In the jury trial of this product liability case, the trial judge allowed the defendant’s motion for a directed verdict *177at the close of the plaintiffs’ evidence. The essence of the negligence and breach of warranty claims2 was that a foundry hook manufactured by the defendant and incorporated into an overhead hoisting mechanism was defective because the hook had an open “C”-shaped configuration and lacked a latch which would have closed the hook into an “0”-shaped configuration. According to the plaintiffs’ theory of the case, a latch would have prevented the hook from disengaging in the process of lifting of heavy loads. The load in this case was a heavy piece of metal foundry equipment called a cope.
The directed verdict was based on a failure in the expert testimony concerning causation, as well as a failure to demonstrate that the proposed redesigned hook with a latch would have prevented the accident. “When the precise cause is left to conjecture and may be as reasonably attributed to a condition for which no liability attaches as to one for which it does, then a verdict should be directed against the plaintiff.” Currie v. Lee Equip. Corp., 362 Mass. 765, 768 (1973), quoting from Ryan v. Fall River Iron Works Co., 200 Mass. 188, 192 (1908). This is such an instance and, accordingly, we affirm the judgment.
1. Background facts. In order to understand the nature of the alleged design defect, the trial evidence on causation, and the related expert opinions, we provide a description of the foundry hook and the interconnected hoisting device. This is followed by a summary of the trial evidence concerning the manner in which the hoist and foundry hooks were being operated by the plaintiff Pedro Fidalgo and a coworker on the day of the accident.
a. The foundry hook. In this complex industrial lifting system, a hoist was attached to a bridge crane, which moved along rails across the ceiling of the foundry building. As the hoist descended from the ceiling, its vertical component divided into an iron beam spreader bar. Two chains hung from each end of the iron beam spreader bar. A foundry hook was attached to the end of each of the two chains. In the hoisting maneuver, the foundry hooks were to be attached to the approximately one-*178half ton cope which was to be lifted and rolled. The cope was the top part of a mold known as a flask; the bottom part of the flask was called the drag. To configure the flask and join both of its components, the cope was to be hoisted, rolled, and placed on the drag. Specifically, in the hoisting mechanism, the foundry hooks were supposed to envelop two cylindrical metal posts called trunnions,3 which were welded onto, and extended from, the two sides of the cope. In essence, the foundry hooks were that component of the hoist mechanism that connected to the cope, so that the cope could be lifted off the ground by the chains.
In its original position before being lifted, the cope rested in a vertical position on the floor of the foundry building. The hoist would be used to lift the vertical cope and, while it was suspended, the workers would roll the cope to a horizontal position. Then, the bridge crane would be operated to move the cope along the ceiling rails to the place where the bottom drag piece was positioned horizontally on the floor. The cope would be lowered from the hoist and stacked on top of the drag. As so combined, the cope and drag would create a metal flask. Later in the foundry process, the metal flask would be filled with sand and used to make a mold. Ultimately, molten metal would be poured into the flask mold to form the desired shape for the forged product.
There were two critical steps in the hoisting process that are pertinent to this case. The first critical step was that the two foundry hooks had to be firmly ensconced around the two cylindrical trunnion posts extending from the cope sides. Mis-positioning of a foundry hook at an angle or on the ridge-like flange near the trunnion end posed the risk of disengagement. The second critical step was that the workers had to ensure that the two chains to which the foundry hooks were attached remained taut. If a chain were inadvertently to slacken during *179the hoisting process, there was also risk that the hook might move from the correct position and not remain secure around the trunnion. Then, when the chain became taut again, the hook could reattach improperly, which would also give rise to the risk of disengagement during the lifting and rolling maneuver. If there were a failing at either of these two critical junctures, there was, of course, a very real and significant risk that the cope could drop out of the hoist and fall. To that end, workers engaged in the hoisting maneuver were supposed to double check the original hook positioning to ensure proper positioning on, and a snug fit around, the cape’s trunnions. In addition, the workers were supposed to operate the hoist’s electric motor carefully to maintain sufficient tension, so that the chains remained taut without any slackening throughout the entire process. These two points, as will be seen, were the subject of material evidence in the trial testimony of Pedro Fidalgo and his coworker concerning what happened on the day of the accident. That evidence, in turn, affected the bases of the expert opinions.
b. The accident description. The trial testimony was as follows. The plaintiff Pedro Fidalgo, a worker at the Palmer Foundry, was seriously injured4 when the one-half ton cope disengaged from the subject foundry hooks, as it was being lifted and rolled and fell from the hoist, pinning Fidalgo beneath. On the day of the accident, Fidalgo was working with David Pinto. Both men testified at trial. Although there was some overlapping of the workers’ functions in the hoisting process, Fidalgo on that day was responsible for attaching the foundry hooks to the trunnion posts on the cope. Pinto assisted with the hoist motor controls. Each described their respective acts that day in attaching the foundry hooks, running the motor to keep the chains taut, and lifting and rolling the cope. According to both Fidalgo and Pinto, each followed all the standard procedures for this maneuver. From Fidalgo’s end, he was certain he had properly positioned both foundry hooks in the correct position, securely surrounding the left and right trunnions. After hooking both trunnions, Fidalgo checked to ensure that both hooks were properly around the trunnions. Fi-*180dalgo observed absolutely no slack in the chains while performing this function. Pinto then operated the electronic motor controls to lift the cope. Pinto confirmed Fidalgo’s testimony that the hooks were properly engaged prior to lifting the cope and that no slack developed in the chains. On the day of the accident, the cope lifted smoothly off the ground for a short distance. Then, suddenly, as the cope was being rolled, it fell from the left hook, hitting Fidalgo and pinning him beneath it.
2. The expert evidence. The plaintiffs tendered two experts: Richard Sisson, Ph.D., a professor of mechanical engineering at Worcester Polytechnic Institute; and Robert Holt, a mechanical engineer. The experts opined how the accident was caused. The experts also opined that the “C”-shaped configuration in the foundry hook manufactured by the defendant was defective, and each expert proposed redesign of the foundry hook with a latch. a. The “tip loading” theory. Both Sisson and Holt tendered opinions that the accident most probably occurred by chain slackening after the initial engagement. Then, when the chain became taut again, the foundry hook became improperly positioned, contacting the trunnion only with the hook’s tip. So positioned on the hook tip, the bowl of the hook did not engage and surround the trunnion post. The experts called this “tip loading.” With such precarious loading only on the hook tip, as the hoist moved, the hook tip slipped off the trunnion. Thereupon, the cope disengaged and rapidly descended onto Fidalgo.5
Although the experts rested their opinions on the theory of tip loading, both experts conceded that tip loading of the hook’s metal tip on a cylindrical metal trunnion was a phenomenon not susceptible to being replicated. Indeed, neither expert offered any reliable evidence of how, as a matter of physical science, the rounded metal tip of the hook could be balanced on the round metal trunnion post so as to support the approximately *181one-half ton metal cope, either after chain slackening or during any point in the hoisting maneuver.6
Sisson offered no evidence of any tests performed to determine whether it was physically possible for the tip of a hook to support and lift the cope. Without offering such testing verification and calculations, Sisson guessed that tip loading might not happen “maybe ninety-nine times out of a hundred.” The record reveals no support for this statistical guess.
Holt also adopted the tip loading theory. However, when pressed in cross-examination, Holt essentially conceded that the possibility of tip loading as the cause of this accident was remote. Despite Holt’s assertion that tip loading “happens a lot in industry,” he was forced to acknowledge that the phenomenon of double tip loading might never be replicated. Holt speculated that, theoretically, it could take “three weeks” to replicate tip loading of even one trunnion and that “you may never be able to” reproduce the phenomenon of double tip loading. Holt, like Sisson, offered no testing verification or calculations to support his double tip loading theory.
b. The flaws in the proposed latch redesign. On direct examination, both experts also opined that had the foundry hook borne a latch, the latch would have held the hook fast by surrounding the trunnion. Then, according to both experts, when the chain slackened, tip loading would have been avoided and the cope would not have fallen.
On cross-examination, however, both experts conceded that, in circumstances of improper loading of the trunnion in the hook, the redesigned latched hook they proposed would not have prevented the hook from slipping off the trunnion. This is so because a redesigned latched hook would require increased dimensions to accommodate the latch, and the resulting opening *182in the hook circumference would be so large that the trunnion could have slipped out of the latched hook, and the cope would have fallen anyway.
Beyond that flaw, the two experts’ redesigns for a latched hook left much to the imagination. Sisson had attempted only very rough sketches, but even the rough sketches he had done were so inadequate that Sisson suggested that better pictures of hooks with latches were available in catalogues. Sisson stated that he had not gone “through the whole process of designing the fastener for this hook,” leaving that as the “sort of thing [the professor] would give as a homework assignment.” All that Sisson displayed as evidence was what the judge charitably characterized as a “very crude model,” consisting of “two thumb tacks and two pieces of paper.” Holt proffered no model. Holt had only general dimensions stored in his mind, which he thought might possibly work. Even measured by these lineless dimensions, Holt admitted the imagined redesigned latch would not have prevented the hook from slipping off the end of the trunnion.7
3. The directed verdict. In granting a directed verdict, the judge identified these — and more — critical flaws and omissions in the plaintiffs’ proof. The judge rejected the experts’ theory of tip loading as too speculative to support a reasonable inference regarding the accident’s cause. Indeed, she found the experts’ explanations — all of which were theoretical and stated only in possibilities — farfetched. She characterized the untested theory of double tip loading as “absurd” and “beyond speculation.”
a. Causation issue. A directed verdict is warranted “notwithstanding that an expert has testified that the cause of the accident was the defendant’s negligence, if it is demonstrated that the opinion is based on speculation alone.” Currie v. Lee Equip. Corp., 362 Mass. at 768. In proof of causation, the plaintiff *183“must show that there is a greater probability than not that the accident resulted from the defendant’s negligence.” Enrich v. Windmere Corp., 416 Mass. 83, 87 (1993). This the plaintiffs’ experts failed to do. “An expert opinion, stated in terms of possibilities, does not satisfy the plaintiff’s burden of proof, namely, to establish by a preponderance of the evidence the existence of a design defect which caused the plaintiff’s injuries.” Goffredo v. Mercedes-Benz Truck Co., 402 Mass. 97, 102-103 (1988). Enrich v. Windmere Corp., supra at 87.
b. Design issue. With respect to the design issue, the judge correctly determined that the evidence failed to show that the redesigned latched hook proposed by the experts would have prevented the accident. Indeed, the judge found Sisson’s testimony “to be strange, ... for lack of a better phrase,” and of “absolutely no help whatsoever, in terms of producing any credible evidence that there was a latch that could be made for this hook.” As the judge analyzed the experts’ purported redesign testimony: “[B]y the time you had the latch engagement, the drop would have exceeded the three-quarter inch safety zone by many inches, and the accident would have happened exactly the same way as it happened here. It would have been an absolutely useless latch for no purpose whatsoever.” See generally Colter v. Barber-Greene Co., 403 Mass. 50, 57 (1988), quoting from Uloth v. City Tank Corp., 376 Mass. 874, 881 (1978) (to go to the jury, the plaintiff must show “an available design modification which would reduce the risk”). Simply put, the expert testimony in this case failed to demonstrate that a latched hook was a feasible design for this operation. See generally Back v. Wickes Corp., 375 Mass. 633, 642 (1978), quoting from Barker v. Lull Engr. Co., 20 Cal.3d 413, 431 (1978) (among the factors to be weighed by the jury in considering a product’s design is “the mechanical feasibility of a safer alternative design”).
In light of these failures of proof, the judge correctly determined that “the evidence, construed most favorably to the plaintiff[s], could not support a verdict for the plaintiff[s].” Goffredo v. Mercedes-Benz Truck Co., 402 Mass. at 101, quoting from Poirier v. Plymouth, 374 Mass. 206, 212 (1978). Cf. Lally v. Volkswagen Aktiengesellschaft, 45 Mass. App. Ct. 317, 324 *184(1998) (“While it is trac that generally a plaintiff may rely on inferences to defeat a motion for judgment notwithstanding the verdict, those inferences cannot be based on mere speculation”). “The inferences must be reasonable and must be based on probabilities rather than possibilities and may not be the result of mere speculation and conjecture.” Goffredo v. Mercedes-Benz Truck Co., supra at 101.
c. The failure to warn. The plaintiffs also argue that a warning should have stated that, with slackening of a chain, there might be disengagement of the hooks. The judge found “a complete lack of proof, that. . . [any] reasonably required warning, would have made any difference in this case whatsoever.” Fidalgo specifically testified that he did not think he needed to be instructed to properly attach the hook before lifting or that slack in the chains was to be avoided. The duty to warn does not attach “where the plaintiff appreciated the danger substantially to the same extent as a warning would have provided.” Carey v. Lynn Ladder & Scaffolding Co., 427 Mass. 1003, 1004 (1998). The judge considered this admission by Fi-dalgo, noting that “[t]here’s no need to warn him of that. He knew that was the essence of what he was checking for.” “A duty to warn is not imposed by law as a mindless ritual.” Killeen v. Harmon Grain Prod., Inc., 11 Mass. App. Ct. 20, 24 (1980).
d. Evidence of industry standards. Through their experts, the plaintiffs introduced standards for overhead hoists promulgated by the American Society of Mechanical Engineers (ASME). The standards stated that hooks should be equipped with latches unless the application made the use of a hook impractical. On appeal, the plaintiffs argue that the jury should have been permitted to decide whether this standard was probative of the defendant’s negligence and defective design. Industry standards may be some evidence of negligence. See Poirier v. Plymouth, 374 Mass. at 211. See also Resendes v. Boston Edison Co., 38 Mass. App. Ct. 344, 358 (1995). However, in this case, the trial judge properly discounted the ASME industry standards because the standards applied to integrated hoisting systems and not to the individual hook employed in such systems. In this case, the hook was purchased separately and added to the hoisting mechanism by the owner of the Palmer Foundry.
*185To this, we add that the judge rightly determined that the ASME standards did not cure the fatal deficiencies in the plaintiffs’ evidence relative to causation and defective design. That is, the evidence in the plaintiffs’ case-in-chief, inclusive of the ASME standard concerning integrated hoist systems, viewed in totality and construed in the light most favorable to the plaintiffs, did not provide a basis upon which a jury could reasonably determine that a latch on the foundry hook would have prevented the cape’s falling from the hook and pinning Fi-dalgo beneath it, resulting in his injuries. See Falvey v. Hamel-burg, 347 Mass. 430, 435 (1964) (violation of a statute, while evidence of negligence, does not make a case for the jury without proof that it was causally connected to the injury).
Based on the foregoing, we conclude that the defendant’s motion for a directed verdict was properly allowed, and we therefore affirm the judgment.
Judgment affirmed.
Also raised was a claim that a warning was necessary to advise of the dangers of using a foundry hook in suspended lifting operations.
At the end of the trunnion posts, a round cap-like shape was attached to enlarge the trunnion circumference. This added a small three-quarter inch rim-like flange, which was supposed to keep the foundry hook from slipping off the trunnion end. However, this small flange rim was not sufficiently large to stop the hook from falling off the trunnion. After the accident, a larger rectangular metal plate was added to the trunnion end in order to more effectively block the hook from slipping off.
Fidalgo’s injuries necessitated numerous surgeries, and he was hospitalized for five months.
We also note, although the outcome does not turn on this point, that the experts’ opinions directly contradicted the testimony of both Pinto and Fidalgo that the hooks were properly engaged and no slack ever occurred in the chains. As such, the experts’ opinions appeared to be based on assumptions not supported by the evidence.
The experts described two other possibilities for how the accident happened but concluded that tip loading, after slackening of the chain, was the most likely cause. Sisson opined that the accident was caused by tip loading on the left foundry hook. Holt’s opinion was that there was double tip loading on both hooks on both trunnions. Holt rejected Sisson’s single-sided tip loading theory of causation as unlikely because the spreader bar on the hoist would be visibly lopsided. That circumstance was inconsistent with the accounts by Fidalgo and Pinto, neither of whom testified to having seen such lopsidedness.
On cross-examination, Holt acknowledged that the trunnion was defectively designed and that a larger flange at the trunnion end would have prevented the hook from disengaging, even if there had been improper loading of the hook on the trunnion. In fact, as noted above, after the accident there was a subsequent remedial repair to the trunnion structure which involved attaching a solid metal plate to the trunnion end in order to prevent the hook from falling off the end.