Plaintiff Linda McCathem brought this product liability action against Toyota Motor Corporation, Toyota Motor Sales, U.S.A., Inc., and Broadway Toyota (collectively Toyota) for damages arising from severe permanent injuries she sustained when the 1994 Toyota 4Runner in which she was a passenger rolled over at high speed. At trial, the jury returned a verdict for plaintiff, and Toyota appeals. We conclude that the trial court did not err in denying Toyota’s motions for directed verdict and judgment notwithstanding the verdict (JNOV), because plaintiff submitted evidence from which a jury could conclude that the 1994 4Runner failed to meet reasonable consumer expectations, and was, therefore, defective. We further conclude that the trial court did not abuse its discretion in admitting plaintiffs evidence of “other similar incidents” or in denying defendant’s motion for a new trial. Accordingly, we affirm.
Because the jury returned a verdict in plaintiffs favor, we view the evidence and all inferences that may reasonably be drawn from the evidence in the light most favorable to plaintiff. Greist v. Phillips, 322 Or 281, 285, 906 P2d 789 (1995). In 1994, plaintiffs cousin, Elizabeth Sanders, leased a 1994 Toyota 4Runner sport utility vehicle.1 On May 28,1995, plaintiff and her young daughter and Sanders and her young daughter, were returning to Portland from Idaho in Sanders’s 1994 4Runner. At about 9:30 that evening, the 4Runner was traveling at approximately 55 miles per hour on a straight, flat portion of Washington State Highway 395, a two-lane highway with wide, paved shoulders. Sanders was driving, plaintiff was in the front passenger seat, and the two children were in the back seat. Sanders saw oncoming headlights in her lane of travel, and to avoid a collision with the oncoming vehicle, sharply steered her vehicle to the right on to the paved shoulder. To prevent the 4Runner from going off the pavement on the right side of the highway, Sanders countersteered quickly to the left, sending the vehicle across both her lane of travel and the northbound lane. *205She then countersteered again, back to the right, to avoid leaving the road on the left side. After that third steering input, the 4Runner began to roll. It rolled over twice and then came to rest holly upright in the center of the highway, without ever having left the pavement. In the course of the rollover, the roof over the front passenger seat was crushed. Plaintiff suffered a broken neck and was permanently paralyzed from the neck down. Sanders and the children suffered no permanent physical injuries.
In January 1996, plaintiff brought this action against Toyota, pleading claims of negligence and strict products liability. Plaintiff alleged, inter alia, that the 1994 4Run-ner was “dangerously defective and unreasonably dangerous in that it was unstable and prone to rollover” as designed and sold.
The case was tried to a jury over three weeks in March and April of 1997. Plaintiff offered detailed and often technical testimony from a variety of expert witnesses, including Thomas Fries, a mechanical engineer specializing in accident reconstruction, Leon Robertson, a statistician specializing in injury statistics, and Simon Tamny, another accident reconstruction expert. Fries presented his analysis of the mechanics of the rollover of Sanders’ 4Runner and testified that, in his opinion, the sole cause of the rollover was “the geometry of the vehicle” — there was no evidence of braking, off-pavement travel, or a “rim trip.”2 Robertson presented evidence regarding the correlation between the height of a vehicle’s center of gravity and its rollover resistance. Tamny testified that, in his opinion, the 1994 4Runner was unreasonably dangerous because tire friction forces generated by hard steering could — by themselves — cause the vehicle to roll over instead of sliding to a stop. Tamny also testified that the design of the 1996 4Rminer — which was lighter, had a wider track width,3 and a lower center of gravity than the 1994 4Runner — achieved a substantially increased degree of stability. Toyota conceded that the design changes *206made in the 1996 model could have been incorporated into the design of the 1994 4Runner.
Plaintiff also presented evidence that Toyota marketed the 1994 4Runner for highway driving and commuting, as well as for off-road use. Specifically, plaintiff submitted examples of Toyota’s print and television advertising of the 1994 4Runner, which contained specific representations of 4Runners engaged in sharp steering maneuvers.
At the close of plaintiffs case, Toyota moved for a directed verdict, asserting that plaintiff had failed to prove her negligence claim and had failed to prove a design defect and causation in her strict liability claim. The court granted a directed verdict for Toyota on the negligence claim, rejected the second argument, and deferred ruling on the third. At the close of all evidence, Toyota unsuccessfully renewed its motion for a directed verdict, arguing again that plaintiff had failed to prove design defect and causation. The jury returned a verdict for plaintiff on the strict liability claim, concluding that the design of the 1994 4Runner was defective, and awarded plaintiff $5,400,000 in economic damages and $2,250,000 in noneconomic damages. Toyota subsequently objected to the proposed money judgment on the ground that ORS 18.560 “capped” plaintiffs noneconomic damages at $500,000. The trial court denied that objection and entered a judgment for plaintiff for $7,645,000, plus interest and costs.4
Toyota timely filed a motion for JNOV, or, in the alternative, a new trial. In support of its motion for JNOV, Toyota reiterated its arguments that plaintiff failed to prove design defect — and particularly that the design of the 1996 4Runner represented a safer reasonable alternative design— or that the design of the 1994 4Runner caused her injuries. In support of its motion for a new trial, Toyota pointed, inter alia, to newly discovered evidence of two rollovers involving 1996-model ¿Runners. The trial court did not rule on Toyota’s motion, and it was denied by operation of law. ORCP 63 D.
*207On appeal, Toyota raises four assignments of error. First, Toyota argues that the trial court erred in denying Toyota’s motions for directed verdict and JNOV because plaintiffs evidence of product defect was insufficient to support the verdict. Second, Toyota argues that the trial court abused its discretion in denying its motion for a new trial because Toyota’s “newly discovered evidence” of the two rollovers involving 1996-model 4Runners would so materially undermine plaintiffs theory that the 1996 4Runner represented a safer practicable alternative design as to change the result if a new trial were granted. Third, Toyota argues that the trial court abused its discretion in allowing plaintiff to admit allegedly irrelevant and unfounded expert testimony regarding 15 “similar” rollover accidents, some of which, Toyota contends, was inadmissible hearsay. Fourth, Toyota argues that the trial court erred by not limiting plaintiffs noneconomic damages to $500,000, as required by ORS 18.560(1).
Toyota acknowledges that, for purposes of our review, its fourth assignment of error as to the statutory damage cap is controlled by our decision in Lakin v. Senco Products, Inc., 144 Or App 52, 925 P2d 107 (1996), rev allowed 325 Or 438 (1997). Toyota raises that issue here to preserve it, pending the Supreme Court’s decision in Lakin. Accordingly, and without further discussion, we affirm the trial court’s refusal to “cap” plaintiffs recovery of nonecon-omic damages. We proceed to Toyota’s other assignments of error.
I. THE “CONSUMER EXPECTATION” TEST: CONTENT AND OPERATION
The disposition of each of Toyota’s assignments ultimately depends on a correct understanding of the controlling standard of strict products liability under Oregon law: the “consumer expectation” test. No Oregon appellate decision in more than 30 years has addressed the meaning and operation of that test. Accordingly, to focus the present dispute— and before considering the particulars of Toyota’s arguments — we explore the development, content, and application of the consumer expectation test.
*208 A. The Test Generally
Oregon is one of roughly a dozen jurisdictions that adhere to the consumer expectation test as the standard for determining strict products liability in manufacturing and design defect cases.5 &Although, as explained below, the precise formulation of that test may differ from state-to-state, in each instance the standard was derived from Comment i of section 402A of the Restatement (Second) of Torts (1965), which describes the meaning of an “unreasonably dangerous” product:
“The article sold must be dangerous to an extent beyond that which would be contemplated by the ordinary consumer who purchases it, with the ordinary knowledge common to the community as to its characteristics.”
Thus, to establish a defective design under Comment i, a plaintiff must prove (1) what an ordinary consumer would expect from the allegedly defective product, and (2) that the product failed to meet those expectations. The critical question in this case — and, indeed, in most cases involving application of the consumer expectation test — is how does a plaintiff prove reasonable consumer expectation? That is, what does that term mean and, by extension, what proof of “expectation” is legally sufficient?
*209Outside of Oregon, courts have identified, or developed, two distinct answers to those questions: the “representational” approach and the “consumer risk-utility” approach. Each of those formulations requires different proof.
Under the “representational” approach, product defect is established by proving that the manufacturer specifically represented to the consuming public that the product would be able to perform certain functions, when, in fact, it could not, resulting in the plaintiffs injury. That approach springs from the warranty-related underpinnings of strict products liability — common-law concepts of implied warranties of fitness and merchantability — and it endeavors to protect consumer expectations reasonably flowing from the manufacturer’s representations. Smith v. American Motor Sales Corp., 215 Ill App 3d 951, 955, 576 NE2d 146 (1991), describes that approach:
“[W]e acknowledge the development of a ‘representational theory' of products liability [that] focuses on the images of a product that a manufacturer portrays, in particular, the representations of the product’s safety for all intended uses, and its effect on consumers. In those cases in which a manufacturer explicitly or implicitly represents that its product possesses certain characteristics that it does not actually possess, the representational theory would impose liability on the manufacturer for the resulting injury that reasonable use of the product causes.”6 (Citations omitted; emphasis in original.)
See also Leichtamer v. American Motors Corp., 67 Ohio St 2d 456, 424 NE2d 568, 578 (1981) (“The commercial advertising of a product will be the guiding force upon the expectations of consumers with regard to the safety of a product, and is highly relevant to a formulation of what those expectations might be.”).7
*210Considerations of cost and utility are immaterial to the representational approach. That is, it does not make any difference what it would cost to build a product that could perform as represented — or even whether such a product could ever feasibly be built. The only issue is whether the product performed in conformance with the manufacturer’s representations. Similarly, causation is virtually self-defining: If the plaintiff was injured because the product did not perform as represented, then causation is shown.
In contrast to the “representational” approach, the “consumer risk-utility” approach does turn on considerations of product cost and benefit, as assessed from the perspective of a reasonable consumer. See, e.g., French v. Grove Mfg. Co., 656 F2d 295, 299 (8th Cir 1981) (applying Arkansas law); Potter v. Chicago Pneumatic Tool Co., 241 Conn 199, 694 A2d 1319 (1997); Sumnicht v. Toyota Motor Sales, U.S.A., Inc., 121 Wis 2d 338, 360 NW2d 2 (1985). While the “representational” approach is ultimately grounded in warranty-like principles, the consumer risk-utility approach proceeds from the premise that ordinary consumers reasonably expect products to be designed in the safest feasible and practicable manner. Thus, proof that a product could have feasibly and practicably been designed more safely is proof that the product failed to meet reasonable consumer expectations.
Seattle-First National Bank v. Tabert, 86 Wash 2d 145, 542 P2d 774, 779 (1975), exemplifies the consumer risk-utility analysis:
“[EJvaluation of the product in terms of the reasonable expectations of the ordinary consumer allows the trier of fact to take into account the intrinsic nature of the product. The purchaser of a Volkswagen cannot reasonably expect the same degree of safety as would the buyer of the much more expensive Cadillac. It must be borne in mind that we are dealing with a relative, not an absolute concept.
*211“In determining the reasonable expectations of the ordinary consumer, a number of factors must be considered. The relative cost of the product, the gravity of the potential harm from the claimed defect and the cost and feasibility of eliminating or minimizing the risk may be relevant in a particular case. In other instances, the nature of the product or the nature of the claimed defect may make other factors relevant to the issue.”8
Two other material distinctions between the representational and the consumer risk-utility approach bear emphasis. First, while proof of a safer practicable alternative design is immaterial to the former, it is essential to the latter. That is, under the consumer risk-utility approach, no reasonable consumer would, or could, expect a product that is not feasible. Second, proof of causation is qualitatively different. Unlike the representational approach, where defect (i.e., failure to perform as represented) and causation can conflate, causation under the consumer risk-utility approach involves a comparison between the product as provided and the alleged safer practicable alternative. That is, was the difference in design between the actual product and the alleged alternative a material cause of the plaintiffs injuries — or, conversely (from the defense perspective), would the plaintiff have been injured even if she had been using the alleged “safer” alternative?
Our review of the law of other “consumer expectation” jurisdictions suggests that most jurisdictions have opted for one approach or the other, but not both, perhaps because of those differences. Nevertheless, we do not understand the two to be, necessarily, mutually exclusive. The two approaches promote distinct, but not conflicting, policies; *212they require different, but not inconsistent, proof. We perceive no reason why a plaintiff, with legally sufficient proof, could not proceed under both approaches.
We turn to whether Oregon endorses either approach, or both.
B. The Consumer Expectation Test in Oregon: Rise, Fall, and Revivification
The evolution of product liability law in Oregon has been marked successively by a decade of intensive development, through judicial decisions between the mid-1960s and the mid-1970s;9 legislative codification of section 402A of the Restatement (Second) of Torts (1965) and its Comments a to m in 1979, ORS 30.920; and 20 years of subsequent quiescence, at least in published decisions.
During the initial phase, the Supreme Court first embraced the consumer expectation test, see Heaton v. Ford Motor Co., 248 Or 467, 470, 435 P2d 806 (1967), and then discarded it in favor of a “reasonable manufacturer” test. See Phillips v. Kimwood Machine Co., 269 Or 485, 494, 525 P2d 1033 (1974). In the second phase, the legislature, in direct response to judicial adoption of the “reasonable manufacturer” test, codified Comment i. See ORS 30.920. In the third, we and the Supreme Court have reaffirmed the primacy of the consumer expectation test,10 but neither court has addressed the substance of that test.
Over 30 years, there are but two substantive guideposts as to the content and contours of “consumer expectation” in Oregon. The first is Heaton, the only reported decision that grapples in any meaningful way with a plaintiffs obligations of proof and with the jury’s ultimate function. The second is the enactment of ORS 30.920 in 1979.
In Heaton, the plaintiffs pick-up truck, while traveling on a paved highway at normal speed, ran over a rock that was five or six inches in diameter. The truck continued *213to drive smoothly for another 35 miles and then, for no apparent reason, veered off the road and tipped over. The plaintiff sued the manufacturer, Ford, alleging that the impact with the large rock caused the wheel to fall apart and that the wheel was defectively designed and unreasonably dangerous because it “failed to perform as an ordinary consumer would have expected it to perform.” Heaton, 248 Or at 472. The trial court granted the defendant’s motion for an involuntary non-suit, and the Supreme Court affirmed.
The plaintiffs fundamental contention in Heaton was that the mere fact that the accident had happened was proof that the wheel had failed to perform as an ordinary consumer would have expected it to perform. The plaintiff did not offer any proof regarding the risks and benefits of the allegedly defective wheel design or the availability of a safer alternative design. That is, the plaintiff did not offer the sort of proof necessary to establish product defect under the consumer risk-utility formulation. After observing that “[t]he user has the right to expect a reasonably safe design * * Heaton, 248 Or at 471 (emphasis added), the court concluded that the plaintiffs proof was legally insufficient:
“Where the performance failure occurs under conditions with which the average person has experience, the facts of the accident alone may constitute a sufficient basis for the jury to decide whether the expectations of an ordinary consumer of the product were met. High-speed collisions with large rocks are not so common, however, that the average person would know from personal experience what to expect under the circumstances. Nor does anything in the record cast any light upon this issue. The jury would therefore be unequipped, either by general background or by facts supplied in the record, to decide whether this wheel failed to perform as safely as an ordinary consumer would have expected. To allow the jury to decide purely on its own intuition how strong a truck wheel should be would convert the concept of strict liability into the absolute liability of an insurer.
"* * * * *
“In the defective-product area, courts have already decided how strong products should be: they should be strong enough to perform as the ordinary consumer *214expects. In deciding what the reasonable consumer expects, the jury is not permitted to decide how strong products should be, nor even what consumers should expect, for this would in effect be the same thing. The jury is supposed to determine the basically factual question of what reasonable consumers do expect from the product. Where the jury has no experiential basis for knowing this, the record must supply such a basis. In the absence of either common experience or evidence, any verdict would, in effect, be the jury’s opinion of how strong the product should be. Such an opinion by the jury would be formed without the benefit of data concerning the cost or feasibility of designing and building stronger products. Without reference to the relevant factual data, the jury has no special qualifications for deciding what is reasonable.” Heaton, 248 Or at 473-74 (original emphasis underlined; footnotes and citations omitted; emphasis added).
As we understand. Heaton, and particularly the emphasized language, the court rejected the plaintiffs premise that a jury should be permitted to intuit how a product should perform, or should have been designed, without reference to proof of actual practicability. That is, permitting the imposition of liability based on “pie-in-the-sky” speculation, unanchored to concrete considerations of cost and utility, would transform “strict liability into * * * absolute liability.” 248 Or at 473.
What Heaton did not resolve — at least, not explicitly — was whether a plaintiff who did, in fact, adduce such proof, including proof of a safer practicable alternative, could prevail. Nevertheless, the propriety of such an approach is implicit in the court’s preoccupation with Heaton’s lack of proof of feasibility. Even more, it comports with the court’s overarching observation: “The user has the right to expect a reasonably safe design * * Heaton, 248 Or at 471. At the very least, Heaton does not preclude the “consumer risk-utility” approach.
Heaton also touched, briefly, on the “representational” approach. In rejecting the plaintiffs contention that Ford’s generic advertising of its trucks as “rugged” constituted a legally sufficient basis of consumer expectation, the court observed:
*215“A general impression of durability, however, does not help a customer to form an expectation about the breaking point of a wheel. A ‘rugged’ Ford truck could be expected to negotiate rough terrain, including five-or-six-inch rocks, at appropriate off-the-road speeds, but it does not follow that a user could expect the same thing at highway speeds.” Heaton, 248 Or at 475.
Implicit in that analysis is that sufficiently specific representations of product performance by a manufacturer can be a legally sufficient basis of reasonable consumer expectation. See id. at 471 (“Unreasonable, in this context, means dangerous to an extent beyond that which is contemplated by the ordinary purchaser. Such a definition is conceptually related to the traditional warranty of merchantable quality in the law of sales.”) (citation omitted).
Although the Supreme Court applied the consumer expectation test without elaboration in post -Heaton cases,11 it ultimately discarded the consumer-oriented test for the “reasonable manufacturer” test. Under that test, as announced in Phillips:
“The question of whether the design is unreasonably dangerous can be determined only by taking into consideration the surrounding circumstances and knowledge at the time the article was sold, and determining therefrom whether a reasonably prudent manufacturer would have so designed and sold the article in question had he known of the risk involved which injured plaintiff.” 269 Or at 494 (emphasis added).
Thus, the “reasonable manufacturer” test had two essential factors. First, it shifted the orientation of the inquiry from the consumer’s perspective to the manufacturer’s. Second, it imputed prescience of the product’s risks, and consequent injuries, to the manufacturer.
Phillips and its companion case, Roach v. Kononen/Ford Motor Co., 269 Or 457, 525 P2d 125 (1974), also identified seven factors to be considered in applying the “reasonable manufacturer” test:
*216“(1) The usefulness and desirability of the product— its utility to the user and to the public as a whole.
“(2) The safety aspects of the product — the likelihood that it will cause injury, and the probable seriousness of the injury.
“(3) The availability of a substitute product which would meet the same need and not be unsafe.
“(4) The manufacturer’s ability to eliminate the unsafe character of the product without impairing its usefulness or making it too expensive to maintain its utility.
“(5) The user’s ability to avoid danger by the exercise of care in the use of the product.
“(6) The user’s anticipated awareness of the dangers inherent in the product and their avoidability, because of general public knowledge of the obvious condition of the product, or of the existence of suitable warnings or instructions.
“(7) The feasibility, on the part of the manufacturer, of spreading the loss by setting the price of the product or carrying liability insurance.” Phillips, 269 Or at 499 n 13 (quoting Roach, 269 Or at 464).
As noted, several of those factors — certainly, the first four-have been incorporated into the “consumer risk-utility” approach as recognized in other consumer expectation jurisdictions. See, e.g., Tabert, 86 Wash 2d at 154. Nothing in those risk-utility factors is unique to the “reasonable manufacturer” test.
In 1979, the Oregon Legislature enacted ORS 30.920, which provides:
“(1) One who sells or leases any product in a defective condition unreasonably dangerous to the user or consumer or to the property of the user or consumer is subject to liability for physical harm or damage to property caused by that condition * * *.
“* * * * *
“(3) It is the intent of the Legislative Assembly that the rule stated in subsections (1) and (2) of this section shall be construed in accordance with Restatement (Second) of Torts section 402A, Comments a to m(1965).” (Emphasis added.)
*217The legislative history of that statute, as recounted in Ewen v. McLean Trucking Co., 300 Or 24, 32, 706 P2d 929 (1985) and Burns v. General Motors Corp., 133 Or App 555, 560-61, 891 P2d 1354 (1995), shows that the legislature intended to reject the “reasonable manufacturer” test and “designate the consumer expectation test as the sole test for a dangerously defective product under Oregon law.” 133 Or App at 561 (emphasis in original). Our review of that history indicates that the source of dissatisfaction with Phillips, and the “reasonable manufacturer” test generally, was the first of the two essential factors identified above — i.e., the shift from a consumer-oriented to a manufacturer-oriented test.12 Nothing in the text of ORS 30.920 or, as nearly as we can ascertain, its legislative history, precludes reference to risk-utility considerations from a consumer’s perspective.13 Rather, the statute merely requires that the Oregon law of strict liability “be construed in accordance with” the Restatement’s comments, including Comment i. ORS 30.920.
From our reading of Heaton and ORS 30.920, we draw two conclusions: (1) Heaton recognized the availability of the “representational” approach to consumer expectation upon proper proof and, at the least, did not foreclose future recognition of a “consumer risk-utility” approach in appropriate cases. (2) ORS 30.920 did not, and does not, codify Heaton in the sense of freezing Heaton’s analysis as the exclusive standard for assessing and proving consumer expectation under Oregon law. Rather, the statute explicitly contemplates judicial development of the law “in accordance with” Comment i. ORS 30.920.
We are, thus, free to recognize the “consumer risk-utility” approach as an alternative, or conjunctive, to the *218“representational” approach implicitly recognized in Heaton. We do so for several reasons. First, the “representational” approach, as analyzed in Heaton and in other jurisdictions, is quite constrictive, requiring a substantial correspondence— albeit not an exact equivalence — between the manufacturer’s representations and the conditions of the product’s nonperformance. If the representational approach were exclusive, in those cases in which the manufacturer made only general representations — or, indeed, no representations at all then the manufacturer could escape strict liability. Thus, to treat that approach as the sole method of proving product defect would substantially compromise the consumer-protective underpinnings of strict liability. Conversely, the premise of the “consumer risk-utility” approach — that consumers can reasonably expect a product to be designed in the safest practicable manner — comports with Oregon’s articulation of those policies: “The user has a right to expect a reasonably safe design * * Heaton, 248 Or at 471. We thus conclude that a plaintiff in a defective design case can, with sufficient proof, proceed under either a “representational” theory, a “consumer risk-utility” theory, or both.
As noted previously, the nature of a plaintiffs proof of defect and causation under the two approaches differs dramatically. Here, although the parties did not so explicitly frame their presentations, plaintiff offered proof corresponding to each theory (e.g., evidence of Toyota’s representations and evidence of the availability of a safer practicable alternative), and Toyota responded in kind. The jury was instructed generally on the consumer expectation test, without distinction between the alternative approaches,14 and returned a general verdict. Thus, we are unable to determine whether the jury’s verdict was based on Toyota’s representations, risk-utility considerations, or both. Given that procedural posture, we assume, without deciding, that appellate *219review of the general verdict should comport with the dictates of Whinston v. Kaiser Foundation Hospital, 309 Or 350, 358-59, 788 P2d 428 (1990) (when multiple specifications of negligence submitted to jury and jury returns general verdict, evidence must be sufficient on every specification). Accordingly, in the discussion of Toyota’s assignments of error which follows, we necessarily consider the sufficiency of plaintiffs proof under both theories.
II. TOYOTA’S MOTIONS FOR DIRECTED VERDICT AND JNOV
Toyota’s first assignment of error asserts that the trial court should have granted its motions for directed verdict and JN OV because plaintiffs proof was deficient in three respects: (1) Plaintiffs evidence of a “safer reasonable alternative design” — and particularly that the 1996 4Runner represented such an alternative — was insufficient. (2) Plaintiffs evidence was insufficient to permit a reasonable jury to conclude that the design of the 1994 4Runner was defective under Oregon’s consumer expectation test. (3) Plaintiffs evidence of causation was insufficient.
In reviewing the trial court’s denial of Toyota’s motions for directed verdict and JNOV, we consider the evidence in the light most favorable to plaintiff and must affirm unless we conclude that there is no evidence from which the jury could have found the facts necessary to support the verdict for plaintiff. Brown v. J. C. Penney Co., 297 Or 695, 705, 688 P2d 811 (1984). We do not weigh the evidence: “The jury weighed the evidence, judged the credibility of the witnesses and resolved all conflicts in the evidence.” Id. at 705. If more than one inference could reasonably be drawn from the evidence, we assume the inference most favorable to the verdict. Id.
A. Safer Practicable Alternative Design
Toyota argues that plaintiff failed to submit evidence from which a jury could find that there was a safer practicable alternative design. In Toyota’s view, Wilson v. Piper Aircraft Corp., 282 Or 61, 577 P2d 1322 (1978), stands for the proposition that a trial court may submit a product design defect case to the jury only if the plaintiff submitted *220sufficient evidence of a feasible, practicable,15 safer alternative design. Id. at 67-68. Plaintiff responds that such proof is not required under Oregon’s consumer expectation test. Plaintiff further contends that, in all events, she adduced sufficient proof of a safer practicable alternative design.
Each party is correct — to an extent. As discussed above, 160 Or App at 210, proof of a safer practicable alternative is immaterial to the “representational” formulation of the consumer expectation test. That formulation is concerned solely with the manufacturer’s representations and consumers’ consequent expectations; alternative designs, whether practicably available or not, are inapposite to that inquiry. Conversely, as explained more fully below, proof of a practicable safer alternative design is integral to the “consumer risk-utility” approach. Reasonable expectation under that formulation is, necessarily, defined by reference to, and circumscribed by, practicability.
The source of the “practicable alternative” analysis in Oregon is Wilson, a case decided under the “reasonable manufacturer” test. There, the plaintiffs, who contended that an aircraft crash was caused by engine failure resulting from carburetor icing, sued the manufacturer, asserting that the aircraft was dangerously defective because a reasonable manufacturer would have designed it with a fuel-injected engine that was not susceptible to engine icing. The jury returned a verdict for the plaintiffs, but the Supreme Court reversed, holding that the plaintiffs’ evidence of a safer reasonable alternative design was insufficient. Wilson, 282 Or at 66-71. The court acknowledged that the plaintiffs had offered proof that a fuel-injected engine was both safer and *221technically feasible, but it emphasized that an alternative design is not reasonable unless it is also “practicable.” Id. at 68. Thus, the court held, in deciding whether to submit a design defect case to the jury, courts should consider “whether the proposed alternative design has been shown to be practicable” Id. (emphasis added). The court stated that, in complex cases where “practicability cannot be properly weighed solely on the basis of inference and common knowledge[,] * * *[i]t is not proper to submit [the case] to the jury unless the court is satisfied that there is evidence from which the juxy could find the suggested alternatives are not only technically feasible but also practicable in terms of cost and the over-all design and operation of the product.” Id. at 69. Because the plaintiffs failed to adduce any evidence regarding the “effect the substitution of a fuel injected engine * * * would have had upon the airplane’s cost, economy of operation, maintenance requirements, over-all performance, or safety in respects other than susceptibility to icing,” the plaintiffs’ evidence was insufficient. Id. at 70.
As noted, Wilson was a “reasonable manufacturer” case. Nevertheless, its underlying concerns properly inform the “consumer risk-utility” approach. No reasonable consumer could expect a manufacturer to provide an ostensibly “safer,” but actually impracticable, product. Moreover, the requirement of proof of practicability addresses the concerns anticipated in Heaton: Without such evidence, the jury’s inquiry is potentially open-ended and speculative. “Reasonableness,” in this context, must be limited by practicability. We thus conclude that a plaintiff who proceeds under the consumer risk-utility approach must present evidence from which the jury could find the availability of a safer practicable alternative design. See, e.g., Hoyt v. Vitek, Inc., 134 Or App 271, 282, 894 P2d 1225 (1995) (affirming dismissal of design defect claim because utility of product was great and the plaintiff failed to prove the availability of a safer practicable alternative).
Here, plaintiff argued that the 1996 4Runner design — with a lower center of gravity and a wider wheel track — was substantially safer than the 1994 4Runner design in all respects and that it was both feasible and practicable for Toyota to incorporate those design changes in *2221994.16 Toyota contends that plaintiff, like the plaintiffs in Wilson, failed to present any evidence about the practicability of implementing the 1996 4Runner design in 1994. Specifically, Toyota contends that plaintiff failed to submit evidence that (a) the alternative design would have preserved the specialized functions that sport utility vehicles are intended to serve; (b) the alternative design could have been implemented in the 1994 4Runner without impairing the vehicle’s utility or increasing its cost to consumers; and (c) the alternative design could have been implemented in the 1994 4Runner without increasing other safety risks.
We disagree. Plaintiff presented sufficient evidence from which the jury could conclude that, at the time Toyota manufactured and marketed the 1994 4Runner, the design of the 1996 4Runner was safer, technically feasible, and “practicable in terms of cost and the over-all design and operation of the product.” Wilson, 282 Or at 69. That evidence included: (1) Testimony by Tamny, the accident reconstruction expert, that the 1996 4Runner’s lower center of gravity and wider wheel track made it “very unlikely” that a 1996 4Runner would have rolled over in the circumstances in which plaintiff was injured. (2) Evidence that Toyota actually produced and marketed the alternative design in 1996, thus demonstrating that the safer design was “practicable” in terms of the cost, over-all design, and operation of the product without compromising its fundamental function. (3) Testimony by Robertson and Tamny that the design modifications to the 1996 4Run-ner did not interfere with its sports utility functions in any way, nor did they reduce the vehicle’s safety in nonrollover accidents. We thus conclude that plaintiff presented sufficient evidence from which a jury could find that the 1996 4Runner design was a safer practicable design available at the time that Toyota manufactured the 1994 4Runner.
*223 B. Sufficiency of Proof of Consumer Expectation
Toyota next argues, and plaintiff disputes, that plaintiff failed to present legally sufficient proof of consumer expectations of the performance of the 1994 4Runner. We disagree, and conclude that plaintiff submitted sufficient evidence to prove that the 1994 Runner failed to meet consumer expectations under both the consumer risk-utility approach and the representational approach.
Our analysis under the consumer risk-utility formulation is short and simple: Plaintiffs evidence as to the availability and practicability of the 1996 4Runner design at the time that Toyota designed and manufactured the subject 1994 4Runner was sufficient to create a jury question as to whether, given considerations of cost, impact on overall utility, and safety, a reasonable consumer would have expected Toyota to have marketed that alternative.
Whether plaintiffs proof was sufficient to support a verdict on the representational theory entails a more extended discussion. At the outset, we reject Toyota’s argument, invoking Heaton, 248 Or at 474, that plaintiffs proof was insufficient in that she
“submitted no poll, no study, no psychological profile of the average person’s expectation, and no other evidence about the ordinary consumer’s expectations regarding rollovers generally or rollovers in circumstances as severe as those that occurred in this accident. Plaintiff also submitted no statement ever published to consumers by any vehicle manufacturer, any government official or agency, any safety researcher, any automotive analyst, any scientist, any engineer, or any other person on the face of the earth, telling consumers that any utility vehicle ever made could not be made to roll over on flat, dry pavement or would not be likely to roll over if subjected to a series of extreme, abrupt steering maneuvers.”
Read as a whole, Heaton does not require such proof. Rather, as we have emphasized, Heaton, by necessary implication, recognized that proof of consumer expectation can be based on manufacturer representations so long as they correspond sufficiently to the circumstances of the accident. 160 Or App at 214.
*224As noted previously, the sufficiency of a plaintiffs proof under the representational approach depends on the degree of correspondence or similarity between the manufacturer’s representations of product performance and the circumstances of the product’s alleged nonperformance. See 160 Or App at 218. On one hand, as Heaton recognized, to impose liability based on generic representations could transform strict liability into absolute liability. 248 Or at 473. A representation that a car is “safe” does not support a reasonable expectation that the vehicle can negotiate and withstand all foreseeable road hazards any more than does a representation that a pick-up is “rugged.” Conversely, to restrict the representational approach to those circumstances where the manufacturer’s representations correspond exactly with the circumstances of the accident would effectively insulate manufacturers from liability. The circumstances of accidents are so variable that a manufacturer will rarely, if ever, represent that the product will perform safely under precisely those conditions.
In truth, it is difficult, if not impossible, to define a principled point of evidentiary sufficiency on that continuum. We also acknowledge that a standard of “substantial similarity” may seem to be so indefinite as to describe a result rather than an analysis. Nevertheless, we believe that is the correct standard: A plaintiffs proof under the representational theory is legally sufficient if a reasonable juror could find that there is such a relationship between the manufacturer’s representations and the circumstances of the product’s nonperformance that a reasonable consumer would have believed that the product could perform safely under those circumstances.
Here, Toyota did not, of course, explicitly represent that the 1994 4Runner would not roll over under the exact circumstances prescribed here: A vehicle traveling at legal speed, on a dry flat highway, swerves suddenly to avoid an oncoming vehicle in its lane of travel and then, immediately, twice sharply countersteers to stay on the roadway. Rather, plaintiffs proof turned on Toyota’s more generalized representations showing the vehicle’s maneuverability. Although the question is close, we conclude that that evidence, as described below, was sufficient.
*225Plaintiff submitted evidence that, although the 4Runner was originally designed for off-road driving, Toyota marketed the 1994 4Runner for highway driving. Don Cecconi, Toyota’s national merchandising manager for the United States, testified that, in 1990, Toyota introduced a new generation of the 4Runner that was designed to be more comfortable and appealing for everyday use than the previous model. The reason for that shift was that Toyota had become aware that, on average, fewer than three percent of 4Runner buyers ever used their 4Runners for the off-road purposes for which they were designed. Toyota began to gear its advertising of the 4Runner to appeal to the rapidly growing market of wealthy, older drivers who use SUVs primarily as commuting and passenger cars. Toyota’s commercial advertising and brochures for the 1994 4Runner showed the 4Runner being used for long-distance family trips, commuting, and highway driving. Cecconi testified that Toyota was aware that consumers thought of the vehicle’s height and weight as safety features but did not make any effort, in its promotional materials, to communicate to consumers that those very features created a high risk of rollover. He testified that, to the contrary, Toyota promoted the 4Runner as maintaining good control in a variety of driving environments, including during sharp turns and evasive maneuvers, albeit less extreme than those which occurred in plaintiffs accident.
More specifically, plaintiff also submitted examples of Toyota’s print and television advertising of the 1994-model 4Runner containing specific representations of 4Runners engaged in sharp steering maneuvers. The “anti-lock brake system” section of the 1994 4Runner marketing brochure contains a computer graphic illustration of a 4Runner traveling on a two-lane road. Dotted lines indicate that the 4Run-ner’s successful course of travel involves swerving left to avoid an obstacle in the right lane of travel and then swerving back to the right to avoid an obstacle in the left lane of travel.17 Another print ad for the 1994-model 4Runner contains a similár icon depicting a 4Runner traveling on a paved, *226two-lane road. The illustration indicates that the 4Runner successfully travels across a small body of water or ice in the right lane, and then, just on the other side of that obstacle, the 4Runner veers sharply into the left lane to avoid debris from a landslide.
Toyota’s television advertising of 1994-model 4Run-ners also contained numerous images of 4Runners engaging in sharp turns and maneuvers. For example, in one ad, the narrator describes 4Runners as if they are an animal species. As the narrator states that 4Runners “adapt[ ] quickly to changing environments,” the screen shows a 4Runner turning sharply to the left and then to the right to avoid trees in its path. The ad then shows two children in the back seat of the 4Runner, states that the 4Runner’s "primary instinct is to care for its young,” and closes with a shot of a 4Runner rounding a sharp turn on a paved road. Another ad shows a 4Runner traveling on a paved road with a man driving, a woman in the front passenger seat, and two children in the back seat. The ad cuts back and forth between images of people enjoying recreational activities and images of 4Runners making sharp turns.18
*227We agree with plaintiff that her evidence showed that Toyota specifically marketed the 1994 4Runner as an appropriate vehicle for highway driving and specifically depicted 4Runners engaging in sharp turns and evasive maneuvers. The jury could conclude that, given Toyota’s representations, an ordinary consumer would reasonably expect a 1994 4Runner traveling at legal speed not to roll over following foreseeable evasive maneuvers, such as three sharp turns on a flat, dry, paved highway.
Toyota argues, nevertheless, that plaintiffs evidence of Toyota’s representations was insufficient to survive the directed verdict motion because Toyota posted a notice inside the vehicle, specifically warning of the possibility of a rollover in the event of sharp turns or other abrupt maneuvers, and that warning was repeated in the 4Runner’s owner’s manual. Toyota asserts that, given the evidence of those warnings, no reasonable jury could return a verdict in plaintiffs favor— i.e., the jury could only conclude that a reasonable consumer would expect and anticipate that a rollover could occur in the circumstances presented here and, thus, the consumer expectation test could not be satisfied.
Federal law requires all sport utility vehicles to have a permanent rollover warning both in the passenger compartment and in the owner’s manual. 49 CFR § 575.105. Accordingly, the following warning was posted on the visor on the passenger side of Sanders’ 1994 4Runner:
“CAUTION. This vehicle will handle and maneuver differently from an ordinary passenger car, in driving conditions which may occur on streets and highways and off road. As with other vehicles of this type, if you make sharp turns or abrupt maneuvers, the vehicle may rollover or may go out of control and crash. You should read the on-pavement and off-road driving tips in the Owner’s Manual, and WEAR YOUR SEATBELTS AT ALL TIMES.”19 (Emphasis added.)
*228Toyota argues that the visor warning made the risk of rollover so apparent and obvious that the only expectation that the ordinary consumer of a 1994 4Runner could reasonably have had was that it would roll over following abrupt steering and sharp turns. Thus, in Toyota’s view, although the vehicle was unstable and prone to rollover following foreseeable evasive maneuvers, the visor warning necessarily precluded imposition of strict liability.
We disagree. Toyota’s marketing representations of the 1994 4Runner making sharp turns and maneuvers justified a reasonable expectation that that vehicle could perform such maneuvers. The question of whether, or to what extent, the visor warning — which, after all, was yet another representation by Toyota — effectively subverted or negated such an expectation in the mind of an ordinary consumer was a matter for the jury to decide.20
We thus conclude that plaintiffs evidence of Toyota’s representations was sufficient to support liability under the representational approach. Conversely, Toyota’s evidence, including the visor warning, although probative of consumer expectations, was not so preclusive as to permit only one result on that issue. Consequently, the court did not err in denying a directed verdict on the basis of the alleged insufficiency of plaintiffs proof of reasonable consumer expectation.
*229 C. Causation
Finally, with respect to the denial of the directed verdict, Toyota argues that plaintiff failed, as a matter of law, to prove that the allegedly defective design of the 1994 4Runner caused her injuries. Specifically, Toyota argues that plaintiff was required to prove that a SUV with a safer reasonable alternative design, viz., the 1996 4Runner, “would not have rolled over under circumstances as severe as those in this accident.” Toyota argues that plaintiff’s expert testimony that the 1996 4Runner was “unlikely” to roll over in circumstances like those in this accident was not sufficiently definitive.
Plaintiff responds that Oregon law does not require explicit “direct” expert testimony that, if the product had been properly designed, the accident would not have occurred. Rather, plaintiff reasons, the necessary showing of defect-related causation can be made through a combination of expert testimony and the inferences that the jury could properly draw from that testimony.
As framed by the parties, the issue of the sufficiency of plaintiff’s proof of causation thus relates solely to the consumer risk-utility formulation. As noted, causation under the representational approach is established merely by showing that the plaintiff was injured because the product did not perform as represented. See 160 Or App at 211. Any evidence of an alleged safer alternative’s putative performance is immaterial to that inquiry.
Plaintiff’s proof was sufficient to establish causation under the consumer risk-utility approach. See Austria v. Bike Athletic Co., 107 Or App 57, 60,810 P2d 1312 (1991). In Austria, a high school football player suffered severe physical and substantial mental injuries after he sustained a blow to his head during football practice. His parents brought suit against the designer and manufacturer of the football helmet, alleging that the defective design of the helmet caused their son’s injuries. The jury returned a verdict for the plaintiffs. On appeal, the defendants argued that the plaintiffs’ evidence was insufficient to prove causation because it did not establish: “(1) the amount of force received by Richard as *230a result of the blow; (2) the amount of force that he would have received had he been wearing a helmet of alternative design; and (3) the amount of force required to cause the type of head injury that he suffered.” Id. at 59-60. We rejected those arguments:
“We do not agree with defendants’ premise that that particular evidence is necessary to prove causation. They point to no authority to support their argument that there must be direct evidence of each or any of those factors in order to prove causation in a case such as this. There is no reason why the sufficiency of an impact to do damage, or the insufficiency of a product’s resistance to it, cannot be inferred from other probative evidence.
“There was evidence that, when combined with the inferences drawn therefrom, was sufficient to allow the jury to conclude that defendants’ defective helmet caused the injury. * * *
* * * *
“In sum, there was evidence about how and why Richard was struck, what a football helmet is supposed to do to reduce the consequences of a collision, that the hema-toma resulted from precisely the kind of trauma that helmets are designed to prevent, that a properly designed helmet would have greatly reduced the likelihood of a hematoma and that defendants’ helmet was not adequately designed to reduce that likelihood.” 107 Or App at 61.
Austria is controlling here. Plaintiff adduced the following evidence from which the jury could infer that the 1996 4Runner design would have either prevented or “greatly reduced the likelihood,” id., of a rollover in this accident: Robertson, the statistician, testified that a five-state survey of rollover accidents21 established that a vehicle’s “static stability” — the number resulting from dividing half the track width by the height of the center of gravity — determines its propensity to rollover. According to Robertson, the study revealed that a vehicle with a static stability approaching or exceeding 1.2 is unlikely to roll over. The pre-1996 4Runner *231had a static stability of only 1.08 and the second highest rollover rate. The 1996 4Runner design had a static stability of 1.17. Robertson concluded:
“I believe [the 1994 4Runner] is unreasonably dangerous because it has a very high rollover rate commensurate with its low stability, and it could have been modified by * * * eight inches [in track width], and virtually eliminated that difference in the rollover rate * *
Tamny, an engineer who had investigated 60-65 rollover accidents, testified that vehicles with a “dynamic stability” — the static stability, combined with other factors such as tire pressure and size and suspension — of .9 or greater are unlikely to roll over. According to Tamny’s analysis, the 1994 4Runner has a dynamic stability of .8, and the 1996 4Runner has a dynamic stability of .94. Tamny stated that, although the 14 percent difference in dynamic stability might seem insignificant, it put the 1996 design “over the hill” to where it is stable enough to slide to a stop instead of rolling. Additionally, in rollover testing results submitted as evidence, the pre-1996 4Runner rolled over as a result of hard steering alone, but the 1996 4Runner did not. There was no evidence at trial — either from test results or real-world accidents — of a 1996 4Runner rolling over in circumstances like those of this accident.
In sum, the jury could infer that had plaintiff been a passenger in a vehicle with the design of the 1996 4Runner, the increased stability of the 1996 design would have either prevented or “greatly reduced the likelihood,” Austria, 107 Or App at 61, of a rollover.
Toyota attempts to distinguish this case from Austria on the ground that, in its view, the jury in Austria could infer causation from its common knowledge,
“[T]hose [football] contacts were well within the realm of the average person’s common understanding and experience. The average person probably has seen dozens,'if not hundreds or thousands, of tackles on television or in person in pee-wee football, high school football, college football, and professional football, and has attained a reasonably well-informed expectation of what kinds of blows football *232helmets generally are able to protect against. In sharp contrast, this case involves complex technology and engineering concepts (e.g., vehicle width-to-height ratios, steering systems, and lateral g forces, to name but a few) and complex accident circumstances as to which the average person has almost no knowledge or experience.”
That argument is unavailing. Nothing in Austria suggests that our analysis applies only to cases involving “accident facts familiar to the average person.” Moreover, as plaintiff suggests, “the physics of impact diffusion in football helmets are probably as technical as automobile designs.”22
The trial court correctly denied Toyota’s motion for a directed verdict.
III. OTHER SIMILAR INCIDENTS
We turn to Toyota’s third assignment of error, which challenges the trial court’s admission of evidence regarding other rollover accidents involving Toyota 4Runners.23 Plaintiff offered the testimony of Jerry Wallingford, an engineer and accident reconstructionist, that he had reviewed approximately 35 rollovers involving pre-1996 4Runners and had evaluated the cause of the rollovers based on a variety of factors. Wallingford concluded that 20 of the rollovers were substantially similar to the McCathem rollover; the court limited his testimony to address only 15 of those 20 “substantially similar” accidents. Because Wallingford had been to the scene of only one of those 15 accidents, his expert opinion that the accidents were substantially similar to the McCathem rollover was based on factual information extracted from other lawsuits, depositions of drivers and police officers, police reports, police photographs, and photographs by experts. In addition to Wallingford’s testimony, plaintiff offered, and the trial court admitted, videotaped testimony by five witnesses, including two people who had been *233in 4Runner rollovers and a police officer who had investigated a 4Runner rollover.
Toyota argues that the trial court abused its discretion by admitting plaintiff’s evidence of the 15 “similar incidents.” We review the trial court’s rulings pertaining to “substantial similarity’ for abuse of discretion. Davis v. Homasote Company, 281 Or 383, 387-88, 574 P2d 1116 (1978); Lakin, 144 Or App at 61.
Toyota first argues that plaintiff’s “other similar incidents” evidence is irrelevant. However, Toyota neither assigns error to a ruling by the trial court regarding the relevancy of Wallingford’s testimony nor directs us to the portion of the record in which Toyota preserved a relevancy challenge to the “other similar incidents” evidence. Consequently, we do not consider Toyota’s argument that plaintiff’s evidence regarding other similar incidents was irrelevant. See ORAP 5.45(4).
Toyota next argues that, even if the “similar incident” evidence is relevant, there was no foundation for that evidence because plaintiff failed to show that the other accidents were substantially similar to the McCathem rollover. We have reviewed the record, and we conclude that, although Toyota properly preserved a lack of foundation argument with respect to all 15 accidents,24 the trial court did not abuse its discretion by concluding that the accidents were “substantially similar” to the McCathern accident. See Lakin, 144 Or App at 61 (“Trial courts have broad discretion in assessing ‘substantial similarity * * *.”).
Toyota argues that, in order to establish that the other rollover incidents were “substantially similar” to the McCathem rollover, plaintiff was required — but failed — to show that the other incidents and the McCathem accident *234had all of the following factors in common: vehicle model and year, tire size, suspension and modification, roadway surface, vehicle speed, cargo, steering input, driver impairment, tripping mechanisms, and contact with another object. Toyota points out that plaintiffs expert, Wallingford, considered other incidents to be substantially similar to the McCathern rollover if they had only three of the above factors in common: driver impairment, tripping mechanisms, and contact with another object. In Toyota’s view, “such a limited comparison cannot meet the requirement of proving substantial similarity.” We disagree.
“Only substantial similarity, not complete identity of circumstances, is required.” Rader v. Gibbons and Reed Company, 261 Or 354, 360, 494 P2d 412 (1972). Oberg v. Honda Motor Co. 25 exemplifies that distinction. There, the Supreme Court held that evidence of other ATV accidents was “substantially similar” to the Honda ATV accident in that case even though the “other incident” evidence involved accidents that occurred under different conditions and involved ATVs made by manufacturers other than Honda. 316 Or at 268-69; see also Lakin, 144 Or App at 61-62.
A protracted published discussion of the similarity of each of the 15 incidents here would serve little purpose. We will, however, address one specific incident, the Fox rollover, as being generally illustrative of our overall analysis. The Fox rollover involved a 1991 4Runner. Fox testified, by videotape, that when he made a U-turn while driving 20 miles per hour on dry, packed dirt with some loose gravel, the 1991 4Runner he was driving rolled over unexpectedly. The trial court held a hearing to consider whether the Fox incident was “substantially similar” to the McCathern rollover. Wal-lingford testified that, in his opinion, the Fox and McCathern rollovers were substantially similar because rollover testing demonstrated that it was more difficult to roll a vehicle on a dirt or gravel surface; both accidents involved the same model of 4Runner;26 and there was no evidence of driver *235impairment, tripping mechanisms, or braking in either accident. According to Wallingford, the fact that both rollovers occurred as a result of hard steering alone made them “substantially similar.” The trial court admitted the evidence. We conclude that, given Wallingford’s testimony, the trial court’s determination of substantial similarity was not an abuse of discretion. We similarly conclude, without further elaboration, that the trial court properly admitted evidence of the remaining incidents.
Finally, Toyota argues that, even if the court did not err in allowing Wallingford to testify that, in his opinion, the 15 other incidents were substantially similar, Wallingford’s testimony as to the factual information derived from documents on which he relied to form his opinion was inadmissible hearsay. See State v. Knepper, 62 Or App 623, 626, 661 P2d 560 (1983) (“OEC 703 does not authorize an expert witness to tell the jury the inadmissible details of the basis of his opinion.”); Mission Ins. Co. v. Wallace Security Agy., Inc., 84 Or App 525, 528, 734 P2d 405 (1987) (although OEC 703 “recognize[s] that experts often rely on facts and data supplied by third parties * * * that does not give carte blanche to admitting otherwise inadmissible hearsay”) (citation omitted).
Whatever the abstract merits of Toyota’s hearsay argument, we decline to address it because the following sequence of events at trial demonstrates that Toyota waived any error: On direct examination by plaintiff, Wallingford recited facts from the police reports and other hearsay documents underlying his opinion. Plaintiff never offered any of those documents into evidence. However, Toyota objected to Wallingford’s recitation of details of the other incidents as inadmissible hearsay. On cross-examination, Toyota referred to portions of several police reports from Wallingford’s file that, Toyota asserted, included accident descriptions contrary to Wallingford’s conclusions. At the close of Walling-ford’s testimony, Toyota stipulated that the entire contents of Wallingford’s file — a banker’s box full of reports, depositions, and photographs — be admitted into evidence. On appeal, Toyota now explains the motive behind that stipulation by quoting its counsel at trial:
*236“ ‘[Wallingford] basically [has] gotten in all sorts of hearsay before the jury in terms of what happened in all these accidents * * * [and] * * * to say we can’t bring in contradictory evidence from this file that he considered in reaching those opinions * * * prevents any meaningful confrontation as to the substance of what he’s already testified to.’ ” (Emphasis added.)
There is no indication that Toyota asked for a limiting instruction with regard to the jury’s consideration of the documents in Wallingford’s file. That is, although Toyota, and not plaintiff, offered Wallingford’s underlying documentation en masse into evidence, Toyota never requested that the jury be instructed that those documents were to be considered solely for nonsubstantive/impeachment purposes.
Given that sequence of events, Toyota cannot now be heard to complain that the facts in the documents to which Wallingford referred to diming his testimony — documents that Toyota stipulated to as evidence — are inadmissible hearsay. Toyota made a strategic decision to sacrifice preservation of its hearsay objection in order to get the substance of the documents underlying Wallingford’s opinion before the jury. See generally State ex rel Juv. Dept. v. Cook, 325 Or 1, 932 P2d 547 (1997); State v. Lopez, 147 Or App 314, 936 P2d 386, rev den 326 Or 58 (1997) (both applying principle in analogous criminal context). The trial court did not err in admitting the “similar incidents” evidence.
IV. NEWLY DISCOVERED EVIDENCE
We turn finally to Toyota’s second assignment of error, which asserts that the trial court erred in denying Toyota’s motion for a new trial. ORCP 64 B(4). That motion was based, inter alia, on “newly discovered evidence” that, on April 11, 1997, the day after the jury returned its verdict, a 1997 4Runner27 was involved in a rollover accident in California. Toyota learned of that accident on April 16, 1997. Subsequently, in its reply memorandum supporting its motion for new trial, Toyota also relied on “newly discovered evidence” of a second accident, which occurred in October 1996 *237and involved the rollover of a 1996 4Runner in Texas. Toyota did not learn of that accident until it was served with process on May 27,1997, after filing its new trial motion. The parties agree that, to warrant a new trial, newly discovered evidence must meet all six factors in the newly discovered evidence test outlined by the Supreme Court in Oberg. 316 Or at 272. They disagree, however, about whether the “newly discovered evidence” of either the 1996 or 1997 4Runner rollovers satisfies the Oberg test. We review the trial comb’s denial of Toyota’s motion for abuse of discretion. Id. at 273.
Plaintiff argues that, as a threshold matter, Toyota’s evidence of the California accident is not “newly discovered evidence” because it did not occur until after the jury returned its verdict. Whether evidence that did not exist at the time of the trial qualifies as “newly discovered evidence” for the purpose of a new trial under ORCP 64 B(4) is a question of first impression. Nevertheless, the federal courts, in applying analogous provisions of the Federal Rules of Civil Procedure, FRCP 59 and 60(b)(2), have held that newly discovered evidence must be of facts, in existence at the time of the trial. We adopt that same view.
ORCP 64 B(4) specifies as one basis for a new trial:
“Newly discovered evidence, material for the party making the application, which such party could not with reasonable diligence have discovered and produced at trial.”
The language of ORCP 64 B(4) highlights that the purpose of the rule is to allow parties a new trial when, through no lack of diligence on the part of the aggrieved party to discover and produce all relevant evidence, the evidence considered by the jury at trial did not accurately reflect the relevant facts existing at the time of trial. The verb “to discover” is defined as “to make known something secret, hidden, unknown or previously unnoticed” or “to obtain for the first time sight or knowledge of.” Webster’s Third New Int’l Dictionary, 647 (unabridged ed 1993). Thus, the phrase “newly discovered evidence” implies that the evidence existed but was not known or knowable at the time of trial. Moreover, the due diligence requirement of ORCP 64 B would be a non sequitur with respect to newly existing evidence — i.e., even with due *238diligence, evidence nonexistent at trial would surely be undiscoverable for production at trial.
As noted, that construction comports with the approach taken by federal courts interpreting FRCP 60(b)(2), which provides that newly discovered evidence may be grounds for a new trial (FRCP 59) or relief from judgment (FRCP 60) if the evidence “by due diligence could not have been discovered.” See Wright, Miller & Kane, Federal Practice and Procedure, 302 § 2859 (2d ed 1995) (“Newly discovered evidence must be of facts existing at the time of the trial.”). Underlying the federal decisions is a principle of finality. See, e.g., Campbell v. American Foreign S. S. Corp., 116 F2d 926, 928 (1941) (noting that if a motion for a new trial could be granted on the basis of the post-trial occurrence of facts contradicting those demonstrated at trial, then “the litigation would never come to an end”). The same considerations underlie the Oregon rules. See Marshall v. Martinson, 264 Or 470, 477, 506 P2d 172 (1973) (“Efficient judicial administration dictates that motions for new trials because of newly discovered evidence be granted sparingly. Otherwise, there would never be any finality to judicial proceedings”) (emphasis added).
Toyota’s reading of ORCP 64 B(4) and, of necessity, the parallel “newly discovered evidence” provision of ORCP 71 B(1), would substantially subvert that principle of finality. Under ORCP 71 B(1), parties may move to set aside judgments based on, inter alia, “newly discovered evidence,” for up to one year after the entry of judgment.28 If Toyota is correct, then (hypothetically) if the jury had returned a defense verdict in this case and, in April 1998, 11 months after the entry of judgment, a dozen rollover accidents involving 1994 4Runners had occurred under circumstances identical to this accident, then plaintiff could have successfully moved to vacate the judgment and obtain a new trial based on circumstances that did not exist at the time of trial. The text of ORCP 64 B(4) (and ORCP 71 B(1)) does not compel such an open-ended reading of “newly discovered evidence” and, *239given practical and prudential principles of finality, we decline to so expansively construe the rule. The trial court did not err in denying Toyota’s motion for a new trial on the basis of evidence of the April 11,1997, rollover.
Our disposition with respect to the evidence of the 1997 California rollover does not, however, apply to Toyota’s evidence of the October 1996 Texas rollover. In assessing whether the trial court abused its “sound discretion” in denying a new trial based on that evidence, we apply Oberg’s cumulative six-part test:
“ ‘(1) [The alleged newly discovered evidence] must be such as will probably change the result if a new trial is granted; (2) it must have been discovered since the trial; (3) it must be such as, with due diligence, could not have been discovered before the trial; (4) it must be material to the issue; (5) it must not be merely cumulative; (6) it must not be merely impeaching or contradicting of former evidence.’ ” 316 Or at 272 (quoting State v. Davis, 192 Or 575, 579, 235 P2d 761 (1951) (emphasis added)).
In Davis, the Supreme Court explained that, to demonstrate that newly discovered evidence will “probably” change the result of a new trial, the movant bears the burden of submitting “supporting affidavits [that] show a state of undisputed facts which would probably lead an ordinary reasonable person to a different conclusion from that arrived at by the jury * * * ” 192 Or at 579.
We conclude that the first factor is dispositive — that is, that Toyota failed to satisfy that requirement — and, thus, we do not address the others. We agree with the trial court (and Toyota) that, given that the essence of plaintiffs risk-utility theory of consumer expectation was that the 1996 4Runner was a safer practicable alternative design that would not have rolled over in this accident, evidence of an accident that was substantially similar to plaintiffs and that involved a 1996 4Runner “could definitely affect the outcome of a new trial.” The question thus narrows to whether the trial court abused its discretion in concluding that Toyota failed to demonstrate that the Texas accident was so similar to the one in which plaintiff was injured that evidence of that accident would “probably [have] change[d] the jury’s verdict.”
*240Toyota submitted two affidavits regarding the Texas accident: one affidavit from a Toyota lawyer reciting the contents of a petition served on Toyota in connection with the accident, with the petition and a police report attached as exhibits, and the second from an investigating police officer, describing the size of the tires on the 1996 4Runner in the Texas accident. The affidavits establish that the 1996 4Run-ner in the Texas accident was equipped with tires of the same size as those on the 1994 4Runner in this case. The police report indicates that the 1996 4Runner was traveling on a two-lane highway when the driver drove off the left side of the road, corrected to the right and then left, and then rolled over four times. The petition alleges that the 1996 4Runner went out of control and “rolled over on a flat level dry surface.” Neither the affidavits nor the exhibits disclose any other facts regarding the accident.
At the hearing on the motion for a new trial, plaintiff argued that Toyota failed to develop the facts of the Texas accident sufficiently to establish that the Texas accident was so similar to the McCathem accident that evidence of it would “probably affect” the outcome of a new trial. Specifically, plaintiff highlighted that Toyota had failed to develop evidence regarding how fast the Texas 4Runner was traveling, whether the driver was sober, whether there was braking involved in the prerollover maneuvers, whether the vehicle had left the pavement, and whether a rim trip had caused the accident.29 Plaintiff also offered into evidence a photograph of the Texas accident scene. That photograph showed a mark that appeared to be a gouge in the pavement. Plaintiff asserted that a gouge would indicate that the accident was caused by a rim trip and not by tire friction forces, as in this case.
We conclude that the trial court did not abuse its discretion by denying Toyota’s motion for a new trial. We note that the only undisputed facts that Toyota established with *241respect to the Texas accident were that a 1996 4Runner with the same size tires as Sanders’ vehicle rolled over on a flat, dry, paved highway. Toyota was free to develop the facts to establish that the Texas accident had more than those facts in common with the McCathem rollover, but it failed to do so. Although a reasonable trial judge could have granted the motion based on Toyota’s showing, given the deficiencies and questions that plaintiff raised, it was not unreasonable for the court to conclude otherwise. There was no abuse of discretion.
Affirmed.