. Barber-Greene contends that the judge denied erroneously its motion for a directed verdict on the negligence counts because there was insufficient evidence that its conduct proximately caused the plaintiff’s injuries. In determining whether the jury were warranted in finding Barber-Greene negligent, “[t]he question is whether the evidence, construed most favorably to the plaintiff, could not support a verdict for the plaintiff.” Poirier v. Plymouth, 374 Mass. 206, 212 (1978). The plaintiff is entitled to judgment if “anywhere in the evidence, from whatever source derived, any combination of circumstances could be found from which a reasonable inference could be drawn in favor of the plaintiff.” Raunela v. Hertz Corp., 361 Mass. 341, 343 (1972), quoting Kelly v. Railway Express Agency, Inc., 315 Mass. 301, 302 (1943). To withstand a motion for a directed verdict on the issue of proximate causation, the plaintiff need only demon*55strate that there was a greater likelihood that the harm of which the plaintiff complains was due to causes for which the defendant was responsible. Mullins v. Pine Manor College, 389 Mass. 47, 58 (1983). The plaintiff need not eliminate all possibility that the defendant’s conduct was not a cause, but need only introduce evidence from which reasonable jurors could conclude that it was more probable than not that the injuries were caused by the defendant’s conduct. Id. We have reviewed the evidence in the light most favorable to the plaintiff and we conclude there is no error.

On the issue of negligent design, Colter alleged that the design of the twin screw classifier was defective in two major respects. First, Colter argued that the sand classifier was defectively designed because it used outdated power transmission elements which required manual greasing rather than using fully enclosed elements which rotated in their own lubricating bath. The latter design eliminated the need to guard against injury from greasing the gears because the gears were self-lubricating. Given that the machine involved in Colter’s accident used exposed power transmission elements, Colter argued that the sand classifier was deficient in a second respect, namely that the guard that covered the exposed gears was improperly designed. Colter argued that the gear guard was defective because it contained large openings that would allow sand and dirt to enter the gears thus ensuring that the guard would need frequent removal for cleaning, greasing, and other maintenance. Colter also argued that the gear guard was too cumbersome to be removed and replaced easily and that the guard, by its own improper design, invited permanent removal.

Colter presented the expert testimony of Egor Paul, a professor of mechanical engineering, who stated that the sand classifier itself was defectively designed because the power transmission elements were exposed to the dirty, sandy environment of the gravel yard and thus required frequent cleaning and greasing. Paul stated that designs which fully enclosed the power transmission elements in a self-lubricating bath had been available since the 1930’s, and that these designs eliminated the need for greasing the transmission elements.

*56Paul’s testimony was corroborated by Amelio Salera, the president of New England, who stated that his company did not manufacture and sell open-gear sand classifiers because the company considered the closed-gear model to be more safe. Salera stated that he considered the enclosed gear design to be safer because it reduced the number of times any worker risked contact with exposed gears. James Mueller, manager of marketing services for Barber-Greene, stated that Barber-Greene manufactured sand classifiers with enclosed gearboxes in the 1950’s, and that one of the reasons for the enclosed design related to safety concerns.

As regards the gear guard itself, Colter contended that the guard was defective because the gear guard did not completely enclose the gears, but left an opening at the bottom and two additional cutouts in the guard. Professor Paul testified that because the machine was intended for use in a sandy, dusty environment, the openings guaranteed that the gears would need frequent greasing. James Mueller stated that Barber-Greene expected that the sand classifier’s gear guard could need to be removed twice a month so that the machine could be washed, greased, and serviced. Although Worcester’s representative stated that he did not know what happened to the gear guard, on cross-examination Barber-Greene’s representative recognized that the sand classifier’s gears might need washing as many as 624 times between the machine’s manufacture in 1952 and Colter’s accident in 1978. Worcester’s representative, Matteo Trotto, testified that the gears never were washed while Worcester owned the machine, but that they were greased and repaired. Trotto stated that he did not know whether the guard was removed for the greasing, but that the repairs possibly would have required the guard’s removal.

Because the guard would need to be removed frequently, Colter argued that the guard was negligently designed because it was not easily removed and replaced. The gear guard designed by Barber-Greene weighed approximately sixty pounds and was attached to the sand classifier by five bolts. Two workers were required to remove and replace the guard. The gears were mounted on the high end of the machine, and Barber-*57Greene acknowledged that the gears could be mounted as high as forty feet off the ground. Professor Paul stated that a proper design would have included a hinged opening or access panels which would have allowed a worker to grease the gears without removing the guard. The plaintiff’s expert testified that design considerations appropriately account for the frequency of removal. Barber-Greene never tested the guard to determine the ease of removal.

We hold a manufacturer liable for defectively designed products because the manufacturer is in the best position to recognize and eliminate the design defects. See Solimene v. B. Grauel & Co., KG, 399 Mass. 790, 796 (1987). “In evaluating the adequacy of a product’s design, the jury should consider, among other factors, ‘the gravity of the danger posed by the challenged design, the likelihood that such danger would occur, the mechanical feasibility of a safer alternative design, the financial cost of an improved design, and the adverse consequences to the product and to the consumer that would result from an alternative design. ’ ” Back v. Wickes Corp., 375 Mass. 633, 642 (1978), quoting Barker v. Lull Eng’r Co., 20 Cal. 3d 413, 431 (1978). “[Tjhere is a case for the jury if the plaintiff can show an available design modification which would reduce the risk without undue cost or interference with the performance of the machinery.” Uloth v. City Tank Corp., 376 Mass. 874, 881 (1978). As the Appeals Court has noted, it is the jury ’ s function to determine “whether the circumstances of the guard’s removal and the plaintiff’s subsequent injury were reasonably foreseeable.” Fahey v. Rockwell Graphics Syss., Inc., 20 Mass. App. Ct. 642, 648 (1985).9

*58Based on the evidence, the jury were entitled to find that Barber-Greene had negligently designed the twin screw sand classifier. The jury could have found that the open gear design was defective because it allowed sand and dirt to enter the machinery, thus necessitating frequent cleanings and greasings. The jury could have found that the enclosed gear design provided a safer alternative to the open-gear sand classifier. The evidence was undisputed that Barber-Greene manufactured an enclosed gear sand classifier in 1952, and Barber-Greene’s representative could not say which design was more expensive.

The jury also could have found that the open guard was defective as well. Moreover, the jury could have found that, because the guard lacked any access panels or doors through which an individual could clean or grease the gears, the guard had to be removed each time the bevel gears required attention. Because the gear guard was so cumbersome, the jury could have concluded that it was entirely foreseeable that a purchaser of the machine would remove the guard permanently. In the absence of any testimony as to what actually happened to the gear guard while the sand classifier was in Worcester’s possession , the jury were entitled to infer10 that the guard was removed *59for greasing and repair and was not replaced because it was too cumbersome, not efficient, and too costly in the labor required to remove and replace the gear guard.11

On the issue of its negligent failure to warn of the hazards of cleaning the bevel gears without shutting down the machine’s operation, Barber-Greene contends that Colter’s admitted knowledge of the dangers inherent in greasing the machine without the guard relieves the company of liability because Barber-Greene’s failure to warn was not the proximate cause of Colter’s injuries. As this court has recognized, where the danger presented by a given product is obvious, no duty to warn may be required because a warning will not reduce the likelihood of injury. Uloth, supra at 880. See also Fiorentino v. A.E. Staley Mfg. Co., 11 Mass. App. Ct. 428, 436 (1981). Colter admitted he knew it was dangerous to grease the gears while operating the machine. In these circumstances, a warning would not reduce the likelihood of injury. We therefore agree with Barber-Greene that it was entitled to a judgment on this theory as a matter of law.

*602. Proximate cause. On the warranty count, the jury found that Barber-Greene had breached its warranty of merchantability with respect to the design, manufacture, and sale of the sand classifier and that New England had breached its warranty of merchantability with respect to the sale of the machine. The jury found that the defendants’ breaches had proximately caused Colter’s injuries, but found that Colter had actual knowledge of the defective condition of the sand classifier, and that he had proceeded unreasonably to use the machine, thus injuring himself. This finding precluded Colter’s recovery on the warranty count. As explicated in this court’s decision in Correia v. Firestone Tire & Rubber Co., 388 Mass. 342 (1983), a plaintiff’s knowing and unreasonable use of a defective product is an affirmative defense to a defendant’s breach of warranty. See Allen v. Chance Mfg. Co., 398 Mass. 32, 34 (1986).12 Barber-Greene and New England contend that the jury’s affirmative answer to a question on unreasonable use on the warranty count also relieves them from negligence liability because, for both counts, Colter’s own conduct was the proximate cause of his injuries. We do not agree.

Barber-Greene and New England principally rely on language in Correia to support their contention that Colter’s unreasonable use of the sand classifier is, for purposes of negligence, a bar to recovery. This court stated, “[T]he user’s negligence does not prevent recovery except when he unreasonably uses a product that he knows to be defective and dangerous. In such circumstances, the user’s conduct alone is the proximate cause of his injuries, as a matter of law, and recovery is appropriately denied. In short, the user is denied recovery, not because of his contributory negligence or his assumption of the risk but rather because his conduct is the proximate cause of his injuries.” Correia, supra at 356.

Although there is a certain logic to the defendants’ argument that a finding of unreasonable use in a warranty count negates a finding of proximate cause on a negligence count, the argument is flawed because it equates proximate cause with sole *61cause. An examination of the principles underlying liability in negligence and liability in warranty indicates that, while sole proximate cause is a component of the warranty inquiry, negligence liability does not focus on a sole cause of the plaintiff’s injuries.

It is a well-settled proposition that actions for negligence and for breach of warranty impose distinct duties and standards of care. The basic elements of a products liability action founded on negligence are duty, breach of duty, cause in fact, and proximate cause. Epstein, Products Liability: Defenses Based on Plaintiff’s Conduct, 1968 Utah L. Rev. 267, 268. The focus of the negligence inquiry is on the conduct of the defendant. We impose liability when a product’s manufacturer or seller has failed to use reasonable care to eliminate foreseeable dangers which subject a user to an unreasonable risk of injury. Correia, supra at 354. Smith v. Ariens Co., 375 Mass. 620, 624 (1978). Uloth, supra at 878. See Prosser & Keeton, Torts 683 (5th ed. 1984); Twerski, From Defect to Cause to Comparative Fault — Rethinking Some Product Liability Concepts, 60 Marq. L. Rev. 297, 298 (1977). “[A] finding of negligence [is] a statement by the jury about the product and about the manufacturer as well. It signified] that the product was unreasonably dangerous because of its design or because of its failure to be accompanied by an adequate warning, or both. It also signified] that an ordinarily prudent manufacturer would have recognized the product’s shortcomings and would have taken appropriate corrective measures.” Hayes v. Ariens Co., 391 Mass. 407, 410 (1984).

Liability for breach of warranty stands on a much different footing. In Massachusetts, liability for breach of warranty of merchantability is governed by G. L. c. 106, §§ 2-314 — 2-318 (1986 ed.), and this court has noted that these provisions are congruent, in all material respects, with the principles expressed in Restatement (Second) of Torts § 402A (1965), the Restatement’s definition of a seller’s strict liability for harm suffered by a user or consumer of a seller’s product. Correia, supra at 353. Hayes, supra at 412. Back v. Wickes Corp., supra at 640. Unlike negligence liability, warranty liability “focuses on *62whether the product was defective and unreasonably dangerous and not on the conduct of the user or the seller.” Correia, supra at 355. Hayes, supra at 413. Because a breach of warranty does not require a defendant’s misconduct, 13 a defendant may be liable on a theory of breach of warranty of merchantability even though he or she properly designed, manufactured, and sold his or her product. See Restatement (Second) of Torts § 402A comment a (1965); Correia, supra at 353; Hayes, supra at 413. Clearly, liability based on a theory of strict liability differs significantly from liability based on negligence. “A defendant in a products liability case in this Commonwealth may be found to have breached its warranty of merchantability without having been negligent, but the reverse is not true. A defendant cannot be found to have been negligent without having breached the warranty of merchantability.” Hayes, supra at 410.

In a negligence case, the conduct of the plaintiff which will serve to bar recovery is governed by statute. Our comparative negligence statute provides that the plaintiff’s contributory negligence shall not bar recovery if the plaintiff’s negligence was not greater than the total amount of negligence attributable to the parties against whom recovery is sought. The plaintiff’s negligence, if less than the amount attributable to the defendant or defendants, only serves to diminish recovery by the proportion of negligence attributable to the plaintiff. G. L. c. 231, § 85 (1986 ed.). Thus, in a negligence action, the trier of fact must focus on the conduct of both the defendant and the plaintiff *63in determining the extent of each party’s responsibility for the plaintiff’s injuries. The plaintiff’s conduct is not viewed as the sole proximate cause of the injury and does not bar recovery completely unless the plaintiff is more than fifty per cent responsible for his or her own injuries.

Defined most simply, contributory negligence in products liability cases consists of the plaintiff’s failure to discover the product’s defect or to guard against the possibility that such a danger exists. Noel, Defective Products: Abnormal Use, Contributory Negligence, and Assumption of Risk, 25 Vand. L. Rev. 93, 106 (1972); Epstein, supra at 270. Essentially, we require the plaintiff to act reasonably with respect to the product he or she is using. The plaintiff’s contributory negligence is measured objectively.

In warranty, as in negligence, a plaintiff’s conduct may bar recovery from a liable defendant. “The absolute bar to the user for breach of his duty balances the strict liability placed on the seller. Other than this instance, the parties are not presumed to be equally responsible for injuries caused by defective products, and the principles of contributory or comparative negligence have no part in the strict liability scheme.” Correia, supra at 355-356. Because warranty liability focuses on whether the product was defective and unreasonably dangerous and not on the conduct of the user or the seller, “the only duty imposed on the user is to act reasonably with respect to a product which he knows to be defective and dangerous. When a user unreasonably proceeds to use a product which he knows to be defective and dangerous, he violates that duty and relinquishes the protection of the law. It is only then that it is appropriate to account for his conduct in determining liability. Since he has voluntarily relinquished the law’s protection, it is further appropriate that he is barred from recovery.” Id. at 355. The plaintiff’s conduct implies consent to the risk and thus is viewed as the sole proximate cause of the injury. Noel, supra at 129.

Applying the unreasonable use doctrine to actions sounding in negligence is foreclosed by the Commonwealth’s comparative negligence statute. That statute provides that a plaintiff’s *64recovery shall not be barred by his or her contributory negligence unless the plaintiff’s negligence is greater than the amount of negligence attributable to the parties against whom recovery is sought. G. L. c. 231, § 85 (1986 ed.). The statute clearly defines the amount of contributory negligence that will bar the plaintiff’s recovery. The defendants’ suggestion that proof of unreasonable use in warranty should bar completely the plaintiff’s recovery in negligence would, in effect, ordain that the plaintiff’s unreasonable use of a product is, as a matter of law, negligence greater than that of the defendants and the sole proximate cause of the injury. That determination is, however, for the finder of fact to make on a case-by-case basis and is not properly decided by the courts as a rule of law. In a negligence action, it is the function of the triers of fact to determine the percentage of fault attributable to the plaintiff, and the triers of fact are entitled to make that determination based on their perception of the relative fault of the parties. A jury may “find that the plaintiff was barred because of his unreasonable action. That such unreasonable conduct on the part of the plaintiff may in this case also have been found by the jury to be contributory negligence . . . does not eradicate the distinction in the defenses to the two counts. Under the negligence count, the plaintiff is not barred unless his negligence is greater than the negligence of the persons against whom recovery is sought.” (Footnote omitted.) Richard v. American Mfg. Co., 21 Mass. App. Ct. 967, 968 (1986). Accord Briney v. Sears, Roebuck & Co., 782 F.2d 585, 589 (6th Cir. 1986) (applying Ohio law and holding that, while plaintiff’s recovery under a warranty theory was barred by his unreasonable use of the product, plaintiff was entitled to trial on his negligence claim under Ohio’s comparative negligence statute).14

*65We conclude that the jury’s findings that Colter unreasonably proceeded to use the twin screw sand classifier after becoming aware of its defective condition does not bar recovery on his negligence claim as a matter of law. Nevertheless, we believe that application of some type of apportionment principles to warranty cases may be fairer than the current system, and may make results in negligence and warranty counts in the same case more consistent with each other. “[G]iven the wide variety of possible solutions,” see Correia, supra at 356, and the serious policy considerations involved, the Legislature is the appropriate forum to select from among the competing proposals.15 The case is remanded to the Superior Court for further proceedings consistent with this opinion.

So ordered.

(dissenting in part). I agree that the jury’s finding that Colter unreasonably used the machine, knowing it to be defective, does not bar Colter from recovery against Barber-Greene or New England on a negligence theory. I also agree that, as a matter of law, Barber-Greene was entitled to judgment on Colter’s claim of negligent failure to warn. There*66fore, the court would be correct in remanding the case for retrial of the negligent design claim against Barber-Greene had the evidence been sufficient to warrant a finding that Barber-Greene’s negligent design caused Colter’s injury. However, in my view, although there was evidence of negligent design, there was no evidence of a causal relationship between that negligence and Colter’s injury. Therefore, I would order the entry of a judgment for Barber-Greene, and I would remand the case solely for retrial of the negligence claims against New England.1

When Worcester purchased the machine in 1952, it was equipped with a gear guard. By the time New England took the machine from Worcester, the guard was missing. The court concludes that, “[i]n the absence of any testimony as to what actually happened to the gear guard while the sand classifier was in Worcester’s possession, the jury were entitled to infer that the guard was removed for greasing and repair and was not replaced because it was too cumbersome, not efficient, and too costly in the labor required to remove and replace the gear guard.” Ante at 58-59. I disagree. It is true that the jury could have found that a reasonably prudent manufacturer would have designed the machine differently so as to reduce the need to remove the guard in order to make repairs. It is also true that the jury could have found that a reasonably prudent manufacturer would have designed the machine differently so as to eliminate the need to remove the guard in order to clean and grease the gears, and that the jury could have found that Barber-Greene reasonably could have foreseen that, due to efficiency and cost considerations, a user of the machine would choose not to replace the guard after removing it. It follows *67that, if there had been evidence in this case that the guard had been removed in order to facilitate repairs or the cleaning and greasing of gears, which would have been unnecessary had the alternative design been employed, and if there had been evidence that the guard had not been restored due to concerns about efficiency and cost, the jury would have been warranted in finding that Barber-Greene’s negligence caused Colter’s injury. However, there was no such evidence. As the court says, ante at 58, there was an “absence of any testimony as to what actually happened to the gear guard . . . .”

On the evidence, it is entirely speculative whether the guard was off the machine for any of the reasons set forth above. Therefore, although the jury could have found that Barber-Greene took a foreseeable risk in designing its machine, the jury could only have speculated that the injuries suffered by Colter were within that risk. The void in the evidence ought to be fatal to the plaintiff’s case against Barber-Greene. As the court has previously stated, “[i]t is . . . necessary for [a] plaintiff to prove [not only] that the defendant took a risk with respect to the plaintiff’s safety that a person of ordinary prudence would not have taken, [but also] that the plaintiff suffered a resulting injury that was within the foreseeable risk.” Cimino v. Milford Keg, Inc., 385 Mass. 323, 330 (1982).

The court asserts that this case is indistinguishable from Fahey v. Rockwell Graphics Syss., Inc., 20 Mass. App. Ct. 642 (1985). In my view, the cases are critically distinguishable. The distinction focuses on the very point expressed above. In Fahey, the evidence not only disclosed that removal of a machine guard to facilitate speedy production was a risk foreseeable by the manufacturer, but it also disclosed that the plaintiff’s injury resulted from the guard being removed for that very reason. Id. at 645. Unlike here, the injury was shown to have been within the foreseeable risk. As the court correctly notes, ante at 58-59 n.10, the Appeals Court concluded in Fahey that the jury were entitled to decide “whether the circumstances of the guard’s removal and the plaintiff’s subsequent injury were reasonably foreseeable.” Id. at 648. However, in this case, the jury were not in position to do that be*68cause, without knowing what the circumstances of the guard’s removal were, there being no evidence in that regard, the jury could not have inferred whether those circumstances were foreseeable.

Based on the plaintiff’s best evidence, a gambling man with an appreciation of mathematical probabilities might be willing to bet that the guard was removed in order to facilitate repairs or other maintenance that would not have been necessary in the absence of Barber-Greene’s negligence, and was left off the machine as a means of efficiency and economy. However, Colter had the burden of proving by a preponderance of the evidence that his injury was within that foreseeable risk. He did not sustain that burden. Mere mathematical odds, no matter how favorable to a proposition, do not consitute proof thereof by a preponderance of the evidence.

“It has been held not enough that mathematically the chances somewhat favor a proposition to be proved; for example, the fact that colored automobiles made in the current year outnumber black ones would not warrant a finding that an unde-scribed automobile of the current year is colored and not black, nor would the fact that only a minority of men die of cancer warrant a finding that a particular man did not die of cancer. . . . The weight or preponderance of evidence is its power to convince the tribunal which has the determination of the fact, of the actual truth of the proposition to be proved. After the evidence has been weighed, that proposition is proved by a preponderance of the evidence if it is made to appear more likely or probable in the sense that actual belief in its truth, derived from the evidence, exists in the mind or minds of the tribunal notwithstanding any doubts that may still linger there” (citations omitted) (emphasis added). Sargent v. Massachusetts Accident Co., 307 Mass. 246, 250 (1940). Sargent was an action on a policy of life insurance. A majority of the court concluded that the evidence warranted a jury finding “not merely that there was a greater chance that the insured met his death by accident falling within the policy than that he met a different fate, but that death by accident within the policy was in fact indicated by a preponderance of the evidence” (emphasis added). Id. at 251.

*69The view we expressed in Sargent continues to be the law of the Commonwealth. Stepakoff v. Kantar, 393 Mass. 836, 843 (1985). See P.J. Liacos, Massachusetts Evidence 38 (5th ed. 1985). Proof of mathematical probabilities is not enough. Although the jury were warranted in finding that Barber-Greene took a foreseeable risk concerning Colter’s safety that an ordinarily prudent manufacturer would not have taken, the evidence was insufficient to warrant the further finding that Colter’s injury was within that risk. I would order the entry of judgment for Barber-Greene, and, as explained at the outset of this opinion, I would remand the case for retrial of Colter’s negligence claims against New England.