Defendants appeal from an order granting plaintiff s motion for a new trial after the jury returned a defense verdict in this personal injury case. ORCP 64 B. Plaintiff cross-appeals, assigning error to several of the trial court’s eviden-tiary rulings. We reverse on appeal; on cross-appeal, we affirm without discussion.

While plaintiff was snowboarding at a ski area that defendants operate on Mt. Hood, he snowboarded into a wooded area between two ski runs, Whoopee and Reservoir Hill. In doing so, he fell into Reservoir Creek, which runs through the wooded area, and as a result, he was seriously injured. On the Reservoir Hill side, the wooded area was entirely roped off at the time of the accident as a warning to skiers, but on the Whoopee side, there were places where there were no ropes or other warnings. Plaintiff testified that he entered the wooded area at a nonroped portion of the Whoopee side. In his complaint, he alleged that defendants were negligent in, among other things, “failing to install and maintain barriers and/or warnings sufficient to prevent skiers or snow boarders from accidentally skiing or snow boarding into the creek bed from the west or south side.” A major issue at trial was whether plaintiff actually entered the area from the Whoopee side, or, as defendants argued, based on the evidence of snowboard tracks, from the Reservoir Hill side. No witness saw plaintiff enter the area.

During the trial, defendants introduced evidence that plaintiff had smoked marijuana while he was on the way to the ski area the morning of the accident. The evidence included the results of a quantitative test of plaintiffs urine after the accident that showed 603 nanograms of cannabi-noids per milliliter. According to defendants’ expert, Dr. Griffin, that level of cannabinoids indicated that plaintiff was impaired as a result of the marijuana at the time of the accident. Plaintiff first learned of the existence of the quantitative test immediately before trial, during the argument on his motion in limine to exclude all evidence of his marijuana use. He had requested all of his medical records from the appropriate providers during discovery and had received a copy of *434the original screening test of his urine, which merely showed the presence of marijuana metabolites. After receiving that test, he specifically asked the hospital whether it had any record of a quantitative test. The person in charge of the hospital’s records responded that there was no such record and that it was not the hospital’s policy to perform additional tests. However, despite those denials, a laboratory connected with the hospital had in fact performed the quantitative test that defendants introduced. Defendants apparently learned of that test as a result of consulting with an expert who worked at the laboratory that performed it. After learning of the test, defendants subpoenaed the test result from that laboratory without providing notice to plaintiff, either before or after issuing the subpoena, and without having a medical release from him.

Because plaintiff had not received a copy of the test report, he first learned of its existence as well as Griffin’s opinion regarding impairment during the argument on the motion in limine. As a result, plaintiff withdrew his motion in limine, and the evidence was admitted at trial over his objections that defendants had not shown that it was valid scientific evidence. Defendants used the evidence at trial both to discredit plaintiff’s version of how the accident occurred, by suggesting that the effects of the marijuana impaired his ability to remember what happened, and to support their argument that the accident was the result of plaintiffs, rather than defendants’, negligence. After the parties rested, the jury returned its verdict, answering “No” to the following question: “Were defendants at fault in one or more of the ways alleged by the plaintiff, and, if so, was such fault a cause of damages to plaintiff?”1

After trial, plaintiff moved for a new trial under ORCP 64 B(l) - (4). ORCP 64 provides that a trial court may set aside a verdict and grant a new trial on grounds that include:

“B(l) Irregularity in the proceedings of the court, jury or adverse party, or any order of the court, or abuse of *435discretion, by which such party was prevented from having fair trial.

“B(2) Misconduct of the jury or prevailing party.

“B(3) Accident or surprise which ordinary prudence could not have guarded against.

“B(4) Newly discovered evidence, material for the party making the application, which such party could not with reasonable diligence have discovered and produced at the trial.”

The trial court granted plaintiffs motion for a new trial. In explaining its reasoning as to why it was granting the motion, the court prefaced its ruling by remarking:

“I’m going to talk a little about this because I want the Court of Appeals to understand why I’m granting this motion for new trial. I have not granted a motion for a new trial since I have been on the bench.”

The court then stated its reasoning:

“I’m going to start with the evidence. At this point now that the science is in, if the matter was before me prior to trial, unless it was further developed in a different manner than what’s been submitted in the affidavits and briefs, I would not allow Griffin to testify about the effect of marijuana, because I think it is contrary to established science.

“Okay. How do we—how did we get to this point that I did not reach that conclusion until well after the trial? I think there was a problem in the response by Legacy. I think there was a problem in the analysis, the hip shot analysis by Brady [plaintiffs expert witness]. But do we end this trial and then have another lawsuit against the hospital for releasing their records without authority to the defense and, as a second prong, failing to adequately respond to the plaintiffs requests for his own records, and try that out were they—was the hospital at fault in some way for doing that? Did that cause the plaintiff to fail to recover here and therefore the hospital had to recover? I can see the case going down that track. Do we have another lawsuit against Brady for falling below the standard of care of an expert witness in not immediately sizing this up and giving the Court the information that he’s later given? I don’t think that’s good. I think this—that’s one way we could go.

*436“Was the subpoena response under [ORCP 55 [2] proper—not response, rather the procedure. I don’t believe so. I think that [ORCP 55 I] controls and that there should have been the notice and then we wouldn’t have gotten into this soup because the plaintiff would have known about this record prior to trial and evidence would have been developed. And on that point, this case just could not have been better prepared by—I think by either side.”

After digressing, the court summarized:

“So I think that the evidence on the use of marijuana would not have come into the case at all, that the reason that the plaintiff wasn’t able to present the science that reaches—leads me to that conclusion is because of conduct at the hospital, conduct of Brady and initially the irregularity in failing to follow the subpoena process.”

The above-quoted material demonstrates that the trial court based its grant of a new trial on three grounds. First, the court ruled as a matter of law that defendants’ evidence regarding plaintiffs use of marijuana before the accident occurred and Griffin’s opinion that plaintiff was impaired by that use at the time of the accident were inadmissible because Griffin’s opinion did not constitute valid scientific evidence. As will be explained later in this opinion in more detail, that conclusion constitutes legal error. Second, the trial court believed incorrectly that the purported procedural irregularities regarding Legacy’s response to plaintiffs subpoena constituted a justification under ORCP 55 I to grant a new trial. Third, and finally, the trial court ruled that a new trial would avoid a malpractice claim against Brady, plaintiffs expert witness, because of, according to the trial court, Brady’s “hip shot analysis” at trial.

On appeal, defendants make six assignments of error: (1) They assert that the trial court erred in granting plaintiffs motion for a new trial because the jury did not reach the question of comparative negligence, the question *437on which they claim that plaintiffs newly discovered evidence would have been probative. (2) They argue that plaintiff waived any error that was the basis of his motion for a new trial by withdrawing his motion in limine. (3) They claim that the trial court’s ruling that defendants violated ORCP 55 I was error. (4) They assert that the evidence offered by plaintiff is not “newly discovered evidence” within the meaning of ORCP 64 B(4). (5) They contend that

“a trial court may only supercede a jury verdict and order a new trial if there is no evidence to support the verdict. * * * [T]he heart of plaintiffs argument was that his expert disagreed with defendants’ expert regarding correlation between total cannabinoid levels and impairment.”

(6) Finally, they contend that the trial court erred when it granted partial summary judgment to plaintiff on defendants’ affirmative defense that plaintiff had released them from liability because he purchased a ticket containing a release provision and then used the ticket to ski on defendants’ ski lifts. Because defendants’ third and fourth assignments of error are dispositive, we do not consider their other assignments of error.

We return to the reasons for which the trial court granted a new trial. The trial court did not express which provisions of ORCP 64 B, in its view, authorized the grant of a new trial.3 We will therefore explore all the possibilities relied on by plaintiff in his motion. In their third assignment of error, defendants argue that the trial court erred in ruling that their action in issuing a subpoena for the result of the second, quantitative test without complying with the requirements of ORCP 55 I was “irregular” or constituted “misconduct,” thus implicating both ORCP 64 B(l) and (2). The relevant portions of ORCP 55 I provide:

“(2) If a patient or health care recipient is represented by an attorney, a true copy of a subpoena duces tecum for *438medical records of a patient or health care recipient must be served on the attorney for the patient or health care recipient not less than 14 days before the subpoena is served on a custodian or other keeper of medical records. Upon a showing of good cause, the court may shorten or lengthen the 14-day period.

* * * *

“(4) The requirements of this section apply only to subpoenas duces tecum for patient care and health care records kept by a licensed, registered or certified health practitioner as described in ORS 18.550, a health care service contractor as defined in ORS 750.005, a home health agency licensed under ORS chapter 443 or a hospice program licensed, certified or accredited under ORS chapter 443.”

ORCP 55 H(2) establishes similar requirements that apply to a subpoena duces tecum for hospital records when no hearing is scheduled and the subpoena directs delivery of the records to the attorney or party issuing the subpoena. However, ORCP 55 H(2) does not require advance notice to the plaintiff if the subpoena requires delivery of the record to a court or other authority that is conducting a deposition or a judicial or other formal hearing. It appears from the record that the laboratory informally told defendants of the result of the quantitative test before trial but did not provide them a copy. The only subpoena that defendants issued to the laboratory required its representative to bring the test result to the trial, although the record on appeal does not indicate whether the representative was to deliver it to the court or to defendants. Possibly for those reasons, plaintiff did not rely on ORCP 55 H(2) at the trial court, and he does not refer to it in this court.

It is undisputed that defendants did not comply with ORCP 55 I when they issued the subpoena for the second test. If they had done so, plaintiff would have known of the test, despite the hospital’s previous denial of its existence. Defendants argue however, that the laboratory that conducted the test was not part of the hospital where plaintiff was treated but was a separate laboratory that specialized in drug testing. To the contrary, there is evidence that the laboratory was part of the same hospital complex and that the *439laboratory’s records were part of the records of that complex. Nevertheless, ORCP 55 I, the rule on which plaintiff relies, does not apply to the laboratory. That rule, by its express language, applies to health care practitioners, not to institutions that keep records separately from practitioners. Although health care practitioners treated plaintiff at the hospital, the hospital, not the practitioners, kept the records of the laboratory tests. Because ORCP 55 I did not apply to the subpoena in question, there was no basis for the trial court to conclude that defendants’ failure to give 14 days notice of the subpoena constituted either an irregularity in the proceedings or misconduct of the prevailing party.4 The trial court erred when it relied on ORCP 55 I as a basis for granting of a new trial.

Plaintiff also relies on federal regulations concerning the confidentiality of information used for the purposes of diagnosis and treatment of drug abuse. 42 USC § 290dd-2; 42 CFR §§ 2.11 - 2.13. However, those provisions apply only to health professionals whose primary function is to provide alcohol or drug abuse diagnosis, treatment, or referral. 42 CFR § 2.12(e)(1). In this case, plaintiff did not receive drug abuse treatment or referral. Despite the positive result on the initial screening test, plaintiff stated that he did not believe that he had a chemical dependency problem and refused any treatment. Because there is no indication in the record that the quantitative test constituted a diagnosis under the federal rules, the test result was not protected. We conclude that the federal requirements are not a basis for finding that defendants acted irregularly or engaged in misconduct.

With regard to the trial court’s grant of a new trial to avoid a malpractice claim against Brady, plaintiffs expert witness, the court held in State v. Arnold, 320 Or 111, 120-21, 879 P2d 1272 (1994), that ORCP 64 B(4) is a statute like any other statute. Accordingly, this court’s task is to discern the intent of the legislature regarding the grounds upon which a new trial may be granted. Significant to this case, none of the *440enumerated grounds in ORCP 64 B authorize a court to grant a motion for a new trial to avoid a malpractice claim against an expert witness; after all, defendants are not required to endure a new trial after the jury has returned a verdict in their favor because of the performance of their adversary’s expert witness. The trial court erred by granting a new trial on a basis that is not legally cognizable.

Before we analyze the trial court’s first reason for granting a new trial, we pause to observe that, based on the above analysis, the only ground on which the grant of a new trial in this case could be upheld is ORCP 64 B(4). ORCP 64B (1) and (2) are eliminated as possibilities because plaintiff does not claim any irregularity or misconduct on the part of defendants or the jury except for the discovery issue regarding ORCP 55 I, an issue that has already been disposed of in favor of defendants. ORCP 64 B(3) authorizes trial courts to grant a new trial because of accident or surprise that ordinary prudence could not have guarded against. The general rule is that a party may not use a surprise as a basis for seeking a new trial unless the party moved for a continuance so as to respond dining the trial, thus possibly eliminating the need for a new trial. State v. Gardner, 33 Or 149, 152-53, 54 P 809 (1898); see also Transamerica Title Ins. Co. v. Millar, Inc., 258 Or 258, 263-64, 482 P2d 163 (1971); Turman v. Central Billing Bureau, 279 Or 443, 450, 568 P2d 1382 (1977) (holding that a party that is aware of a ground for a new trial during trial may not speculate on the result by failing to bring the issue to the court’s attention, intending to use the issue to seek a new trial if the outcome is not favorable). Here, although plaintiff was aware of Griffin’s testimony during trial, he never moved for a continuance so that he could respond differently than in the manner that Brady, his expert witness, testified. That leaves only ORCP 64 B(4) as a possible basis for granting the new trial.

In Arnold, the court specified,

“[W]e hold that evidence that may justify a court in granting a new trial [under ORCP 64B (4)] must meet the following requirements:

“(1) It must be such as will probably change the result if a new trial is granted;

*441“(2) It must be such as, with reasonable diligence, could not have been discovered before or during the trial;

“(3) It must be such that it cannot, with reasonable diligence, be used during trial;

“(4) It must be material to an issue;

“(5) It must not be merely cumulative;

“(6) It must not be merely impeaching or contradicting of former evidence.”

320 Or at 120-21 (footnote omitted).

At trial, the jury heard conflicting expert opinions from Griffin and from Brady, plaintiffs expert, about whether impairment at the time of the accident could be determined from the available evidence; yet the jury returned a verdict for defendants. If the evidence of plaintiffs marijuana use in Griffin’s opinion regarding plaintiffs impairment was properly admitted at trial (contrary to the trial court’s post-trial ruling), then the question arises whether plaintiff is otherwise entitled to a new trial under ORCP 64 B(4). Consequently, the analysis regarding the trial court’s riding about the admissibility of plaintiffs use of marijuana and Griffin’s opinion as to plaintiffs impairment breaks into two sub-parts: (1) whether the trial court erred in admitting the evidence of plaintiffs marijuana use and Griffin’s opinion at trial of the effect on plaintiff from that use; and (2) assuming that that evidence was admissible, whether plaintiff has otherwise demonstrated an entitlement to the grant of a new trial under ORCP 64 B(4).

To resolve whether Griffin’s opinion testimony about plaintiffs impairment at the time of the accident due to marijuana ingestion was properly admitted requires an application of the Supreme Court’s case law regarding the admission of scientific evidence. In State v. O’Key, 321 Or 285, 899 P2d 663 (1995), the court held that the admissibility of scientific evidence is governed by OEC 401, 702, and 403.5 Under OEC 401, evidence is relevant if it has the tendency to make the existence of any fact that is of consequence to the determination of the action more probable or less probable than it *442would be without the evidence. Here, defendants offered the evidence of plaintiffs impairment due to marijuana for two relevant reasons, to impeach his recollection of how the accident occurred and to prove that he was at fault in causing his injuries.

The next inquiry is under OEC 702, which defines the permissible sphere of expert testimony as whether “scientific, technical or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue [.]” The O’Key court applied a multifactor test that includes whether the theory or technique could be tested; whether the theory or technique had been subjected to peer review and publication; whether there is a known or potential rate of error in the scientific process; whether there exist operational standards that regulate the technique or theory; whether the theory or technique is accepted in the relevant scientific community; whether the expert’s qualifications qualify him as an expert in the field; whether the use of the theory or technique is widely used, the extent to which other courts admit its use, its novelty, the extent to which the theory or technique relies on a subjective interpretation; and the presence or lack of existence of safeguards. None of the above factors is decisive, nor is the list exclusive. In addition, the evidence must meet the helpfulness requirement of OEC 702.

According to the record, defendants’ expert, Griffin, is a medical doctor, board-certified in occupational medicine, and a review officer for OSHA compliance. He specializes in reviewing drug test results and has been specially trained in the effects of marijuana on the human body, including the decline of the effects of marijuana over time. He also supervises laboratory testing procedures for compliance with regulations and testing procedures. As we noted previously, Griffin testified that he reviews about 11,000 drug test results annually and regularly deals with the kinds of test results that occurred in this case. Out of the results that he had reviewed in the ten months preceding trial, about 990 were positive marijuana tests, but fewer than 20 of those involved test results above 500 nanograms/ml. Plaintiffs test level was 603 nanograms/ml at 8:45 p.m. on the day of the accident, which was, in Griffin’s opinion, “an exceptionally *443high level” in comparison with the test results that he typically evaluated. Plaintiff had consumed marijuana of an unknown quantity sometime the morning of the accident, and the accident occurred sometime around noon on that day. According to Griffin, the correlation between impairment and cannabinoid levels in urine samples has been the subject of published studies and those studies have been subject to peer review. Griffin submitted copies of those studies to the trial court before it ruled on plaintiffs motion for a new trial.

Plaintiff in his motion for a new trial, however, characterized Griffin’s testimony at trial as a bald assertion that a measure of total cannabinoids that he considered high must mean that plaintiff was impaired at the time of his injuries. According to plaintiff, the scientific literature is “unanimous” that such a correlation cannot be scientifically made. In his reply brief to the trial corut on the motion for new trial, plaintiff criticized the conclusions that Griffin drew from each of the studies on which Griffin relied and, on appeal, plaintiff incorporates by reference those criticisms.6

Ultimately, the question is whether plaintiffs criticisms of Griffin’s opinion go to the admissibility of his testimony or to the weight that the trier of fact gives it. In that regard, the case of Jennings v. Baxter Healthcare Corp., 331 Or 285,14 P3d 596 (2000), informs the issue. At issue in that case was the admissibility of the testimony of a board-certified neurologist with an advanced degree in neurophysiology. He testified, based on his clinical experience with patients, that the silicone from the plaintiffs breast implants had migrated throughout her body and caused her personal injuries. On appeal, the Supreme Court upheld the admissibility of his opinion as scientific evidence. The court focused on the physician’s methodology that led to his opinion rather than on his ultimate conclusion. The court noted that, although the witness’s hypothesis had not been tested by others, nor had it been subject to peer review or been published, *444OEC 702 does not preclude the admission of novel scientific evidence. Id. at 307. Ultimately, the court concluded that, because the witness’s methodology supported his conclusions, his testimony was admissible as scientific evidence.

The governing principle from Jennings is that scientific evidence will be admissible if the evidence is based on a methodology that has scientific validity, even if other scientists or experts might be unwilling to draw the same conclusion. In this case, Griffin relied on the test of plaintiffs urine after the accident, his clinical experience with such test results, and the findings of certain studies as the basis for his opinion that plaintiff was impaired at the time of the accident. He compared the effect of alcohol on the human body to the effect of marijuana on the human body and testified that there is a known range of marijuana in the body above which “nearly all or substantially all adults would be impaired.” He testified that, in his experience and based on reviews of medical literature, “the range at which one can with confidence state that all or substantially all adults would be impaired is above 400 [nanograms per milliliter total cannabinoids].” He explained, “So if we’re looking at a urine level of 603 nano-grams at 8:45 in the evening, and we can see that this is a level in the urine, we can anticipate accurately that sometime five, six, seven hours back was the peak level in the blood and the peak effect on the individual’s psycho-motor function [.]”

For purposes of discussion, we assume that members of the scientific community would disagree with Griffin that a qualified medical person could opine with reasonable medical probability that plaintiff was impaired at the time of the accident based on plaintiffs test result. For example, in his affidavit in support of plaintiffs motion for a new trial, plaintiffs expert explained that the active ingredient of marijuana associated with impairment of the human body is called “Delta 9-THC.” When ingested, that ingredient is metabolized by the body within four hours or less into several metabolites. Once metabolized, the primary metabolite left in the body is known as carboxy-THC and is not known to cause impairment. “[U]rine tests for carboxy-THC may reflect the presence of metabolites from both recent use (use that day) and several days prior, without distinguishing *445which metabolites are from which use.”7 Thus, plaintiffs expert concluded in his affidavit that the carboxy-THC test result relied on by defendants’ expert is not a reliable indicator of the active ingredient of marijuana in plaintiffs body at the time of the accident because the test does not discern when the marijuana was ingested, nor does it indicate the amount of the active ingredient of marijuana in the body as distinguished from other metabolites or cannabinoids of marijuana that do not cause impairment.8

In response to the post-trial challenge made by Brady to Griffin’s trial testimony, Griffin filed a controverting affidavit in which he attached copies of scientific literature, including a study that suggested, based on tests of airplane pilots, that the “carry-over effects of low to moderate social doses [of marijuana] may produce significant impairment for as long as 24 h [sic] on complex human performance involving machines.” Griffin also averred in his post-trial affidavit that “a mine level of 603 nanograms per mil total cannabinoids means” that the active cannabinoid level in that same sample ranges from a “6-to-l to a 10-to-l ratio.” Griffin reiterated, “Based upon my experience, a urine level of603 nanograms per mil is very high.” Finally, he said in his affidavit:

“Based upon the studies cited above, the studies cited in my testimony, plaintiffs admission that he smoked marijuana that morning, my training and experience, it is my opinion that plaintiff was impaired at the time of his accident. My opinion is based upon a reasonable medical probability.”

We conclude on this record that Griffin’s opinion satisfies the admissibility requirements for scientific evidence because of the methodology used by him to arrive at his conclusion. His opinion is based on scientific facts derived from scientific studies and his own clinical experience as well as on empirical reasoning. As we understand his point, from the *446fact that, at 8:45 p.m. on the day of the accident, plaintiff’s urine contained 603 nanograms/ml, it can be inferred that his urine also contained active cannabinoids at that time based on a known ratio. In addition, Griffin’s opinion takes into account the fact that the effect of marijuana peaks one hour after consumption and then subsequently dissipates. Accordingly, Griffin concluded, “So a level of 600 implies to me that any adult at that level would have been impaired.” Thus, in light of the dissipation rate over an eight-hour period, Griffin drew the inference from plaintiff’s test result based on his experience and scientific literature that the active ingredients of marijuana in plaintiff’s body must have been high enough at the time of the accident to have caused him to be impaired.

Griffin’s methodology may be novel, as plaintiff asserts, and there may be other factors, including the fact that the test performed on plaintiff’s urine measured the residual amount of total cannabinoids ingested over a 72-hour period of time, that challenge the accuracy of Griffin’s opinion. In addition, as plaintiff’s expert’s affidavit demonstrates, there exist scientific opinions contrary to Griffin’s opinion. But differing opinions based on scientific methodology do not necessarily result, under the applicable law, in the exclusion of one of the opinions. Rather, generally, a conflict in scientific opinion presents a question of weight to be resolved by the trier of fact. Here, we hold that the arguments raised by plaintiff go to the weight to be given Griffin’s testimony and not its admissibility in light of the evidence of the high level of cannabinoids found in plaintiff’s urine some eight hours after the accident and Griffin’s experience in evaluating test levels. As the court said in O’Key, “[t]he scientific method is a validation technique consisting of the formulation of hypotheses, followed by observation or experimentation to test the hypotheses.” 321 Or at 292. In the same manner, “ ‘clinical diagnoses’ bear the mark of science.” Jennings, 331 Or at 304. Griffin’s use of empirical reasoning based on scientific facts and his reliance on his own clinical experience and the scientific literature to which he referred demonstrate a methodology that satisfied the requirements of the law for purposes of admissibility. Thus, as a matter of law, Griffin’s testimony was admissible, and the trial court *447erred when it ruled, based on the information supplied by plaintiff in support of his motion for a new trial, that the evidence of marijuana usage and Griffin’s opinion testimony about plaintiffs impairment was inadmissible. In other words, the trial court correctly admitted into evidence the test result and Griffin’s opinion during the trial, and that evidence would also be admissible in the event a new trial occurs.

We turn now to the second part of the analysis. Having concluded that evidence of plaintiffs marijuana use and Griffin’s opinion regarding the effect of its use on plaintiff are admissible, the issue is whether plaintiff otherwise is entitled to a new trial on this record. ORCP 64 B(4) requires that the evidence that entitles a movant to a new trial must be “newly discovered” evidence, evidence that will probably change the result if a new trial is granted and evidence that reasonably could not have been discovered or used at trial. Arnold, 320 Or at 120. In plaintiffs motion for a new trial, he described the “newly discovered evidence” that justified the grant of a new trial under ORCP 64 B(4):

“Plaintiff discovered the additional drug test on the first day of tr[ia]l. Over the course of trial and afterwards, Plaintiff discovered expert evidence that could show the additional test result was not probative of impairment and should have been excluded along with all other marijuana evidence. Plaintiff could not have reasonably presented the evidence at trial for two reasons. First, plaintiff did not discover much of the evidence (academic literature showing that the urine test could not be used to judge impairment). Second, the expert evidence Plaintiff discovered in the course of trial was discovered only after the time when it would have made a difference, which was in the hearing on Plaintiffs motion in limine.”

From plaintiffs motion, we learn that the purported “newly discovered” evidence was scientific literature. Plaintiff does not tell us in his motion whether the “newly discovered” literature was discovered during or after trial, only that it was discovered “after the time when it would have made a difference.” We turn then to plaintiffs expert and his affidavit in support of plaintiffs motion for a new trial for more answers. Our first task is to identify what scientific literature *448plaintiff claims to be “newly discovered.” In that vein, Brady’s affidavit is careful to aver that his survey of the literature occurred “[d]uring and after [plaintiffs] trial.” Importantly, Brady never expressly identifies which scientific articles he claimed to have discovered post-trial. Our own review of the record reveals that Brady refers in his post-trial affidavit to only one other specific source of scientific literature not mentioned in his affidavit in support of plaintiffs motion in limine made before trial or made available to him by Griffin during trial. That study was conducted by the United States Center for Disease Control and Prevention, and it was published in 1983. Perhaps it can be inferred from the record that that study constitutes the “newly discovered” post-trial evidence, but we are left to imagine how its contents differ from the contents of the literature that Brady surveyed before and during trial. Also, in his affidavit in support of plaintiffs motion for a new trial, Brady referred to studies furnished to him through Griffin’s file, but those studies were furnished to Brady during trial after Griffin finished testifying, and Brady testified regarding them in his rebuttal testimony that followed Griffin’s testimony before the trial ended. Finally, in light of Brady’s assurances to the court pretrial and to the jury that no scientifically accepted method of calculating impairment based on the available data existed, we are inclined to take him at his word that he had undertaken a survey of the existing literature before he made those assertions.

Even if we assume that plaintiff made an adequate identification in his motion of what he claims to be “newly discovered” evidence, there is yet another requirement that plaintiff has not met under ORCP 64 B(4). Based on the purported discovery of newly discovered scientific literature, Brady proposes to testify at a new trial that Griffin’s opinion that the marijuana consumed by plaintiff impaired him at the time of accident has no scientific validity. However, he gave the same opinion in his trial testimony; thus, the issue is whether there is anything new about Brady’s testimony offered in his post-trial affidavit. The dissent argues that Brady’s testimony is “newly discovered” because “Brady gave only negative testimony at trial—that he was unable to find relevant studies—while after the trial he was able to show *449positively that there was no scientific basis for using the quantitative test to indicate impairment.” 195 Or App at 468 (Wollheim, J., dissenting). The dissent also asserts that Brady “did not assert [at trial] that Griffin’s opinion was without a scientific foundation or that it was impossible to determine impairment from a quantitative measurement of metabolites rather than of total cannabinoids.” Id. at 473.

Additionally, the dissent lists five factual statements in Brady’s post-trial affidavit that it says “are in neither the pretrial affidavit nor Brady’s testimony at trial.” Id. at 474. They are, according to the dissent: (1) the fact that excretion of marijuana metabolites into the urine varies dramatically among individuals; (2) because of such variations, it is not possible to correlate a quantitative test of metabolites in urine with the amount of active metabolites in a person’s blood at a particular time; (3) there is no correlation of the active ingredient of marijuana with impairment that can be made in the way that blood alcohol content can be correlated with impairment; (4) the United States Center for Disease Control and Prevention advised in 1983 that a urine test alone cannot indicate impairment; and (5) “[g]iven those facts, together with the lack of any study that would enable an expert to determine impairment from a measure of total cannabinoids, Brady was highly doubtful that there was any scientific method that could lead to an informed conclusion about impairment based on a measure of total cannabinoids in a urine sample.” Id. at 474-75. The dissent concludes that “[t]he post-trial affidavit, thus, contained a substantial amount of new information, most significantly that it is not possible to use even a urine test for the total metabolites to determine impairment.” Id. at 475.

Our review of the record indicates that it does not support the dissent’s assertion that Brady based his post-trial opinion on “new” information. In fact, Brady testified at trial that the quantitative test of plaintiffs urine “clearly * * * does not” measure how much of the active ingredient was ever present. That is the identical assertion that plaintiff made in his motion for a new trial when he said that “the urine test could not be used as a reliable gauge of impairment * * *.” Moreover, Brady was also asked in his testimony at trial about the scientific literature in Griffin’s file and *450whether those articles attempted to connect a quantitative test result with a ‘level of impairment at any particular time.” Brady responded, “No, they don’t.” Then, he was asked whether he was aware of any scientific literature or studies that connect a quantitative test result with level of impairment at any particular time. Brady answered:

“No, I’m not, [plaintiffs counsel]. You’ve asked me this question. I’ve done my best to look around, and I know of no literature, no study that I’m familiar with that correlates the total cannabinoids with the question of impairment.”

In addition, Brady was asked at trial if, in his opinion, an expert in his field could determine with reasonable medical probability from the test result “whether the patient was impaired at some particular time before?” Brady answered, “At a particular time prior to this sample being taken? No, no.” Finally, Brady was asked at trial, “Can you tell from [the test result] how much of that particular metabolite or breakdown product was in the test subject’s urine?” Brady responded, “No. All we know was the total material all together, and this gives us no information about any particular component of that whole big glop.”

In comparison to his trial testimony, Brady averred in his post-trial affidavit:

“I am unaware of, and have not been able to locate, any scientific study that enables an expert to use a measure of total cannabinoids to form an opinion about whether the subject was impaired at some particular prior point in time. Given the problems inherent in judging impairment even from a particular metabolite of active ingredient (as discussed above), and given the inability to distinguish the contribution of each substance included in a measure of total cannabinoids, I am highly doubtful that there is any scientific method that could lead to an informed conclusion about impairment based on a measure of total cannabinoids in a urine sample.”

Later, in his post-trial affidavit, Brady added that Griffin’s opinion “is not an opinion that can be reached through any accepted scientific method of analysis” and that “there is no scientifically accepted method for getting from total cannabinoids in [plaintiffs] urine later that day to *451active ingredients in his blood at the time of his injuries * * * because * * * different people metabolize the active ingredient at different rates and secrete the metabolite in different amounts and at different times.” He explained further in his affidavit,

“Moreover, the measure of carboxy-THC would fail to distinguish use that day that might have caused impairment from use days earlier that could not have caused impairment that day. The result reflected in the test result produced on the day of trial is fraught with the further problem that it is not even a measure of carboxy-THC. It is a measure of total cannabinoids that does not distinguish between active ingredient, carboxy-THC, [or] other metabolites or cannabinoids that do not cause impairment.”

Brady also concluded in that portion of his affidavit that “[n]o accepted scientific method allows such a conclusion “that an opinion as to impairment can be drawn from the measure of total cannabinoids in plaintiff’s urine sample.” Brady also referred in his affidavit in support of plaintiffs motion for a new trial to the literature in Griffin’s file. He opined that

“the literature in Dr. Griffin’s file (1996 and 1998 studies by Huestis, Cone and, in the 1996 study, Mitchell) does not support Dr. Griffin’s claim that a measure of total cannabi-noids in a person’s urine can be used to determine whether a person was likely to be impaired at a particular point in time.”

In sum, the record demonstrates that, if a new trial is granted, Brady’s testimony will be a repeat of his trial testimony. Contrary to the dissent’s assertion that Brady did not assert at trial that there was no scientific basis for using the quantitative test to indicate impairment, Brady testified both at trial and in his post-trial affidavit that he was unable to find relevant studies to support Griffin’s opinion and that an expert in his field could not determine with reasonable medical probability whether plaintiff or any person was impaired at a particular time.

Finally, what does the record show, as ORCP 64 B(4) requires, about whether plaintiff—through Brady—exercised reasonable diligence to discover the “new” scientific evidence before the trial ended? Recall that the trial court found *452that Brady had “shot from the hip” during trial. Brady does not explain in his affidavit why he could not have done the necessary research before the trial ended,9 and the evidence submitted by Brady in his affidavit suggests the contrary inference. The most pertinent fact relative to that issue is that all of the studies on which Brady relies were published in scientific journals many years before the trial in this case. The second most pertinent fact relative to the issue is that Brady represented to the court and the jury during the trial that he had previously conducted a complete survey of existing scientific literature on the subject. Brady’s affidavit in support of plaintiffs motion for a new trial does not respond to those facts or explain why his research done during the trial was incomplete. Although Brady says that, after the trial, he “had time to further review the drug test,” he does not explain what circumstances prevented him from taking the time to do the required research before trial ended.

The chronology of the trial further supports the court’s belief that Brady “shot from the hip.” Plaintiff became aware of defendants’ evidence no later than the morning of June 26, 2001, at the hearing on his motion in limine. The trial also began on that day. The trial ended on July 2, 2001, after a weekend recess. Plaintiff called Brady as his final rebuttal witness on June 29, and Brady testified before the jury that Griffin’s opinion had no scientific basis. Plaintiff never explains in light of those facts why the survey of scientific literature that produced the “newly discovered” evidence *453could not have been discovered before the end of trial through the exercise of reasonable diligence.

Additionally, the dissent’s assertion that Brady was unable to reach a conclusion that there was no scientific support for Griffin’s opinion until after the trial is also contradicted by Brady’s affidavit made before trial. In that affidavit, Brady makes the following generalizations:

“8. It is not possible to determine from the type of urine test that was done for [plaintiff] whether [plaintiff] was mentally or physically impaired at the time of his accident. As noted, the type of urine test that was done for [plaintiff] can only determine whether [plaintiff] ingested some amount of marijuana at some indeterminate time within the previous 72 hours.

“9. It also is impossible to determine from the other evidence regarding [plaintiffs] marijuana use whether [plaintiff] was mentally or physically impaired by marijuana at the time of his accident. No reliable scientific method, and no method commonly accepted in the scientific community, allows such a determination from the facts that are available without also knowing the precise details regarding the potency of the particular marijuana and the precise amount ingested.”10

(Emphasis added.) The dissent posits that Brady was only aware of the screening test performed on plaintiffs urine at the time he made those averments, an assertion, which, of course, is correct. But Brady’s affidavit expressly states that he conducted a scientific investigation that extended beyond ascertaining the limits of screening tests. According to his affidavit, it is impossible to determine from the screening test *454and “other evidence” that plaintiff was impaired at the time of his accident. As a result of his investigation, Brady concludes that “[n]o reliable scientific method * * * commonly accepted in the scientific community * * * allows such a determination from the facts * * *.” Presumably, in light of his sweeping statements about the scope of his investigation, Brady was aware of or should have been aware of, the existence of the kind of quantitative test performed on plaintiffs urine.

Plaintiffs argument to the trial court in support of plaintiffs motion of a new trial further informs the issue of whether Brady acted with reasonable diligence during the trial. Counsel told the court:

“[W]e knew we had been surprised by the additional drug test result, but we didn’t know we had been harmed by it. We didn’t know that given some additional time to consider and analyze this result, to look at the literature and so on, we would have able to conclude, you know, this is really no more probative than anything else in this case, we can still get all this marijuana evidence excluded. We didn’t know that the surprise had hurt us * * *. As time went on, and we heard the other testimony unfold, and we looked at literature in the course of and after trial, our opinion on that issue changed.”

(Emphasis added.) Counsel’s statement tells us that plaintiff made the strategic decision to proceed with the trial on the first day of trial and then, at some later point in time, decided that plaintiff had been harmed by Griffin’s testimony. Taking plaintiff at his word, the earliest that plaintiff could have realized that he had been harmed would have been after Griffin testified on direct examination in the latter stages of trial. Nonetheless, plaintiff does not explain why Brady could not have begun his research on the first day of trial when he learned about Griffin’s testimony and finished any additional research before he testified on rebuttal days later or why plaintiff could not have asked for a continuance to complete his research.

Arnold is again particularly instructive regarding the requirement of “the exercise of reasonable diligence,” as applicable to this case. 320 Or at 119-22. The defendant in that case, after being convicted of sexual abuse of a child, *455appealed and assigned as error the denial of her motion for a new trial under ORCP 64 B(4) based on the discovery of new evidence. At trial, a prosecution witness testified about her background in child development and about statements made to her by the victim regarding the defendant’s alleged inappropriate touching. On cross-examination, defense counsel examined the witness extensively about her educational qualifications. Immediately after the witness testified, the defendant’s investigator contacted the college that the witness said she had attended and learned that the college had no record of her attendance at that institution. Although defense counsel was aware of that information, he did not bring the information to the attention of the trial court, he did not ask for a continuance, and he did not call the investigator as a rebuttal witness. In her motion for a new trial based on newly discovered evidence, the defendant asserted that, after a complete investigation had been completed, it was evident that the witness had lied about her educational background. The trial court denied the defendant’s motion on the ground that defense counsel’s decision not to raise the matter during trial precluded the grant of the motion. On appeal, this court reversed. On review, the Supreme Court reversed our ruling and remanded for the consideration of the defendant’s other assignments of error.

In reaching its decision, the Supreme Court considered the state’s argument that the defendant’s “newly discovered” evidence did not meet the requirements of ORCP 64 B(4) and that, therefore, the trial court had no discretion to exercise and no authority to grant the defendant’s motion for a new trial. Arnold, 320 Or at 121. The court pointed out that, before the case was submitted to the jury, defense counsel had specific information that the college that the witness said she had attended denied that she had ever attended there and that counsel knew how to obtain that evidence through a subpoena. Moreover, defense counsel never asked for a continuance or for the opportunity to issue a subpoena to procure the evidence. The court held that a new trial should not have been granted under those circumstances. It explained:

“This case is a dispute over how much evidence had to be known and had to be usable during trial to make the balance of the evidence not such as would justify the award of *456a new trial. Even if some additional evidence is discovered after trial, it does not justify the award of a new trial if the evidence that was known during trial could have been used during trial for substantially the same purpose as the additional evidence that is not discovered until after trial.”

Id. (emphasis in original).

The circumstances in Arnold parallel the circumstances in this case. Plaintiffs counsel knew of the issue days before the case was submitted to the jury, and plaintiffs expert witness, Brady, like any expert in his field, must have known how to conduct a survey of pertinent scientific literature. Also, as in Arnold, plaintiff never asked for a delay in the trial to procure additional evidence before the case went to the jury. It is evident from his testimony and his pretrial affidavit that Brady believed before and during the trial that there existed no test that correlated the amount of the active ingredients of marijuana in a person’s body with impairment at a particular time. Even if there was some scientific literature discovered post-trial, that literature could not justify the grant of a new trial under ORCP 64 B(4) because Brady used the literature in Griffin’s file and his own knowledge of the subject matter for substantially the same purpose as the “newly discovered” literature would be used in a new trial. In light of the public’s interest in the finality of the judicial process, the trial court was left with no discretion to exercise in this matter under ORCP 64 B(4). We conclude that plaintiffs attempt to satisfy the legal requirements of ORCP 64 B(4) is legally inadequate on the face of the record for each one of the above reasons.

According to the dissent, “the majority’s errors fall into three basic categories. First is its failure to follow the applicable standard of review.” 195 Or App at 476 (Wollheim, J., dissenting). In the dissent’s view of the proper standard of review, we are to resolve all factual issues in favor of plaintiff. Thus, the dissent’s opinion is replete with references about what “the trial court could have found.” See, e.g., id. at 470. In fact, the trial court made no findings of fact.11

*457The dissent also argues that we must defer to the trial court’s exercise of discretion in this case, based on the implicit findings of fact that the dissent attributes to the trial court. However, the dissent misconstrues the nature of the discretion granted to the trial court under ORCP 64 B(4). A trial court has no discretion to grant a new trial under ORCP 64 B(4) unless the legal requirements of the rule are met. “ORCP 64 B(4) is a statute.” Arnold, 320 Or at 119. “[E]vidence that may justify a court in granting a new trial must meet [the requirements set out above].” Id. at 120. The threshold issue in Arnold was whether ORCP 64 B(4) applied to evidence discovered during the trial. After resolving that issue, the Arnold court explained:

“Having held that ORCP 64 B(4) applies to evidence discovered during the trial, we proceed to consider the state’s argument that defendant’s ‘newly discovered evidence’ would not have justified the award of a new trial under ORCP 64 B(4) and, therefore, that the trial court had no discretion to exercise, and, thus no authority to grant defendant’s motion.”

320 Or at 121. The court concluded:

“[T]he post-trial discovery of the latter information does not satisfy the requirement that the evidence must be such as, with reasonable diligence, could not have been discovered and produced at trial.”

Id. at 122. In other words, the determination of whether the requirements of ORCP 64 B(4) are met is not the kind of decision that involves more than one legally permissible choice to be made by the trial court. See State v. Rogers, 330 Or 282, 312, 4 P3d 1261 (2000) (“If there is only one legally correct outcome, then ‘discretion’ is an inapplicable concept.”). Rather, Arnold teaches us that the decision to be made by the court under ORCP 64 B(4) is initially a legal determination about the adequacy of the movant’s grounds for a new trial.

As we recently said in State v. Woodman, 195 Or App 385, 387, 97 P3d 1263 (2004), while the grant of a new trial is reviewed for an abuse of discretion, “to the extent that the *458grant is based on the interpretation of law, we review for legal error.” See also Bennett v. Farmers Ins. Co., 332 Or 138, 151, 26 P3d 785 (2001) (“When the trial court’s order of a new trial is based on an interpretation of the law, we review that order for errors of law.”). Here, because the trial court based its grant of a new trial on its interpretation of the applicable law regarding the admissibility of scientific evidence, the purported discovery violation under ORCP 55 I, and the desire to avoid a malpractice claim against Brady, our standard of review is for errors of law.

The dissent also argues that the issue of whether Griffin’s testimony was admissible scientific evidence and whether the court erred in finding that it would have rejected that testimony at trial if Brady had been able to present the information that he discovered after trial are issues that defendants do not raise on appeal. The dissent’s argument fails to appreciate the express argument made under defendants’ fourth assignment of error and its necessary implications. In that assignment, defendants first assert that “[t]he trial court erred in granting the plaintiffs motion for a new trial because the supplemental opinion of plaintiffs expert is not newly discovered evidence.” That argument clearly raises the issue whether the trial court ruled correctly under ORCP 64 B(4). Next, in its brief on appeal, defendants expressly point to the portion of the transcript previously quoted in this opinion where the trial court states:

“At this point now that the science is in, if the matter was before me prior to trial, unless it was farther developed in a different manner than what’s been submitted in the affidavits and briefs, I would not allow Griffin to testify about the effect of marijuana, because I think it is contrary to established science.”

(Emphasis added.) We cannot decide whether plaintiff is entitled to a new trial under ORCP 64 B(4) without first deciding whether the trial court’s ruling regarding the scientific validity of Griffin’s opinion was correct.12 For purposes of *459this case and the trial court’s ruling, Brady’s “newly discovered” evidence must constitute evidence that makes Griffin’s opinion inadmissible as a matter of law; otherwise, Brady’s testimony amounts to the same evidence that he presented at trial, and plaintiff is not entitled to a new trial under ORCP 64 B(4). Thus, determining the correctness of the trial court’s ruling regarding the validity of Griffin’s scientific opinion, a ruling expressly challenged by defendants in their fourth assignment of error, is a necessary predicate to deciding whether the trial court’s grant of a new trial under the subsection of ORCP 64 B was correct.

Moreover, parties to an appeal cannot prevent us from noticing and invoking the requirements of a statute merely because they have failed to assert its application. Miller v. Water Wonderland Improvement Dist., 326 Or 306, 309, 951 P2d 720 (1998); Burk v. Hall, 186 Or App 113, 118, 62 P3d 394, rev den, 336 Or 16 (2003). Plaintiff can prevail on appeal under ORCP 64 B(4) only if the rule’s requirements are met and its requirements compel us to determine if there exists “newly discovered evidence.” It follows from defendants’ assignment of error that we must necessarily decide whether the trial court’s subsequent reversal of its own evidentiary ruling made during trial was correct.

Finally, the dissent argues that we err because, according to the dissent, we infer improperly that Brady should have been able to find the purported new information during trial and because we impermissibly rely on the statements of plaintiffs counsel to draw that inference. 195 Or App at 476-77 (Wollheim, J., dissenting). Apparently, the dissent’s notion flows from its belief that we are required to draw all factual inferences in favor of plaintiff. Once again, the dissent’s argument is premised on a misunderstanding of the proper standard of review. As stated above, the trial court made no findings nor is there any concrete suggestion in the record that it drew any inferences pursuant to ORCP 64 B(4). Additionally, the burden was on plaintiff to make an adequate showing in the trial court that he had fulfilled all the legal requirements under the rule. In fact, for the reasons mentioned previously, plaintiff made no prima facie showing of compliance with the rule’s requirements; that is the point of our references to both Brady’s failure to explain why the *460research could not have been completed before trial and to plaintiffs counsel’s explanation of that failure.

In summary, the trial court granted a new trial premised on three legal rulings: (1) Griffin’s opinion does not constitute valid scientific evidence; (2) ORCP 55 I was violated; and (3) if a new trial is not held, plaintiff may sue Brady for malpractice. A trial court does not have discretion to commit errors of law. It must exercise its discretion within a range of legally permissible choices. None of the above grounds relied on by trial court involves the exercise of discretion. From the time that plaintiff filed the pretrial motion in limine until the arguments were made in support of his motion for a new trial, plaintiff never wavered in his position that the evidence of his marijuana usage and Griffin’s testimony was inadmissible. Indeed, he offered Brady’s testimony at trial on that very issue. In contrast, the concept of “newly discovered” evidence contemplates new evidence discovered post-trial that reasonably could not have been discovered before or during trial with the exercise of reasonable diligence and that is likely to change the result of the trial. It is plaintiff’s burden to demonstrate that the requirements of the rule have been satisfied. In the absence of satisfying the requirements of the rule, he is not entitled to a new trial as a matter of law. That is exactly what occurred here, and the record is clear that the trial court’s reasoning that led it to grant a new trial was predicated on its rulings of law, rulings that are erroneous. Consequently, reversal of the grant of a new trial is required, and the judgment for defendants must be reinstated.

Reversed on appeal; affirmed on cross-appeal.

dissenting.

The majority reverses the trial court’s grant of a new trial in this case on an issue that defendants did not raise on appeal, and it decides that issue incorrectly. None of the issues that defendants do raise justifies reversal. This court should affirm the trial court’s order.

I begin with the facts, both those involved in plaintiffs claim and those that occurred during the trial and the post-trial proceedings. Plaintiff was injured while snowboarding at a ski area that defendants operate on Mt. Hood. He snowboarded into a wooded area between two ski runs, Whoopee and Reservoir Hill, he fell into Reservoir Creek, which runs through the wooded area, and was seriously injured. Plaintiff testified that he entered the wooded area at a nonprotected portion of the Whoopee side rather than at the fully protected Reservoir Hill side. In his complaint, he alleges that defendants were negligent in, among other things, “failing to install and maintain barriers and/or warnings sufficient to prevent skiers or snow boarders from accidentally skiing or snow boarding into the creek bed from the west or south side.” At trial defendants denied that plaintiff entered the area from the Whoopee side and argued, based in part on the evidence of snowboard tracks, that he actually entered from the Reservoir Hill side. No witness saw plaintiff enter the area. The jury answered “No” to the following question: “Were defendants at fault in one or more ways alleged by the plaintiff, and, if so, was such fault a cause of damages to plaintiff?”

During the trial, defendants introduced evidence that plaintiff had smoked marijuana while he was on the way to the ski area the morning of the accident. That evidence included the results of a quantitative test of plaintiffs mine after the accident that showed 603 nanograms of cannabi-noids per cubic centimeter. According to defendants’ expert, Dr. Griffin, that level of cannabinoids indicated that plaintiff was impaired as a result of the marijuana. During discovery, plaintiff had requested all of his medical records from the appropriate providers. He had received a copy of the original *462screening test of his urine, which merely showed the presence of marijuana metabolites, not their amounts. He specifically asked the hospital whether it had any record of a quantitative test. The person in charge of the hospital’s records responded that there was no such record and that it was not the hospital’s policy to perform additional tests. Despite those denials, a laboratory connected with the hospital had in fact performed the quantitative test that defendants introduced. Defendants apparently learned of that test as a result of consulting an expert who worked at the laboratory that performed it. They then subpoenaed the test result from that laboratory without providing notice to plaintiff, either before or after issuing the subpoena, and without having a medical release from him.

Plaintiff filed a motion in limine to exclude evidence of his use of marijuana. He relied on the opinion of his expert, Dr. Brady, that a positive result on a screening teest could not show that plaintiff was impaired at the time of his injury. During the argument on the motion in limine on the first day of the trial, defendants revealed the existence of the quantitative test. That revelation took defendants and Brady by surprise. Because he had not conducted research on the issue, Brady was not able at that time to state that an expert could not conclude that plaintiff was impaired at the time of the accident. Plaintiff therefore withdrew the motion in limine, and the evidence was admitted at trial over plaintiffs objections that defendants had not shown that it was sufficient to indicate impairment. Defendants used the evidence both to discredit plaintiffs version of how the accident occurred, by suggesting that the effects of the marijuana impaired his ability to remember what happened, and to support their argument that the accident was the result of plaintiffs, rather than defendants’, negligence.1

During and after the trial, plaintiff investigated the circumstances under which defendants received the quantitative test report, and Brady conducted research into its scientific significance. Brady was ultimately able to determine *463that there was no scientific basis for concluding that the total amount of cannabinoids in plaintiff’s urine, which reflected the total material from the marijuana smoke rather than only the psychoactive component and its metabolites, could indicate whether plaintiff was impaired at the time of his injury. Plaintiff moved for a new trial on the ground that the trial court should not have admitted evidence concerning his marijuana use because there was no scientific basis for determining from the evidence that plaintiff was impaired.

ORCP 64 provides that a trial court may set aside a verdict and grant a new trial on grounds that include:

“B(l) Irregularity in the proceedings of the court, jury or adverse party, or any order of the court, or abuse of discretion, by which such party was prevented from having fair trial.

“B(2) Misconduct of the jury or prevailing party.

“B(3) Accident or surprise which ordinary prudence could not have guarded against.

“B(4) Newly discovered evidence, material for the party making the application, which such party could not with reasonable diligence have discovered and produced at the trial.”

Plaintiff moved for a new trial on each of those grounds. The trial court granted the motion without indicating the ground on which it relied. In doing so, it found that the evidence that plaintiff presented to support the motion showed that Griffin’s testimony about the effect of marijuana was contrary to established science. The court also found that the reason that plaintiff did not present that evidence before trial was in part the conduct of the hospital, in part the conduct of Brady, and in part the conduct of defendants in not following the proper subpoena procedure for obtaining the report. It concluded that the evidence could have had a significant effect on the trial, in part because it could have led the jury to find that plaintiffs marijuana use, rather than anything that defendants did, was the cause of plaintiffs injury and in part because inappropriate references to marijuana would unfairly impair plaintiffs right to a fair determination of the case.

*464Whether to grant a motion for a new trial is a discretionary decision for the trial court. See Ertsgaard v. Beard, 310 Or 486, 496, 800 P2d 759 (1990); State v. Miller, 167 Or App 72, 75-76, 1 P3d 1047, rev den, 330 Or 553 (2000). We will affirm the trial court’s exercise of its discretion to grant a new trial “if any of the grounds argued in support of the motion is well taken and the error is prejudicial. Williams v. Laurence-David, 271 Or 712, 718, 534 P2d 173 (1975),” Schacher v. Dunne, 109 Or App 607, 609, 820 P2d 865 (1991), rev den, 313 Or 74 (1992). A party seeking to reverse such an order must show that none of the grounds specified in the motion is valid, and we give the trial court every benefit of the doubt in that regard. Hightower v. Paulson Truck Lines, 277 Or 65, 69, 559 P2d 872 (1977). At least to the extent that the trial court based its decision on the ground of newly discovered evidence, we defer to its explicit and implicit factual findings. See OEC 104.

Defendants raise several issues that, if correct, would require reversing the trial court’s order. Before discussing my disagreements with the majority’s analysis, I first consider those issues. Defendants first assert that the evidence of plaintiffs marijuana use was relevant only to defendants’ affirmative defense of comparative fault. If so, the jury’s finding that defendants were not negligent means that it never reached the issue of comparative fault. In that case the evidence of plaintiffs marijuana use would be harmless as a matter of law if it was relevant only to defendants’ affirmative defense. See, e.g., Hatfield v. Gracen, 279 Or 303, 311, 567 P2d 546 (1977) (failure to give instruction on the plaintiffs contributory negligence harmless when jury found the defendant not negligent); Stiles v. Freemotion, Inc., 185 Or App 393, 400-02, 59 P3d 548 (2002), rev den, 335 Or 504 (2003) (erroneous instructions on inherent risks of skiing in action against maker and seller of snowboard related only to the plaintiffs comparative fault and were harmless); Wohlers v. Ruegger, 58 Or App 537, 541-42, 649 P2d 602 (1982) (only possible prejudice from admission of complaint in another case related to the plaintiffs comparative negligence, and the jury found that the defendant was not negligent).

The problem with defendants’ argument is that, as their own argument to the jury shows, the evidence of plaintiffs marijuana use was directly relevant to the fundamental *465question of whether defendants’ negligence was a cause of plaintiffs injuries. If the jury found that, as plaintiff testified, he entered the wooded area from the Whoopee side at a point where there were no ropes or other warnings, it might well have found defendants negligent in a way that harmed plaintiff. However, if it found that he entered from the Reservoir Hill side, where there were ropes the entire way, such a finding was impossible under plaintiffs specification of negligence. The only relevant evidence in plaintiffs favor on that crucial point came from his own testimony. Thus, it was important for defendants to give the jury a reason to discount that testimony as unreliable.

At the very close of his argument to the jury, defendants’ attorney told the jury:

“I’m sure that the plaintiff truly believes what he says. We’re not saying he’s a liar or he’s not, you know, some bad person. I’m sure that’s what he believes. But what we don’t—all we can do is look at this scene. There are no physical facts that confirm what he says. We know—it’s up to you to evaluate the evidence from the expert—but we know that he smoked marijuana that morning, and we know there’s evidence that he was impaired. We don’t—you know, maybe that affected his perception, his time, distance, vision and so forth. Maybe he doesn’t, because of that, because of whatever reason, doesn’t have a good recollection of what happened.”

(Emphasis added.) Defendants, thus, asked the jury to infer that plaintiffs use of marijuana could have affected his ability to know what he was doing and to remember what he did, and, thus, could have led him to believe something that was not true. The jury could conclude from the quantitative test, with the significance that Griffin gave it in his testimony, that plaintiffs ability accurately to remember what happened was impaired. Thus, the evidence of the quantitative test, together with the meaning that Griffin gave it, could have affected the verdict in a way that the instructions and evidence at issue in the cases on which defendants rely did not. Defendants’ first assignment of error is not a basis for reversing the grant of a new trial.

Defendants next assert that plaintiff failed to preserve any error because he withdrew his motion in limine rather than asking for a continuance when defendants *466informed the court of the second urine test. Defendants argue that plaintiff thereby waived any claim of surprise. They also point out that plaintiff failed to object to the admission of the test result on the grounds that he raised in his motion for a new trial. The difficulty with defendants’ argument is that plaintiff did not need to show that the trial court erred during the trial in order to seek a new trial on the grounds that he raised. There was thus no claimed error for him to preserve.

What a party must do during trial in order to move for a new trial afterwards varies depending on the ground on which the party relies. None of the grounds on which plaintiff relies has an implicit or explicit preservation requirement, and it would be inconsistent with at least some of them to impose such a requirement. For example, a claim of newly discovered evidence under ORCP 64 B(4) is necessarily based on evidence that the party discovered after the trial had ended or, at the least, could not have presented during the trial. It is usually impossible to preserve that ground for a new trial during the trial itself. In contrast, the two grounds for a new trial on which plaintiff did not rely both require preservation, either implicitly or explicitly, because they involve actions that occurred at the trial and that could be grounds for an appeal. A motion under ORCP 64 B(5), based on “[insufficiency of the evidence to justify the verdict or other decision, or that it is against law,” implicitly requires a previous motion for a directed verdict or a peremptory instruction raising the sufficiency issue. Arena v. Gingrich, 305 Or 1, 8 n 1, 748 P2d 547 (1988); Edward D. Jones & Co. v. Mishler, 161 Or App 544, 564-66, 983 P2d 1086 (1999). A motion under ORCP 64 B(6), based on “[e]rror in law occurring at the trial and objected to or excepted to by the party making the application,” expressly requires that the error be preserved. Because plaintiff did not rely on those grounds, and because he does not assert—and does not need to assert—that the trial court erred during the trial, he did not have to preserve any alleged error.

It is true that a party that is aware dining trial of one of the grounds for a new trial that is listed in ORCP 64 B(l) to (4) may not speculate on the result by failing to bring the issue to the court’s attention, intending to use the issue to seek a new trial if the outcome is unfavorable. See, e.g., *467Turman v. Central Billing Bureau, 279 Or 443, 450, 568 P2d 1382 (1977) (trial court did not err in denying defendant’s motion for new trial when defendant had ability during trial to learn from its own files of alleged juror misstatements but did not raise issue until after verdict). However, nothing in the record suggests that plaintiff speculated on the outcome in this case. The trial court expressly stated that plaintiff did everything possible to raise the issues on which he relied in his motion for a new trial; according to the court, doing more would have verged on being obstructionist and unprofessional.

ORCP 64 B(3), which permits the trial court to grant a new trial because of accident or surprise that ordinary prudence could not have guarded against, does have a kind of preservation requirement when the issue is surprise. The general rule is that a party that is surprised at trial may not use that surprise as a basis for seeking a new trial unless the party moved for a continuance to provide an opportunity to respond to the surprising event. See State v. Gardner, 33 Or 149, 152-53, 54 P 809 (1898); see also Arbogast et al v. Pilot Rock Lbr. Co., 215 Or 579, 594-95, 336 P2d 329 (1959) (applying principle to motion to reopen case in suit in equity). Plaintiff did not seek a continuance and, therefore, was not entitled to a new trial on the ground of surprise.

Defendants also argue that the trial court erred in ruling that their action in issuing a subpoena for the result of the second, quantitative test without complying with the requirements of ORCP 55 I (2001)2 was irregular or constituted misconduct, thus implicating both ORCP 64 B(l) and (2). I agree with the majority that there was no irregularity and that defendants did not commit misconduct, although I do not necessarily agree with all of its analysis.

The decisive question, then, is whether the trial court correctly granted a new trial on the ground of newly discovered evidence. Defendants argue that the supplemental opinion that Brady gave concerning the significance of the quantitative test was not newly discovered evidence under ORCP 64 B(4). Defendants point out that Brady testified at *468trial that he had been unable to find any studies that correlate total cannabinoids with impairment. They also argue that plaintiff did not act with reasonable diligence in seeking the quantitative test before trial. The nature of the majority’s analysis requires me to discuss those points, and the facts on which they are based, at some length. In short, the problem with defendants’ first argument is that Brady gave only negative testimony at trial—that he was unable to find relevant studies—while after the trial he was able to show positively that there was no scientific basis for using the quantitative test to indicate impairment. The problems with their second argument include that, before trial, the appropriate person in the records department at the hospital expressly denied that there was a quantitative test and that the evidence permitted the trial court to find that Brady worked diligently to determine the meaning of that test once he learned of it but did not complete his research until the trial was over.

Brady described the relevant events in his affidavit. He first learned of the quantitative test when defendants referred to it during the argument on plaintiffs motion in limine. When they represented to the court that the quantitative test indicated a level of 603 nanograms in plaintiffs urine, Brady believed that that amount referred to the total metabolites, not to all cannabinoids including inert substances, because most laboratories seek to measure only the metabolites of the active ingredient of marijuana when they perform quantitative tests. Based on that belief, he initially concluded, without time to conduct significant research, that reasonable experts could disagree on whether the test result showed probable impairment.

Brady obtained a copy of the test result during the lunch hour recess but was unable, before the hearing on the motion in limine resumed, to determine the relationship between the total cannabinoid level that the test showed and the levels of active ingredients. As a result, he told plaintiffs attorneys that he could not testify that there was insufficient information from which an expert could conclude that plaintiff was impaired.3 During and after the trial, he continued to *469review scientific literature and consulted with other experts concerning the meaning of the quantitative test. Based on that research, Brady ultimately concluded that there is no scientifically accepted basis for using the test result to determine whether it was likely that plaintiff was impaired at the time of the injury. If he had had the test result a reasonable time before trial, he would have been able to examine it and the relevant literature without the crisis atmosphere that the surprise disclosure created, and he would have reached his conclusions before trial rather than afterwards.

The science on which Brady relied involves the nature of marijuana and the way that the body processes its active ingredient. In summary, his review of the literature led him to conclude that there is no scientifically accepted method forjudging impairment even from a measurement of metabolites of the active ingredient. There are significant individual differences in the speed and other characteristics of the metabolism of the active ingredient and in the amount that is available in any particular sample of marijuana. Brady was unable to find any scientific study that would allow someone to use a measure of total cannabinoids, rather than metabolites of the active ingredient, to determine impairment. For the reasons that I describe later in this opinion, the trial court could conclude, based on Brady’s post-trial affidavit, the studies that accompanied it, and on Griffin’s response, that there was no scientific basis for the opinion that Griffin gave.

The majority appears to hold that Brady’s post-trial opinion was not newly discovered evidence, although it is not entirely clear on this point. At times it appears to confuse the question of the admissibility of Griffin’s testimony at trial with the question of whether Brady discovered new evidence after trial. That confusion is reflected in the majority’s decision to discuss the admissibility of Griffin’s testimony based primarily on the evidence at trial, when the trial court ultimately rejected it because of the newly discovered evidence that plaintiff presented after trial. In order to avoid that confusion, I will first explain why the post-trial evidence was newly discovered. After doing so, I will then discuss the trial court’s decision to grant a new trial based on that evidence.

*470The majority denies that Brady reached an opinion after trial on the basis of full research that was substantially different from what he was able to say before the trial ended. Rather, the majority argues that Brady either did or should have known before the end of the trial substantially everything that he described after trial. In order to respond to the majority’s arguments, I will describe the development of Brady’s opinion from before the trial to his final conclusion after the trial was over. The trial court could have found, contrary to the majority’s position, that Brady exercised reasonable diligence throughout his research. The majority also does not recognize the crucial aspect of Brady’s post-trial opinion in the trial court’s judgment, which was that it deprived Griffin’s evidence of scientific support and made that evidence inadmissible. Thus, Brady’s post-trial opinion does not simply contradict Griffin’s opinion concerning the weight to give the quantitative test in determining whether plaintiffs use of marijuana impaired him at the time of his injury. Rather, Brady’s post-trial analysis, and Griffin’s inadequate response, permitted the trial court to conclude that the jury should not have heard anything whatever about plaintiffs use of marijuana. The trial court could have found that nothing that Brady said during trial had that effect.

Brady gave three opinions during the course of the trial and post-trial proceedings. His first opinion, which is contained in his pretrial affidavit, applied only to the screening test, not to the quantitative test; he did not know about the quantitative test when he executed that affidavit. Because a screening test is fundamentally different from a quantitative test, Brady’s first opinion is largely irrelevant to the validity and admissibility of the quantitative test. His second opinion, which is contained in his trial testimony, described the problems with using the quantitative test that he had been able to identify at that time. His final opinion, which is contained in his post-trial affidavit, relied on evidence that he did not use (and, thus, that the court could find he had not yet discovered) at trial and, for the first time, explained why Griffin’s opinion about the quantitative test was without scientific basis.

Before the trial, both plaintiff and Brady reasonably relied on the hospital’s assurance that the only test that it *471had conducted was the original screening test. A screening test simply determines that metabolites of the active ingredient in marijuana are present in the urine; it does not attempt to determine the amount of those metabolites, of the active ingredient, or of other cannabinoids. Brady’s pretrial statements applied only to that screening test.4 Thus, in his affidavit in support of plaintiffs motion in limine, Brady concluded that it was not possible to determine “from the type of urine test that was done for Mr. Mitchell”—that is, from the screening test—whether plaintiff was mentally or physically impaired at the time of his injury. He pointed out that the screening test could determine only whether a person ingested marijuana within the previous 72 hours. It could not determine whether the person was impaired at the time of the test or at any other particular moment. Brady then added that it was also impossible to determine from the other evidence regarding plaintiffs marijuana use whether he was mentally or physically impaired at that time.

The majority quotes those and other paragraphs of the pretrial affidavit without recognizing that they refer only to the screening test and other information that was available before trial; they have nothing to do with Brady’s later conclusions concerning the quantitative test. Indeed, Brady’s statements in his pretrial affidavit were contingent on his belief that no quantitative test existed and that it was too late to perform one. His pretrial affidavit, thus, showed only that the screening test and other available information were insufficient to determine plaintiffs impairment; it said nothing about the usefulness of a quantitative test for that purpose. For those reasons, the pretrial affidavit does not detract from the trial court’s finding that Brady’s post-trial opinion concerning the quantitative test was newly discovered evidence.

*472An example of the majority’s misunderstanding of the pretrial affidavit is its discussion of paragraph 9. In that paragraph Brady first stated that it was not possible to determine from the other evidence (that is, from the evidence other than the screening test) whether plaintiff was mentally or physically impaired at the time of the accident. He then added, in a sentence that the majority italicizes:

“No reliable scientific method, and no method commonly accepted in the scientific community, allows such a determination from the facts that are available without also knowing the precise details regarding the potency of the particular marijuana and the precise amount ingested.”

Quoted in 195 Or App at 453 (emphasis in majority). The majority reads that sentence as stating that there is no test to correlate impairment with the amount of active ingredient in a person’s urine and suggests that Brady should have been aware of the quantitative test when he made it. Two fundamental things about that quotation show that the majority is wrong. First, Brady emphasized the inability to determine impairment “from the facts that are available”—that is, from the facts as he knew them at the time of the affidavit, before he learned of the quantitative test.5 His affidavit did not question the possibility of correlating impairment with the amount of active ingredient, if that had been known. Second, Brady suggested that knowing the potency of the marijuana and the precise amount ingested would make it possible to determine impairment. However, in his post-trial affidavit he indicated that it is not possible to correlate the amount of active ingredient in the blood with impairment in the same way that it is possible to correlate the amount of alcohol in the blood with impairment. Thus, Brady’s subsequent research led him to a conclusion that was different from what he had suggested before trial.6 Contrary to the majority’s suggestion, the quotation indicates that Brady did not know before trial what he discovered afterwards.

*473When Brady learned of the existence of the quantitative test, he in effect told plaintiffs lawyers that the information in his pretrial affidavit did not apply to that test and that he would have to do additional research. He conducted that additional research during and after the trial. His first discussion of what he found came three days later when he testified as the last witness in the trial. That testimony was based on what he had discovered to that point; in effect, it was a progress report.

The focus of Brady’s testimony was explaining the difficulty of using a measurement of total cannabinoids to determine impairment. He explained that total cannabinoids include all of the material in the marijuana plant, not simply the unknown small portion that is the active ingredient, and pointed out that the research on which Griffin relied in his testimony did not measure total cannabinoids. When plaintiffs lawyer asked Brady if he was aware of any studies that connected the measurement of total cannabinoids to a person’s level of impairment at a particular time, he stated that he had done his best to look around and knew of no study that correlated total cannabinoids with the question of impairment. He did not testify that studies showed that it was impossible to make that correlation. He also stated that an expert could not determine from the test result whether a person was impaired at a specific time before the measurement was taken.

Brady’s testimony, thus, was limited in significant respects. He simply disagreed with Griffin about the meaning of the quantitative test of total cannabinoids. His primary point was that a measurement of total cannabinoids says little or nothing about the amount of the active ingredient or its metabolites or about the person’s impairment at a specific time in the past. He did not assert that Griffin’s opinion was without a scientific foundation or that it was impossible to determine impairment from a quantitative measurement of metabolites rather than of total cannabinoids.* 7

*474Brady’s final conclusions came in his post-trial affidavit, which plaintiff filed in support of his motion for a new trial. That affidavit contained important statements that Brady had not previously made and that were essential to the trial court’s decision to grant the motion. The trial court could infer that, if Brady had known during trial what he said in his affidavit, he would have included those things in his testimony; it could therefore infer that Brady did not reach the crucial conclusions until after the trial was over. In the affidavit Brady, explained that, “[d]uring and after Mr. Mitchell’s trial,” he reviewed the quantitative test result, reviewed scientific literature, and consulted with other experts about whether it was possible to use a measurement of total can-nabinoids to determine whether plaintiff was impaired at the time of his injury. Brady’s crucial conclusion was that “there is no accepted scientific basis for using the test result * * * to determine whether Mr. Mitchell was likely to have been impaired at the time he was injured.”8

The post-trial affidavit includes several factual statements that are in neither the pretrial affidavit nor Brady’s testimony at trial. The trial corut could find that they are, at least in part, the result of his post-trial research: (1) The timing and level of the excretion of marijuana metabolites into the urine varies dramatically among individuals. (2) Because of those variations it is not possible to correlate even a quantitative test of metabolites in the urine with the amount of active ingredient in a person’s blood at some previous time. (3) Even if it were possible to correlate a urine test for the metabolite with the amount of active ingredient in the blood at a particular time, it would still be necessary to correlate the amount of the active ingredient with impairment in the way that the amount of alcohol in the blood can be correlated with impairment. (4) The United States Centers for Disease Control and Prevention, referring to tests for metabolites, has advised that a urine test alone cannot indicate performance impairment or assess the degree of risk associated with a person continuing to perform tasks. (5) Given those *475facts, together with the lack of any study that would enable an expert to determine impairment from a measure of total cannabinoids, Brady was highly doubtful that there was any scientific method that could lead to an informed conclusion about impairment based on a measure of total cannabinoids in a urine sample.

Brady’s post-trial statements thus went significantly beyond his skepticism in his testimony at trial about using a test of total cannabinoids—the great majority of which are inert—to determine impairment. As a result of his post-trial research, his skepticism about Griffin’s testimony had changed to certainty that it was without scientific basis.

In his post-trial affidavit Brady also relied on the information that he had originally described in his pretrial affidavit and in his trial testimony. He concluded, based on all of the information that he described, that the quantitative test of total cannabinoids failed to provide a scientifically valid basis for forming an opinion about whether plaintiff was impaired at the time of his injury. He also described his reasons for believing that Griffin’s analysis was totally lacking in any scientific basis. In doing so he repeated that there is no scientific method for getting from the amount of total cannabinoids in plaintiffs urine at a later time to the amount of active ingredient in his blood at an earlier time. However, he now emphasized that even a measure of the total metabolites would not be a basis for a reliable estimate of the amount in plaintiffs mine. Finally, he described why the literature in Griffin’s file did not support Griffin’s opinion.9

The post-trial affidavit, thus, contained a substantial amount of new information, most significantly that it is not possible to use even a urine test for total metabolites to determine impairment. That new information was crucial to Brady’s opinion that Griffin’s testimony was without scientific foundation. Brady’s opinion on that point, in turn, was the basis for the trial court’s conclusion that it should have excluded Griffin’s testimony, along with all other evidence of plaintiffs marijuana use, from the trial. And that conclusion, *476finally, was the fundamental reason that the trial court granted the motion for a new trial. The trial court could properly have found that Brady did not discover, and in the exercise of reasonable diligence could not have discovered, that new information before the end of the trial.

The majority discusses much of the evidence that I have described but places a different significance on this evidence. The majority’s errors fall into three basic categories. First is its failure to follow the applicable standard of review. Because the trial court granted plaintiffs motion, our standard of review requires us to resolve all of those factual issues in favor of plaintiff. The majority, however, finds the facts itself without regard to whether the trial court could have found different facts. It does not recognize that, in order to hold that the trial court erred in determining that Brady’s post-trial opinion constituted newly discovered evidence, we must be able to conclude that there was no evidentiary support for that determination.10

Second, the majority appears to argue that, because the scientific research on which Brady ultimately relied existed in published form before the trial, his failure to find it before the end of the trial is in itself proof that he did not exercise reasonable diligence in seeking it. Nothing in the record supports the majority’s assumption that the mere fact that a study is published means that an expert should be able to find it instantaneously, immediately distinguish it from all other possibly relevant studies, examine its methodology and results, and thereupon describe its precise significance to the issue at hand.* 11 There is no rule of law that establishes a *477reasonable time for completing those tasks, and there is no evidence that the time that Brady took was unreasonable. The development of Brady’s views in itself permits the finding that he continued to discover additional information during the period between when he learned of the existence of the quantitative test and when he prepared his post-trial affidavit. It also permits the finding that his decisive conclusions did not come until after the trial had ended. On this record, that is sufficient to support a finding that he acted with reasonable diligence.

Finally, the majority treats statements that plaintiffs attorney made during the trial as showing that the evidence that Brady presented after the trial was not newly discovered. However, the attorney’s statements are argument, not evidence. Plaintiffs attorney hoped to convince the trial court of his position based on the evidence that was then in the record. The trial cotut, however, rejected those arguments at the time, thus indicating that it believed that there was insufficient evidence to support them. Plaintiff discovered the evidence that convinced the trial court only after the trial ended; that was when the trial court accepted arguments that it had previously rejected. The fact that the court responded differently to plaintiffs arguments after trial from how it responded during trial indicates that it believed that plaintiff presented new evidence after trial.

The trial court concluded that, if it had had the benefit of Brady’s opinion before or during trial, it would not have allowed Griffin to testify concerning plaintiffs impairment and that it would not have permitted any evidence of plaintiffs use of marijuana. It then granted the motion.

In reviewing a decision to grant a new trial under ORCP 64 B(4), we apply the well-established criteria for determining whether to grant a trial based on newly discovered evidence in light of the facts that the trial court could have found. With one minor exception, those criteria have not changed for over a century and are the same under the rule as they were under the former statute. The Supreme Court restated them most recently in State v. Arnold, 320 Or 111, 120-21, 879 P2d 1272 (1994):

*478“(1) It must be such as will probably change the result if a new trial is granted;

“(2) It must be such as, with reasonable diligence could not have been discovered before or during the trial;

“(3) It must be such that it cannot, with reasonable diligence, be used during trial;

“(4) It must be material to an issue;

“(5) It must not be merely cumulative;

“(6) It must not be merely impeaching or contradicting of former evidence.”

(Footnote omitted.) See also Oberg v. Honda Motor Co., 316 Or 263, 272, 851 P2d 1084 (1993), rev’d on other grounds, 512 US 415, 114 S Ct 2331, 129 L Ed 2d 336 (1994) (rule is materially identical to previous statute; relies on decisions under statute to interpret rule); State of Oregon v. Davis, 192 Or 575, 579, 235 P2d 761 (1951) (stating same criteria under predecessor to former ORS 17.610(4)); State v. Hill, 39 Or 90, 94-95, 65 P 518 (1901) (establishing criteria).

Although motions for a new trial on the ground of newly discovered evidence are generally disfavored, Davis, 192 Or at 579, each motion must rest on its own particular facts and circumstances, and the court should grant a new trial if the undisputed facts would probably lead an ordinarily reasonable person to a different conclusion from the one that the jury reached. Watrous v. Salem Brewery Ass’n, 151 Or 294, 302, 49 P2d 375 (1935). With those considerations in mind, I examine the newly discovered evidence on the motion for new trial in light of the applicable criteria.

Brady’s opinion that it is not scientifically possible to determine from the quantitative test whether plaintiff was impaired, and the reasoning on which he relied, meet the final three criteria for newly discovered evidence. Whether plaintiffs marijuana use impaired his ability to perceive and remember what happened was material to the issue of whether defendants were negligent in a way that caused him damage. The evidence was not cumulative because there was no other evidence in the record that effectively challenged Griffin’s opinion. The testimony that Brady gave during trial *479was both weaker and qualitatively different; at most it showed a disagreement between experts, not that Griffin’s opinion was without scientific validity. Finally, the evidence did not merely impeach or contradict former evidence. Bather, it showed that the former evidence was not admissible and thus should not have been presented at trial.

The second and third criteria relate to plaintiff’s diligence in obtaining the evidence. The facts on those points are uncontested. As a result of the hospital’s failure to disclose the quantitative test to plaintiff when he asked about it, before trial plaintiff believed that the screening test was the only test in existence. Brady was prepared before trial to testify concerning the evidentiary value of that test. After defendants surprised plaintiff by disclosing the quantitative test, Brady conducted additional research during and after the trial. He ultimately discovered the studies that allowed him to reach the conclusions that he described in his affidavit in support of plaintiffs motion for a new trial. The trial court could find that plaintiff acted with reasonable diligence but was still unable to discover the evidence before or during trial; thus, the court could find that the evidence could not have been used during trial.

These facts distinguish this case from somewhat similar cases involving newly discovered expert testimony in which the Supreme Court affirmed the trial court’s decision not to grant a new trial.12 Marshall v. Martinson, 264 Or 470, 506 P2d 172 (1973), was an automobile accident case. The crucial issue was which driver was in his proper lane at the time of the accident. Some of the evidence on that point concerned whether the defendant’s automobile collided with a guardrail before or after hitting the plaintiffs vehicle. Before trial, experts for both the plaintiff and the defendant examined the vehicles with that issue in mind and found pieces of creosoted wood from the guardrail that each claimed supported his particular opinion. After a verdict for the defendant, the plaintiffs expert again examined the defendant’s *480vehicle and found additional evidence that, he asserted, gave further support for his opinion. The plaintiff sought a new trial based on that evidence. Id. at 476. The Supreme Court pointed out that the plaintiffs expert had had every motive and opportunity to discover the evidence in question before the trial. For that reason, the court could not say that the trial court abused its discretion in denying the motion for a new trial. Id. at 477-78.

Arnold was a prosecution for child sexual abuse.13 The CSD social worker who had interviewed the child and was a primary prosecution witness testified that she had earned an associate’s degree at Citrus College in California and that she had audited seven classes at Southern Oregon State College (SOSC). She had not listed those experiences in her employment application, and defense counsel cross-examined her at length concerning the discrepancy. During the trial, defendant’s investigators contacted Citrus College and learned that it had no record of the investigator ever having attended it under either her married name or her birth name. The college stated that it would confirm that information only if it were required to do so by a subpoena. The defendant’s counsel knew those things during trial but did not bring them to the attention of the prosecutor or the trial court, nor did it make any other effort to introduce them into evidence. After the jury returned a guilty verdict, the defendant moved for a new trial, asserting that the investigator had lied about her qualifications. The defendant pointed to the information from the community college and also pointed out that SOSC had no record of the investigator’s attendance at the classes that she described and that the investigator did not even have a high school diploma, as she had claimed on her employment application. Arnold, 320 Or at 113-16.

The trial court denied the motion, emphasizing that the defendant discovered the evidence of the witness’s false testimony in the middle of the trial and could have presented it to the jury. Because the defendant learned about the evidence during trial, it should have notified the court and sought a continuance in order to secure the evidence for use *481at the trial. The court apparently recognized that the defendant discovered the other evidence concerning the witness’s background after trial, but it concluded that that evidence would not have probably changed the result. Arnold, 320 Or at 116-17. On appeal, we reversed, holding that the defendant’s inability to challenge the witness directly on cross-examination, together with the difficulties involved in obtaining an out-of-state witness during the trial, justified the defendant’s failure to seek a continuance to obtain the evidence from Citrus College. State v. Arnold, 118 Or App 64, 69, 846 P2d 418 (1993), rev’d, 320 Or 111, 879 P2d 1272 (1994).

On review, the Supreme Court reversed our decision and affirmed the trial court’s judgment. The primary issue that it considered was whether evidence discovered during trial can be newly discovered evidence for the purpose of a motion for a new trial. The foundation of the Supreme Court’s opinion was its conclusion that the defendant had in fact discovered the evidence during trial. The Supreme Court held that evidence discovered during trial cannot be newly discovered evidence. It emphasized that a party seeking a new trial must show that the party could not, acting with reasonable diligence, have discovered and produced the evidence at trial. Arnold, 320 Or at 119-20. It held, consistently with the trial court’s findings, that the evidence in Arnold was not newly discovered under that rule.

Marshall and this case are similar in that in each an expert discovered additional evidence after the trial. However, in Marshall the expert had a full opportunity to discover the evidence before trial and had every reason to do so. The expert did not conduct a truly diligent search until after the verdict against the plaintiff In contrast, in this case, despite plaintiffs diligence in conducting discovery, Brady did not know that the quantitative report existed until the very beginning of the trial. As soon as Brady learned about the report, he began doing the necessary research to evaluate its meaning in light of existing scientific principles and studies, but he was not able to complete that research and reach his conclusions until after the trial ended. He acted with diligence that the expert in Marshall did not use.

*482Arnold and this case are similar in that in each the parties began investigating the issues that they raised after trial while the trial was in process. However, in this case Brady did not discover the crucial scientific information until after the trial. After trial, not during it, Brady learned that there is no basis for using a measurement of the metabolite to determine impairment; after trial, not during it, he could say that there was no scientific basis for Griffin’s conclusions concerning the quantitative test. Unlike Arnold, in this case Brady did not discover, and could not have discovered, the crucial evidence dining trial. When the trial ended, plaintiff still did not know what Brady would discover or when he would discover it. Plaintiff specifically did not know that Brady would conclude that the quantitative test was as inadmissible as was the screening test.

That is what his plaintiff’s attorney told the court:

“[W]e knew we had been surprised by the additional drug result, but we didn’t know we had been harmed by it. We didn’t know that given some additional time to consider and analyze this result, to look at the literature and so on, we would have [been] able to conclude, you know, this is really no more probative than anything else in this case, we can still get this marijuana evidence excluded. We didn’t know that the surprise had hurt us. * * * As time went on, and we heard testimony unfold, and we looked at literature in the course of and after trial, our opinion on that issue changed.”

(Emphasis added.) The majority quotes that statement but does not understand it. 195 Or App at 454. What plaintiffs attorney told the court is that plaintiff at first did not believe that the surprise had hurt him because he did not believe that he could have gotten the quantitative test excluded even if he had known of it before trial. What changed after the trial was that plaintiff learned from Brady that he could have had all of the marijuana evidence excluded despite the existence of the quantitative test; at that point, and not before, plaintiff knew that the additional drug result had hurt him. Plaintiff could not have used the evidence that he knew during the trial for substantially the same purpose—getting all of the evidence of marijuana use excluded—as he used the evidence that he discovered after trial. That evidence, thus, was newly discovered evidence under Arnold, and the trial court acted *483consistently with that decision when it granted plaintiffs motion.

The remaining question is whether Brady’s opinion meets the first criterion for newly discovered evidence, that it would probably change the result of the case. The foundation of the majority’s opinion appears to be its conclusion that the trial court erred in concluding that Brady’s post-trial opinion rendered Griffin’s trial testimony inadmissible under the rules concerning scientific evidence. Not only is that conclusion wrong, but it resolves an issue that defendants did not raise, that the parties have had no opportunity to brief to us, and that is therefore not before us on appeal. A party that does not raise an issue in its opening brief has waived that issue. See State v. Jones, 184 Or App 57, 60 n 2, 55 P3d 495 (2002). The majority goes beyond the function of an appellate court when it reaches out to decide an issue that the appellant neither raised nor briefed.

“A party may make many arguments in the trial court, in order to ensure that it has raised all issues, but may decide on appeal to focus on those that it thinks are most likely to be successful. Except for errors of law apparent on the face of the record, ORAP 5.45(2), it is not our role as an appellate court to ignore an appellant’s choice.”

State v. Rivas, 100 Or App 620, 622, 788 P2d 464, rev den, 310 Or 122 (1990).

Defendants made six assignments of error, none of which involves the correctness of the trial court’s ruling that Brady’s post-trial evidence made Griffin’s trial testimony inadmissible. Defendants cited many cases in their opening brief; none deals with evaluating the admissibility of scientific evidence.14 Plaintiff did briefly discuss the admissibility of scientific evidence in response to defendants’ fifth assignment of error and in support of their third assignment of *484error on cross-appeal, but in both cases the issues were necessarily limited to the evidence at trial.

In defendants’ fifth assignment, they asserted that plaintiffs motion for a new trial was actually based on ORCP 64 B(5), which provides for a new trial if the evidence was insufficient to support the verdict. Although under the rule the only issue is the sufficiency of the evidence at trial, plaintiff responded in part by defending the trial court’s post-trial exclusion of Griffin’s testimony. Defendants did not respond to that point in their reply. In their third assignment on cross-appeal, plaintiff argued that the trial court should have stricken Griffin’s testimony during trial. In responding, defendants relied in part on the leading Supreme Court case on scientific evidence, State v. O’Key, 321 Or 285, 899 P2d 663 (1995), in defending the validity of his testimony. Again, they did not attack the trial court’s post-trial decision, which was different from its decision during the trial. Even if defendants had done so, their attack would not have been in support of any assignment of error that they made and, thus, there would be no basis for us to consider it.

The majority seems to recognize those facts but argues that, because the trial court ruled on the scientific validity of Griffin’s testimony as part of granting a new trial under ORCP 64 B(4), and because defendants assign error to the grant of a new trial under that provision, we cannot decide whether the grant of a new trial is correct without deciding whether the trial court’s ruling is correct. 195 Or App at 458. The majority thus suggests that assigning error to a trial court decision automatically raises all issues involved in that decision, whether or not the appellant discusses or otherwise refers to them in its brief. That position is inconsistent with accepted appellate practice, as we explained in Rivas. In Rivas the state appealed an order suppressing evidence. The trial court did not explain the basis for its decision. In our original opinion we held that the trial court erred as to some of the evidence because the officer had consent to be where he was at the time that he arrested the defendant and found that evidence. State v. Rivas, 99 Or App 23, 781 P2d 364 (1989). On reconsideration defendant pointed out that, although the state had made that argument *485to the trial court, it did not make it on appeal. We therefore reconsidered our original decision, held that we erred in discussing the issue of consent, and affirmed the trial court as to that part of the evidence. Rivas, 100 Or App at 622-24. If, as the majority now holds, a challenge to a trial court ruling necessarily encompasses a challenge to every decision that is a foundation for that ruling, our decision on reconsideration in Rivas—and the normal requirement that a party raise the issues that it wants us to consider under its assignments of error—is wrong. I do not believe that we should lightly make so radical a change in appellate procedure, and I see no reason to do so here.

The majority also suggests that, because ORCP 64 B is a rule of procedure, cases such as Miller v. Water Wonderland Improvement Dist., 326 Or 306, 951 P2d 720 (1998), require use to reach issues that the parties did not raise. Those cases, however, simply hold that the parties cannot, by the arguments that they choose to make, limit our construction of statutes and administrative rules in a way that prevents us from construing them correctly. Nothing about this case involves the construction of ORCP 64 B(4); the construction of that rule is already well established. The issue is, rather, the application of the rule to this case. In that context the normal rule that an appellant must raise an issue before we will consider it applies.

Even if the majority were correct that we must decide an unraised issue, its result is wrong. The majority appears to rely primarily on Jennings v. Baxter Healthcare Corp., 331 Or 285, 14 P3d 596 (2000), in which the court discussed the criteria for evaluating the validity of novel scientific evidence. In Jennings, the trial court rejected the testimony of an expert neurologist who would have testified, based on his evaluation of 45 women with similar conditions, that the plaintiffs neurological disorder was related to her silicone breast implants. The neurologist had not reached any conclusions about the mechanism that caused the condition, nor had he published his conclusions. The Supreme Court held that his testimony was nevertheless admissible because he followed proper scientific techniques in developing it.

*486Jennings has little if anything to do with this case. Griffin did not rely on any research—published or unpublished—that he had conducted in reaching his conclusions. He did not assert that his testimony was based on novel theories. Rather, he relied on his general expertise in the field of drug examination and on several published research studies with which he had not been personally involved. The strength of his opinion depends on the strength of those studies. Rather, the relevant case is O’Key. In O’Key the Supreme Court evaluated the admissibility of the Horizontal Gaze Nystagmus test that an officer had used in determining that the defendant had driven while under the influence of intoxicants. The court took the opportunity to establish general criteria for the admission of scientific evidence. In doing so, it emphasized that “trial courts have an obligation to ensure that proffered expert scientific testimony that a court finds possesses significantly increased potential to influence the trier of fact as ‘scientific’ assertions is scientifically valid.” O’Key, 321 Or at 293. The court noted that the United States Supreme Court has referred to this role as a “gatekeeper” function. Id. at 301, quoting Daubert v. Merrell Dow Pharmaceuticals, 509 US 579, 113 S Ct 2786, 125 L Ed 2d 469 (1993). It also held that a trial court must not limit itself to determining whether a reasonable person could conclude that evidence is scientifically valid; it must decide the validity issue itself. Id. at 307 n 29.

In O’Key the Supreme Court discussed the criteria that the Court described in Daubert for the admission of scientific evidence under the Federal Rules of Evidence, which are similar in these respects to the comparable Oregon rules. It noted that the first requirement is that the scientific evidence must be “pertinent to the issue to which it is directed.” O’Key, 321 Or at 302. It then discussed the factors that the Court had suggested for determining whether evidence is truly scientific. Id. at 303-05. However, those factors do not come into play until the evidence passes the first requirement, that it be pertinent.

In this case that first requirement is decisive. Brady’s post-trial opinion makes it clear that Griffin’s evidence is not pertinent to the issue to which it is directed— *487whether the measurement of the total cannabinoids in plaintiffs urine indicated that he was impaired at the time of his injury. The thrust of Brady’s post-trial opinion provides strong scientific grounds for concluding that it is impossible to use that test for that purpose. Griffin’s post-trial affidavit supporting his testimony at trial does not confront Brady’s evidence. It consists of copies of three research studies on the relationship between marijuana and physical impairment and Griffin’s statement, unsupported by any scientific research, that the normal ratio between total cannabinoids and metabolites of the active ingredient is between 6 to 1 and 10 to 1. None of the studies that Griffin cited involves a measurement of either total cannabinoids or metabolites of the active ingredient in the urine, and none suggests a relationship between such a measurement at one time and impairment at another. The studies show that marijuana use can lead to impairment under certain circumstances, but they do not show that those circumstances are comparable to the circumstances in this case.

One study was a National Highway Traffic Safety Administration study that measured impairment in people who received measured doses of the active ingredient in marijuana or of a combination of the active ingredient and sufficient alcohol to bring their blood alcohol content to .04 percent. The test subjects began driving 30 minutes after receiving the drugs. There is no indication that they were tested for the amount of cannabinoids or metabolites in their urine or blood. To the contrary, the report commented that “there is no reliable and readily available laboratory marker for THC impairment as there is for alcohol.” This study, if anything, supports Brady’s opinion. A second study examined impairment in college students who were heavy users of marijuana—that is, who had used marijuana at least 22 of the 30 days before the test. Although the students had can-nabinoids in their urine, there was no attempt to correlate the amounts with their impairment. In any case, there is no evidence that plaintiffs marijuana use put him in the same category—or close to the same category—as the tested students.

*488The final study involved airline pilots who smoked a known dose of marijuana and then performed various activities on flight simulators. The study found residual effects that, in some pilots, lasted up to 24 hours, although the effects on a number of specific activities often disappeared after four or eight hours. The fundamental problem with the study for the purposes of this case is that it provides no information on whether the level of cannabinoids in plaintiffs urine has any probative value in determining whether his use of marijuana impaired him at the time of his injury. The evidence does not indicate whether the dose that plaintiff received was comparable to the doses that the pilots received, and the study does not indicate the level of cannabinoids or metabolites in the pilots’ urine.

For these reasons I would hold that the trial court correctly held that, given Brady’s post-trial evidence, and considering what Griffin provided after trial to support his trial testimony, Griffin’s testimony was not pertinent to the issues to which it was directed and thus was not proper scientific testimony. The remaining question is whether excluding Griffin’s testimony, along with all other evidence of plaintiffs use of marijuana, would have affected the outcome of the trial. In its decision the trial court emphasized the impact that evidence of marijuana would have on the jury:

“I would conclude as a trial judge that you get marijuana into the case, if it shouldn’t be there, and it’s just such powerful poison that it’s going to be impair, substantially impair the plaintiff’s right to get a fair determination of the facts of the case.”

It therefore granted the motion for a new trial despite some uncertainty about the strength of plaintiff’s liability case. The trial court acted within its discretion in reaching that conclusion and granting the motion. The trial court was in the best position to evaluate the effect of the improper admission of evidence of marijuana plaintiffs use on the jury’s decision, and I cannot say that its conclusion was unreasonable. Although the court did not expressly state that the result would probably have been different without that evidence, defendants did not ask it to decide that issue expressly. The trial court concluded that the evidence was so prejudicial that *489the only remedy was to grant plaintiff a new trial. In doing so it exercised its discretion with care and consideration,15 and I cannot say, in light of the issues as the parties presented them, that it erred in doing so.16

Finally, in their sixth assignment of error defendants attack the trial court’s pretrial grant of plaintiffs motion for partial summary judgment on defendants’ affirmative defense of release. The majority does not reach that assignment. I would reject it, but it is not necessary for me to state my reasons for doing so. As does the majority, I would affirm on plaintiffs cross-appeal.

For these reasons, I respectfully dissent.

concurring.

I agree with the majoritys disposition of the appeal and cross-appeal. Because I concur in the majority’s analysis that plaintiff failed to demonstrate that the alleged “newly discovered evidence” could not, with reasonable diligence, have been presented at trial, see 195 Or App at 447-56, I would not address whether Arnold’s other, conjunctive criteria were satisfied here. Accordingly, I express no view as to the application of those other criteria in this case.

Landau, J., joins in this concurrence.