NIEMEYER, Circuit Judge:
At dusk on December 27, 1984, in rural Kentucky, Chad Sexton, 13, was riding an off-road motorcycle (or “dirt bike”) when he collided with his friend, Jackie Boggs, who was also riding a motorcycle, coming from the opposite direction. Sexton and Boggs were each traveling about 35 miles per hour, and after the collision Sexton was propelled 183 feet from the point of impact. He sustained a closed-head injury and brain damage causing quadriparesis.
At the time of the accident, Sexton was wearing a helmet manufactured by Bell Helmets, Inc., which had been purchased from a retailer in West Virginia over three years earlier, in May 1981, by his mother for his brother. The Bell helmet remained on Sexton’s head throughout the accident without apparent damage except for scrape marks. Sexton and his parents sued Bell on theories of strict liability, negligence, and breach of warranty, alleging that the helmet was defectively designed and that the warning given on the helmet was inadequate. Sexton, the only remaining plaintiff at the time of trial, contends that the liner of the helmet did not crush adequately to absorb the energy from the impact and that the warning failed to disclose that the helmet was designed for an average adult, rather than a child.
From a jury verdict in favor of Sexton in the amount of $1,551,381, Bell appeals, contending (1) that the trial court improperly failed to instruct the jury on an applicable statutory presumption; (2) that the trial court improperly refused to strike plaintiff’s expert testimony and to grant defendant’s motion for a directed verdict; and (3) that the court erred in refusing to find Sexton contributorily negligent as a matter of law. For the reasons that follow, we vacate the judgment and remand the case for a new trial.
I
Jurisdiction rests on diversity of citizenship, and the trial court determined that the controlling substantive law is that of Kentucky, a determination with which none of the parties takes issue.
Bell contends that the district court erred in refusing to instruct the jury on a presumption created by the Product Liability Act of Kentucky, Ky.Rev.Stat.Ann. (K.R.S.) §§ 411.300 et seq. (Michie 1990). The relevant portion of that act provides:
In any product liability action, it shall be presumed, until rebutted by a preponderance of the evidence to the contrary, that the product was not defective if the design, methods of manufacture, and testing conformed to the generally recognized and prevailing standards or the [333]*333state of the art in existence at the time the design was prepared, and the product was manufactured.
K.R.S. § 411.310(2). Bell points out that at the time of the helmet’s design in 1979 and its manufacture in 1980, three industry standards for the manufacture of helmets existed: (1) the American National Standards Institute Specifications for Protective Headgear for Vehicular Users, Z90.1-1971 (the ANSI Z90 standard); (2) the Federal Motor Vehicle Safety Standards and Regulations for Motorcycle Helmets No. 218 (FMVSS No. 218, also called the DOT standard); and (3) the Snell Memorial Foundation Standard for Protective Headgear 1975 (the Snell 75 standard). The parties agree that the helmet worn by Sexton complied with these standards and that there were no other standards applicable. Bell argues, therefore, that application of the presumption created by the Kentucky statute “required a verdict for Bell” and that the jury should have been so instructed. Brief of Appellant at 13 (emphasis added).
Despite Bell’s argument to the contrary, the Kentucky statute does not provide that if Bell’s product complied with applicable standards, Bell would be entitled to a judgment on a claim that the product was defective. Rather, the statute creates a presumption, which, in the absence of evidence to the contrary, will prevail. In this case, the district court concluded that substantial evidence had been presented that the product was defective, and therefore the presumption was no longer a matter to be considered. Having “burst the bubble” by the introduction of evidence, the plaintiff was entitled to have the question decided on the evidence, and the presumption was no longer applicable. See Mason v. Commonwealth, 565 S.W.2d 140, 141 (Ky.1978) (quoting McCormick’s Handbook of the Law of Evidence § 345, at 821 (2d ed. 1972)).
We can perceive no reason why the trial court would ever have need to instruct the jury on this statutory presumption. If a plaintiff fails to rebut the presumption “by a preponderance of the evidence to the contrary,” a verdict will be directed that the product was “not defective.” K.R.S. § 411.310(2). On the other hand, if a plaintiff rebuts the presumption, the issue of whether the product was defective will be submitted to the jury on the evidence. For this reason the Kentucky courts have not generally permitted advising the jury of any presumptions except in a narrow set of circumstances not applicable here. See Lowe v. McMurray, 412 S.W.2d 571, 574 (Ky.1967) (general prohibition against instructing jury as to questions of law or fact); Mason, 565 S.W.2d at 141 (jury should not be informed of presumption of sanity in criminal cases). But cf. Wells v. Commonwealth, 561 S.W.2d 85, 86 (Ky.1978) (particular statute involving expert testimony creates a presumption of fact and jury may be informed of it).
This analysis prompts us to question the utility of such a statutory presumption, because a defendant always carries a presumption in his favor when the burden of proof is placed on the plaintiff. If the plaintiff fails to present substantial evidence to prove his case that a product was defective, the court will direct the verdict; if the plaintiff presents substantial evidence that carries him over the directed verdict threshold, the evidence will take the case to the jury and, at the same time, overcome the presumption created by the statute. The standard for rebutting the presumption under the statute, i.e., by introduction of evidence that shows a defect by a preponderance of the evidence, see K.R.S. § 411.310(2), is the same as that which makes out a prima facie case for the plaintiff.
Other courts have shared our puzzlement over the need for the statute as presently drafted. See Anderson v. Black & Decker (U.S.), Inc., 597 F.Supp. 1298, 1301 n. 3 (E.D.Ky.1984); Ingersoll-Rand Co. v. Rice, 775 S.W.2d 924, 928 (Ky.Ct.App.1988); see also J.A. 491, 499-500. If the statute was intended to mean what Bell contends, i.e., that the presumption is irre-buttable, then the legislature of Kentucky will have to address the issue and modify the language of the statute. Under the statute in its present form, however, we cannot say that the district court erred in [334]*334refusing to instruct the jury regarding the presumption.
II
Bell also contends that the testimony of Dr. Kenneth Saczalski and Dr. Loma Mid-dendorf, plaintiffs experts who testified that the helmet was defectively designed and that the warning was inadequate, should have been stricken. Preserving the same issue by a different procedural means, Bell also contends that the court should have directed a verdict in its favor because the plaintiff did not prove either a defective design or an inadequate warning.
In the 1970’s Bell redesigned its existing helmet, the Bell Star II, and produced a new helmet, the Bell Star III, the version worn by Sexton, to accommodate greater speeds and higher impacts that were anticipated in the use of helmets. The helmet was advertised for use in automobile and motorcycle racing, although it was also expected to be used on highways and for recreational use. J.A. 651. In the brochure that accompanied its helmets Bell stated:
Your Bell helmet is designed to protect your head in two ways. It spreads the energy of a sharp impact over a wide area so that you are better able to withstand its force. Secondly, it absorbs much of the impact’s energy rather than transmitting it to your head. Helmet damage after an accident does not indicate a deficiency in the helmet. On the contrary, it shows that the helmet performed its function.
Brief of Appellee at 6. The advertised performance criteria were addressed by the differing functions of the helmet’s outer shell and its inner liner. The outer shell is rigid and spreads the energy of a sharp impact; the inner liner crushes under impact in order to absorb shock. If the liner is too soft, it crushes “all the way down” and “bottoms out” and fails to absorb all of the shock. J.A. 235. On the other hand, if the liner is too stiff, it provides absorption capacity for higher impacts but fails to provide enough cushion for lesser impact collisions. J.A. 235, 242. Gradations in between provide differing levels of protection.
Bell made the decision to stiffen the liner of the Bell Star III, over its earlier version, the Bell Star II, to avoid “bottoming out” and therefore to provide greater protection in more serious collisions. The availability of choice in design was recognized even before the 1979-80 period when the Bell Star III was designed. In 1976 Bell’s manager of manufacturing standards, Jim G. Sundahl, wrote in an article:
Bell helmets use liners having non-linear rates. They are carefully designed and molded to be relatively stiff as they begin to crush. As they are crushed further, they do not become too stiff. This means that in a minor impact the Bell helmet transmits a relatively high acceleration to the head, especially in comparison to many helmets approved only to the DOT standard. However, in a more massive impact, the Bell helmet transmits an acceleration level only a bit higher than in the lesser impact. Other helmets designed with softer liners often “bottom” at higher impact levels.
J.A. 242.
It is interesting to note that Bell helmets could be made to more easily pass some helmet tests such as the DOT by using a softer and more resilient liner. Such a liner, however, would not provide as much protection from higher impact energy levels.
J.A. 245.
The parties agree that the Bell Star III helmet used by Sexton was not defectively manufactured and that its design met or exceeded the only three standards in the industry: the ANSI Z90 standard, the DOT standard, and the Snell 75 standard. They also agree that there were no special standards for helmets worn by children. J.A. 251-54.
Dr. Saczalski, who was the only witness proffered by plaintiff on whether the design of the helmet was defective, testified that Bell improperly selected too stiff a helmet, which did not provide enough crushability for low impact situations. [335]*335Summarizing his opinion, he testified, “[T]he liner in this particular helmet was too stiff to provide the crush and the slowing down of Chad’s [Sexton’s] head in a relatively minor impact situation. So that’s my opinion.” J.A. 231. While he acknowledged that the helmet in question showed some crush, he did not measure the degree of that crush and did not test the helmet in this case. J.A. 283-85. He concluded, however, that there was “very little evidence” of crush and that the impact between the helmet and the ground was a “relatively minor accident.”
Focusing on Bell’s conduct, he noted that Bell had various options when designing the helmet to make the liner stiffer or softer. As he stated,
There are a number of options that a design engineer has in putting together all of these pieces or parts of this component here that are going to absorb the energy over the range that one would desire, that one would select, and for the sizes and weights that one would anticipate being used in this helmet.
J.A. 232-33. He argued that Bell made a “conscious decision” to go to a stiffer material that “would exceed or in essence be better than what would be required just by, say, meeting the Department of Transportation standard.” J.A. 235. This decision forms the basis of his opinion that Bell bears responsibility for the injuries suffered by Sexton because it thereby accepted the risk of low impact accidents.
Dr. Saczalski explained that technology existed to design a helmet that accommodated both higher and lower impact incidents. While no one in the industry had designed or manufactured such a helmet, Dr. Saczalski said he had built a prototype to show that it could be done. J.A. 272. He presented no evidence, however, which revealed that Bell or the industry recognized a need for addressing lower impact situations, and he identified no standard or expectation that the Bell helmet did not fulfill. Beginning with assumptions that Sexton’s collision was a low impact one and that the stiffness of the helmet was the cause of Sexton’s injuries, assumptions that Bell contested, Dr. Saczalski confined his opinion to the conclusions that the technology was available and Bell therefore “could” have addressed the risks of Sexton’s accident. J.A. 247, 311. As he testified repeatedly, “The technology was there. They were aware of it. They could have done it.” J.A. 263.
Sexton argues that the testimony of Dr. Saczalski supports his claim that Bell manufactured and sold a helmet with a defective design and that the helmet was therefore unreasonably dangerous for his use. He argues that his proof of a defective condition exceeds that which was described as inadequate in Ingersoll-Rand, 775 S.W.2d at 929, where the court stated:
On remand, unless Rice can present something more than a conclusion that it was theoretically probable that a different design would have been feasible and would have prevented his injury, we are of the opinion that the issue of strict liability should not be submitted to the jury-
Bell, on the other hand, argues that Dr. Saczalski’s testimony amounted to no more than the type of conclusory testimony described in Ingersoll-Rand. It argues that Dr. Saczalski’s testimony, which tended to prove only that known technology was available to protect against low impact accidents and that Bell could have utilized it to have avoided the injuries in this case, failed to prove that the product was defective.
Kentucky adopted Restatement (Second) of Torts § 402A in 1966 as a basis for making a claim for strict liability. Dealers Transp. Co. v. Battery Distrib. Co., 402 S.W.2d 441 (Ky.1966). Although Kentucky also imposes the traditional duty of care on manufacturers to make products reasonably safe, when a product liability claim is based on a design defect, the articulation of liability, whether based on a negligent breach of a duty of care or on strict liability, reduces to the single question of whether the product was defective. A negligence theory focuses on the conduct of the manufacturer in adopting an allegedly defective design, and strict liability focuses on the product itself. See Montgomery [336]*336Elevator Co. v. McCullough, 676 S.W.2d 776, 780 (Ky.1984). But in either case, proof of a defective product is essential to maintaining the cause of action. As noted in Montgomery Elevator, “[t]he sole question in a products liability case is whether the product is defective as defined in Nichols v. Union Underwear [602 S.W.2d 429 (Ky.1980) ].” Id. at 782. See also Jones v. Hutchinson Mfg., Inc., 502 S.W.2d 66, 69-70 (Ky.1973) (In defective design cases, “the distinction between the so-called strict liability principle and negligence is of no practical significance.”). The issue presented therefore is whether the testimony of Dr. Saczalski offered proof that the Bell Star III helmet was defectively designed.
In the leading Kentucky case addressing the question of what constitutes a defective design, Nichols v. Union Underwear Co., 602 S.W.2d 429 (Ky.1980), the court stated:
We are immediately met with the difficult problem of describing the standard to which the fact finder should compare the product to decide whether it was in a “defective condition unreasonably dangerous” when sold. In Ulrich v. Kasco Abrasives Company, Ky., 532 S.W.2d 197, 200 (1976), we pointed out that the inquiry is to be made from the perspective of a “prudent manufacturer of similar products fully apprised of the condition and tendencies of the product when he put it into the stream of commerce .... ”
In design defect cases liability is founded upon the premise that the design itself selected by the manufacturer amounted to a defective condition which was unreasonably dangerous....
Id. at 432-33 (footnote omitted). Although the court thus directs the inquiry to the two core elements of a strict liability claim described in Restatement (Second) of Torts § 402A, that the product be in a defective condition which causes it to become unreasonably dangerous, it fails to give the standard by which a condition of defectiveness is to be measured.
Although a manufacturer has a duty to design a product that is reasonably safe for its foreseeable use, the manufacturer is not required to design “the best possible product.” Jones, 502 S.W.2d at 69. That a product is inherently dangerous, or that it creates a risk of harm, or that it causes injury does not, alone, make it defective. Some products, no matter how carefully manufactured, are inherently dangerous, and virtually every product is capable of causing injury. As is often stated, manufacturers are not insurers against all injury involving their products. Id. at 70. Similarly, proof that technology existed, which if implemented could feasibly have avoided a dangerous condition, does not alone establish a defect. Id. at 70-71. Thus, for example, automobiles in the 1950’s were manufactured without seat belts, even though the technology was available for their installation. Yet, as counsel for Sexton conceded at oral argument, we would not now impose liability on the manufacturer of those automobiles on the basis that they were defective for not having seat belts. Government and industry standards did not require them and consumers did not, at the time, demand or expect them.
It is, no doubt, significant when no one in an industry manufactures a product in a manner that a plaintiff claims is required. While conformity with industry practice is not conclusive of the product’s safety, because an industry could adopt a careless standard, the cases where a member of industry will be held liable for “failing to do what no one in his position has ever done before” will be infrequent. Jones, 502 S.W.2d at 70 (quoting W. Prosser, Handbook of the Law of Torts § 33, at 167 (4th ed. 1971)). It is perhaps more significant that a reasonable purchaser does not demand or expect that the product incorporate a particular aspect of design. Thus, comment g to the Restatement (Second) of Torts § 402A, in explaining the “defective condition” element of strict liability, states, “The rule stated in this Section applies only where the product is, at the time it leaves the seller’s hands, in a condition not con[337]*337templated by the ultimate consumer... (Emphasis added.)
A defect can therefore be identified by measuring the product against a standard articulated expressly by government or industry or established by society in its expectations held about the product at the time of its sale. While government and industry standards are readily identifiable for a given product at a given time, the reasonable expectation of purchasers requires a factual examination of what society demanded or expected from a product. This may be proved from evidence of actual industry practices, knowledge at the time of other injuries, knowledge of dangers, the existence of published literature, and from direct evidence of what reasonable purchasers considered defective at the time. While society demands and expects a reasonably safe product, an examination of societal standards at any given point in time usually reveals an expectation that balances known risks and dangers against the feasibility and practicability of applying any given technology. See Montgomery Elevator, 676 S.W.2d at 780. With respect to the seatbelt hypothetical, society did not consider an automobile manufactured in the 1950’s defective merely because it had no seat belts. Likewise, society does not currently expect automobiles to be manufactured to eliminate all risks of blindspots when operated in reverse. Existing technology would undoubtedly permit rearview video cameras or beeper warnings which could operate while the automobile is in reverse. Nevertheless, the automobile would not be considered defective because it is not equipped with these devices.
In short, a product can only be defective if it is imperfect when measured against a standard existing at the time of sale or against reasonable consumer expectations held at the time of sale.
There is no evidence in this case that purchasers of motorcycle helmets in the 1979-81 period, when the Bell Star III helmet was designed, manufactured and sold, reasonably expected a higher level of protection than that called for by the existing government and industry standards. There also is no evidence that purchasers of helmets expected more protection in helmets worn primarily by children than in helmets worn mainly by adults.
Furthermore, the record in the case before us reveals no government or industry standard that would have required Bell to incorporate Dr. Saczalski’s prototype design into its helmets. The goals for technology at the time the Bell Star III helmet was designed were to accommodate higher impact accidents that were incidental to higher speeds. Nothing in the record suggests that focusing on higher impact accidents created an unacceptable risk for lower impact accidents. As has already been noted, the Bell Star III helmet satisfied or exceeded all three of the existing government and industry standards. No literature was presented to indicate that anyone —industry, government, or consumers— recognized a need for greater protection in low impact accidents than that provided for by these standards.
Dr. Saczalski testified to that which could have been done to avoid plaintiff’s injuries in this case. But that is not the equivalent of what was required to have been done. See Ingersoll-Rand, 775 S.W.2d at 928-29; Jones, 502 S.W.2d at 70. It is true that Dr. Saczalski went beyond simply stating what could have been; he proffered a prototype allegedly based on technology that existed in 1979. The district court relied on this in denying Bell’s motion for judgment notwithstanding the verdict and for a new trial. The district court observed:
Notwithstanding defendant’s arguments to the contrary, the plaintiff’s experts not only presented more than a “theoretical possibility” that a different design would have better protected the plaintiff from the injurious consequences of the accident; one of plaintiff’s experts, Dr. Kenneth Saczalski, also testified as to the nature of such alternative design and to the defendant’s knowledge of the feasibility of incorporating such an alternative design in its product line.
J.A. 808. The court thereby attempted to distinguish the ruling in Ingersoll-Rand, [338]*338where the court concluded that a plaintiff, in making out a product liability case, must “present something more than a conclusion that it was theoretically probable that a different design would have been feasible and would have prevented his injury.” 775 S.W.2d at 929.
Nevertheless, the fact that as of the time of trial Dr. Saczalski was able to develop a prototype using technology available to Bell in 1979 does not lead to the conclusion that the design of the Bell Star III helmet was defective. That an alternative design was “feasible” is at best hindsight. Looking back on the progressions of technology, it is easy to fall into a mind-set of imposing current knowledge and expectations on older standards or expectations. But we cannot, under the law, judge yesterday’s design by contemporary standards or expectations. For this reason, Dr. Saczalski’s testimony falls short of establishing a defect, i.e., an imperfection measured against a then-existing standard or reasonable consumer expectation. We conclude, therefore, that his testimony should not have been considered by the jury in rendering a verdict in this case.
Independent of the claim for defective design, Sexton contends that the warning provided by Bell of a potentially dangerous condition was inadequate. Dr. Loma Mid-dendorf, who provided the testimony that the warning was inadequate, stated:
It’s my opinion that that information should be on the warning or in the product instructions that children should not use this particular helmet, or this helmet is not designed to protect children, or this helmet is designed not to protect you in certain low-impact or relatively minor-impact injury situations.
J.A. 335. She concluded by stating that the warning principles that she described were “well-known and recognized by the engineering, and manufacturing and industrial community in 1979 and before.” J.A. 336.
While it is apparent from a review of Dr. Middendorf’s testimony that it was in part dependent on the design defect advanced by Dr. Saczalski, see J.A. 334-35, 349, we cannot rule as a matter of law whether she could give her opinion in the absence of Dr. Saczalski’s opinion and still satisfy the requirement of Kentucky law that Bell must have had reason to anticipate the danger omitted from its warning. See Garrison v. Rohm & Haas Co., 492 F.2d 346, 352 (6th Cir.1974) (citing Post v. American Cleaning Equip. Corp., 437 S.W.2d 516 (Ky.1968)). A new trial is therefore required for that limited purpose.
Ill
Finally, Bell contends that the district court erred in failing to grant its motion for a directed verdict on the issue of Sexton’s alleged contributory negligence under the applicable statute. It argues that Sexton was negligent as a matter of law and that therefore it was entitled to a verdict under the Kentucky statute, which provides:
In any product liability action, if the plaintiff failed to exercise ordinary care in the circumstances in his use of the product, and such failure was a substantial cause of the occurrence that caused injury or damage to the plaintiff, the defendant shall not be liable whether or not said defendant was at fault or the product was defective.
K.R.S. § 411.320(3).
Kentucky enacted the Product Liability Act in 1978 to restrict and limit common law actions for product liability. Reda Pump Co., Div. of TRW, Inc. v. Finch, 713 S.W.2d 818, 820 (Ky.1986). The limitation adopted in K.R.S. § 411.320(3) was enacted specifically to allow for contributory negligence as a defense in product liability cases. See Hilen v. Hays, 673 S.W.2d 713, 715 (Ky.1984). The availability of contributory negligence in this type of case had been the subject of some debate because of the traditional rule of strict liability, as embodied in Restatement (Second) of Torts § 402A, which precludes contributory negligence as a defense. Id. at comment n.
Summarizing the evidence, Bell contends that Sexton was operating an unlighted, unlicensed, unregistered, uninspected motorcycle, designed solely for off-road use, [339]*339on a public highway at or near sunset. At the time of the accident Sexton was travel-ling near the center of the road on which Jackie Boggs was coming towards him from the opposite direction. When asked why he failed to see Jackie Boggs before the accident, Sexton replied that he “must have been looking off.” J.A. 465. Bell argues that because Sexton failed to exercise ordinary care, and his failure was a substantial cause of the accident, the trial court should have directed a verdict in favor of Bell in reliance on the Kentucky statute. See Reda Pump, 713 S.W.2d at 820.
While the negligence of Sexton arguably may have been a contributing cause of the accident, no evidence has been advanced to show that he negligently used the helmet or that the negligent use of the helmet was a cause of his injuries. Yet both are specific requirements of the statute. It provides that “[i]f the plaintiff failed to exercise ordinary care ... in his use of the product” and if that negligence was “a substantial cause of the occurrence that caused injury or damage” then the defendant shall not be liable. K.R.S. § 411.320(3) (emphasis added). The statute defines contributory negligence as the failure to exercise ordinary care in the use of the product in question.
At the time of the accident, all evidence showed that Sexton was properly wearing the Bell helmet and that it was still strapped to his head after the accident. There is no evidence to suggest that he failed to make any necessary adjustment. Therefore, the trial court did not err in refusing to instruct the jury as a matter of law that Sexton’s negligence was a bar to his product liability claim under the Kentucky statute.
Although the facts do not permit a finding, as a matter of law, that Sexton was contributorily negligent as defined by K.R.S. § 411.320(3), the issue of whether Sexton’s conduct amounted to contributory negligence as a matter of common law is a different question which we need not reach because contributory negligence is not a defense to a claim based on strict liability, see Restatement (Second) of Torts § 402A, comment n, except as provided by K.R.S. § 411.320(3).
In summary, because the jury was improperly permitted to consider the testimony of Dr. Saczalski, we vacate the judgment and remand for a new trial.
VACATED AND REMANDED WITH INSTRUCTIONS.