Brandon Stollings v. Ryobi Technologies, Inc.

725 F.3d 753, 92 Fed. R. Serv. 50, 2013 WL 3964477, 2013 U.S. App. LEXIS 16055

• Court: Court of Appeals for the Seventh Circuit
• Decided: August 2, 2013
• Docket: 12-2984
• Status: Published

Opinion

HAMILTON, Circuit Judge.

On May 9, 2007, Brandon Stollings lost an index finger and portions of other fingers in a table saw accident. Stollings sued the saw manufacturer, Ryobi Technologies, alleging that Ryobi defectively designed the saw because it failed to equip the saw with either of two safety features: a riving knife—a small blade that holds the cut in the wood open to prevent kickbacks—and automatic braking technology—a safety system that automatically stops the saw blade upon contact with human tissue. Stollings contends either safety feature would have prevented the accident. A jury returned a verdict in favor of Ryobi. Stollings has appealed.

Stollings argues that the district court made three reversible errors: (1) failing to stop Ryobi’s counsel from arguing to the jury that Stollings’s counsel brought the case as part of a joint venture with the inventor of the automatic braking technology to force saw manufacturers to license the technology, and admitting hearsay evidence to support this improper argument; (2) excluding the testimony of one of Stollings’s expert witnesses; and (3) giving two erroneous jury instructions. We find that Ryobi’s joint venture argument was improper and prejudicial, so we vacate the judgment and remand for a new trial. Because the remaining issues are likely to resurface if the case is retried, we address them and conclude that the court erred in excluding the expert testimony and in giving the jury a sole proximate cause instruction where Ryobi was not asserting a comparative fault defense or blaming a third party.

I. The Improper Attack on Counsel’s Motives

A. The Accident and Power Saw Safety

We address first Ryobi’s improper attack at trial on the motives of plaintiffs counsel, which requires us to provide the background on the accident and power saw safety. Stollings was injured while operating a Ryobi Model BTS20R table saw. The immediate cause of the injury was a common woodworking hazard known as a kickback. A kickback occurs when the kerf, the gap in the wood created by a saw’s cut, closes around the saw blade in such a way that the force of the spinning blade throws the wood back at the user. If the saw operator is holding onto the wood, the unexpected movement can sometimes force the operator’s hand into the spinning saw blade. This is what happened to Stollings.

Saw manufacturers include safety features to help protect users from kickback injuries. Ryobi equipped the saw with a “3-in-l” guard safety system. This safety system has three components: a splitter, anti-kickback pawls, and a blade shield. The splitter is a piece of plastic that rests behind the saw blade to prevent the kerf from closing around the saw. The anti-kickback pawls are serrated pieces of metal attached to the sides of the splitter that rest on the wood as it moves through the cut to prevent the wood from moving backwards. And the blade shield is a piece of plastic that covers the top of the blade to prevent the user’s hands from coming into contact with the blade. This system complied with the applicable guarding standards published by Underwriters Laboratory—a private company that sets industry safety standards—and the applicable federal Occupational Safety and Health Administration regulations.

The 3-in-l system is effective at reducing injuries when used correctly, but it has shortcomings. The principal problem is that many saw users deliberately disable the 3-in-l guard system. There are two reasons for this. The plastic guard makes certain cuts more difficult to complete, and the guard can become clouded by sawdust and other material, thus obstructing the user’s view of the saw blade as it cuts. The 3-in-l system is also interconnected. When a user removes the guard, he must also remove the splitter and the anti-kickback pawls, leaving the saw blade without any kickback protection. That is what Stollings did. Despite warnings on the saw about the dangers of operating the saw without the guard, Stollings removed the guard and operated the saw without the safety protection. Before a reader concludes that this fact decides the case, though, we should note that Ryobi’s former chief engineer testified that he had removed the 3-in-l system on his own home saw and had instead installed a riving knife.

The jury heard evidence that Ryobi could have equipped its saw with two alternative safety features. The first is a riving knife, which is a cheap piece of metal or plastic similar to a splitter. Like a splitter, a riving knife rests behind the blade and holds the kerf open. Unlike the splitter in the 3-in-l guard system, a riving knife is typically positioned closer to the saw blade, making it more effective at preventing kickbacks. Most important, it is independent of the guard system, so the user has no reason to remove it.

The second additional safety feature is an automatic braking system, colorfully known as flesh detection technology. The automatic braking system prevents injury by stopping and retracting the blade at the moment the blade contacts flesh. The technology works by detecting the human body’s electrical current. When an operator’s flesh contacts the blade, the body’s electrical current triggers the safety system, which applies a brake and retracts the blade beneath the cutting surface. The saw stops within a few milliseconds, fast enough in most cases to leave the operator with only a minor, superficial wound. The technology, however, is not cheap. It would add somewhere between $50 and $150 to the cost of a table saw.

Stephen Gass developed the automatic braking system in 1999. Gass patented the technology and then attempted to license it to table saw manufacturers, including Ryobi. Gass and Ryobi entered negotiations over a licensing agreement, but the negotiations fell through and Ryobi never licensed Gass’s technology. Stollings maintains that Ryobi and other manufacturers decided not to license Gass’s technology for fear of product liability exposure on saws that did not have the technology. Ryobi contends Gass’s terms were unreasonable and the technology was too expensive and unproven. In 2005, Gass founded a competing company named SawStop to manufacture and sell table saws that include his automatic braking system. Gass testified at trial as one of Stollings’s expert witnesses about the feasibility and effectiveness of the automatic braking system. He did not ask for or receive compensation for his testimony.

B. The Trial Attack on Plaintiffs Counsel

In addition to the arguments one would expect Ryobi to make—that the saw complied with industry safety standards and that Stollings was responsible for his injury because he failed to use the 3-in-l safety system—Ryobi framed the case for the jury as a joint venture between Gass and Stollings’s attorneys—Mr. Carpinello and Mr. Sullivan—to coerce Ryobi and other saw manufacturers to license and use Gass’s automatic braking technology. The district judge referred to this as Ryobi’s “conspiracy” theory, though the word conspiracy was not used in the presence of the jury.

Ryobi’s attack on the motives of Stollings’s counsel began in its opening statement. More than half of it was dedicated to the argument that the case was being brought by Stollings’s attorneys to intimidate the saw manufacturers. The implication of the argument was that the jury should not let Stollings’s counsel and Gass play them for chumps. Here are some examples from Ryobi’s opening statement:

“The evidence is going to establish that there’s a joint venture between Mr. Carpinello and Mr. Stephen Gass ..., whereby Mr. Carpinello will file product liability lawsuits against manufacturers that don’t pay Mr. Gass a royalty for his patent....”

“Mr. Carpinello has filed over 90 of these product liability lawsuits ... and Mr. Gass is his expert witness in every one of those cases.... ”

“So what we have here is a patent IP case, an intellectual property case, masquerading as a personal injury case.... ”

“There’s something going on below the surface and that’s why, in [this] opening statement, I want you to be aware of what is going on here. So the joint venture that exists in this case is part of an overall strategy to force the manufacturers to pay Mr. Gass for his technology- • • •”

“So what are we talking about here? We’re talking about an attempt to intimidate manufacturers to pay [Gass] a royalty so they don’t have to be sued by Mr. Carpinello all over the country because they’re making a saw that complies with what the design standards required.”

At the beginning of the opening statement, Stollings’s counsel objected, arguing that these insinuations were prejudicial and not based on any evidence. Ryobi responded that it had evidence of the supposed joint venture. The judge permitted Ryobi to continue, informing the jury that opening statements are not evidence. Ryobi’s “evidence” was an article about Gass from The Oregonian newspaper. Stollings had made a valid hearsay objection to the article before trial, but the court permitted Ryobi to use the article during its opening statement.

After the court overruled Stollings’s objection, Ryobi’s counsel continued: “The evidence will establish that there’s a joint venture between Mr. Carpinello and Mr. Gass.... ” Ryobi’s counsel then told the jury about the Oregonian newspaper article and argued that it was evidence of the joint venture. Ryobi’s counsel used a blown-up version of the article to draw the jury’s attention to what Ryobi believed were the critical points. Pointing to a section of the article that said Gass had been approached by products liability lawyers, Ryobi’s counsel said: “Oh, this thing about the joint venture. Gass says he has been approached by lawyers looking to launch product liability suits that ultimately could force companies to license his technology. That’s part of his strategy.” Ryobi’s counsel then concluded: “And that’s the case that Mr. Carpinello has filed here, just like Mr. Gass said, to launch product liability suits that will ultimately force companies to license his technology. Part of a game plan to sue the manufacturers all across the country, have Mr. Carpinello coming after the manufacturer alleging to a jury that the product is defective and unreasonably dangerous .... ”

Throughout the trial, Ryobi continued to link the motives of Stollings’s attorneys with Gass’s desire to persuade saw manufacturers to license his technology. Ryobi’s counsel made repeated references to the number of saw cases in which Stollings’s attorneys were involved, asked witnesses whether the attorneys were involved in other cases in which the witnesses had testified, and read the names of Stollings’s attorneys off of deposition transcripts from other cases.

Ryobi also emphasized the joint venture theme in its closing argument. “What’s going on in this case?” counsel asked rhetorically: “This is that Oregonian newspaper article. Why is that significant? Because these are quotes from Mr. Gass ... and he doesn’t deny these quotes.” Ignoring the facts that the article did not directly quote Gass and that Gass in fact had denied that the article directly quoted him, Ryobi’s counsel continued:

Gass says he has been approached by lawyers looking to launch product liability suits that ultimately could force companies to license his technology.... I didn’t say that. That’s what Gass says. Approached by lawyers to launch product liability suits that ultimately could force companies to license his technology. That’s why he has—has been sitting here with his partner throughout this trial, not charging Mr. Carpinello or Mr. Sullivan to be here. And he has been their expert in all of these cases that they’ve filed against table saw manufacturers ....

By the end of the trial, the district judge realized that Ryobi’s argument had gone too far. The judge had always recognized that Ryobi was entitled to attack the credibility of Gass for bias, but earlier in the trial the judge had said that an attack on the motives of Stollings’s lawyers “would be out of bounds.” At that time the judge did not believe that such an attack had occurred. Later in the trial, after reviewing the transcripts, however, the judge recognized that the attack had also been aimed directly against Stollings’s lawyers.

The judge then concluded that it would likely be reversible error if he did not allow Stollings to rebut the accusations. Tr. at 1653. Stollings asked for free rein to discuss the facts of the other saw cases to which Ryobi referred, while Ryobi argued that any reference to the other saw cases would be unduly prejudicial. The judge’s solution was an instruction telling the jury that, of the cases Ryobi referred to, only one was decided on the merits, and it resulted in a verdict for the plaintiff. Unfortunately, however, the instruction also told the jurors that they were free to consider Ryobi’s “joint venture” theory in reaching their decision. The instruction said in relevant part:

You’ve heard mention of other table saw cases filed by Mr. Carpinello and Mr. Sullivan against various table saw manufacturers in which Dr. Gass served as an expert witness. Only one of those other table saw cases has been tried to verdict. That verdict was in favor of the plaintiff in that case. You are not to consider, discuss, or speculate about which manufacturer was the defendant in that case; nor may you consider, discuss, or speculate about the brand or model of the saw in that case; nor may you consider, discuss, or speculate about where or when that trial took place. You are not to consider the verdict in that case as having any direct bearing on whether the table saw in this case was negligently designed or unreasonably dangerous.... In the rest of the table saw cases, no court or jury has made any decision one way or the other regarding the merits of those cases.... You may consider this information about the other table saw cases solely for the purposes of: 1. Evaluating the weight of Dr. Gass’s testimony; and 2. Considering whether the other table saw cases brought by Mr. Carpinello and Mr. Sullivan in which Dr. Gass served as an expert witness were brought to force table saw manufacturers to license table saw technology. The weight to be given to this information is up to you.

C. Analysis of the “Joint Venture” Argument

Stollings argues that Ryobi’s attacks on the motives of his counsel deprived him of a fair trial, which requires him to show that Ryobi’s argument was improper and that any errors the district court made in permitting improper lines of argument were not harmless. See Fed. R.Civ.P. 61; Perry v. Larson, 794 F.2d 279, 285 (7th Cir.1986) (to require a new trial, error must be “substantial enough to deny [party] a fair trial”). If an argument is improper, five factors are relevant as we consider whether the improper argument deprived a party of a fair trial: (1) the nature and seriousness of the argument, (2) whether the statement was invited by the opposing party, (3) whether the statement could be rebutted effectively, (4) whether an effective curative instruction was given, and (5) the weight of the evidence. See United States v. Kiebig, 600 F.3d 700, 720-21 (7th Cir.2009). 1 Stollings contends that Ryobi’s attack on the motives of his counsel—exacerbated by the admission of hearsay evidence and an improper summation of the evidence during Ryobi’s closing argument—was improper and sufficiently prejudicial to deprive him of a fair trial. We agree.

Ryobi’s argument that plaintiffs counsel brought this suit as part of a joint venture with Gass to force Ryobi to license Gass’s technology was improper. The first problem with this argument is that it is not relevant to any issue in dispute. The suggestion that the ease was an intellectual property case “masquerading as a personal injury case” did not bear on whether Ryobi designed and sold a defective product. How does a statement about counsel’s motive help a jury decide whether there was an injury? A duty? A breach of that duty? Or causation?

The argument worked by directing the jury’s focus away from the elements of the case to an extraneous and inflammatory consideration. It said to the jury, don’t let Stollings’s lawyers trick you into finding Ryobi liable for his injury because this case is really not about Stollings, it’s about Gass and his technology. Such an argument aimed at a party’s counsel is improper and risks depriving the party of a fair trial. See Gruca v. Alpha Therapeutic Corp., 51 F.3d 638, 645-46 (7th Cir.1995) (remanding for new trial; improper for products liability defendant to highlight negligence of third party when third party’s conduct was not at issue in case); Davis v. FMC Corp., 771 F.2d 224, 233 (7th Cir.1985) (same); see also United States v. Xiong, 262 F.3d 672, 675 (7th Cir.2001) (noting attacks on counsel “can prejudice the [opposing party] by directing the jury’s attention away from the legal issues”).

The second problem is that no admissible evidence supported the argument. Ryobi based its accusation solely on an article about Gass published in the Oregonian newspaper, and the article was inadmissible hearsay. The article reported: “Gass says he has been approached by lawyers looking to launch product liability suits that ultimately could force companies to license his technology.” Ryobi used the article to assert that Gass said he was working with product liability lawyers to force manufacturers to license his technology. This is classic hearsay: an out-of-court statement offered to prove its truth—that is, that Gass made the statement. See Chicago Firefighters Local 2 v. City of Chicago, 249 F.3d 649, 654 (7th Cir.2001) (newspaper article was inadmissible hearsay when offered as proof of article’s contents). Federal Rule of Evidence 802 plainly prohibited admission of the article for this purpose.

The rules of evidence prohibit most hearsay evidence because it is so often unreliable. This concern is illustrated by looking at the ambiguities in the Oregonian article. The author did not use quotation marks to indicate that he was quoting Gass, so it is unclear even whether the author was asserting that the conclusion about forcing companies to license Gass’s technology was Gass’s conclusion or the author’s. (On cross-examination Gass denied that the author was quoting him.) It is also ambiguous whether the article said that the lawyers launched the product liability suits for the purpose of forcing the companies to license Gass’s technology or whether such licensing might simply be a consequence of the lawsuits. The statements in the article were made out of court, and the declarant (the reporter) was not available for Stollings to cross-examine. The judge should have excluded the article. 2

We need not decide if the improper admission of the hearsay article and its use in the opening statement were reversible error on their own. Cf. Jordan v. Binns, 712 F.3d 1123 (7th Cir.2013) (finding that some errors in admitting hearsay testimony were harmless in light of entire trial record). This entire line of argument was sufficiently prejudicial to warrant a new trial. The improper statements were prejudicial because they constituted a substantial part of the case, they were not invited by Stollings, they could not be rebutted effectively, an effective curative instruction was not given, and the weight of the evidence was not overwhelming in Ryobi’s favor. See Klebig, 600 F.3d at 720-21.

To explain these factors, first, the attack on plaintiffs counsel was both serious and substantial. Ryobi dedicated most of its opening statement to planting the seed in the jury’s mind that Stollings’s counsel had a suspect motive in bringing this case. Over half of the opening statement focused on the intellectual property consequences of the case and the existence of the alleged conspiratorial “joint venture” between Gass and Stollings’s counsel. During the trial, Ryobi repeatedly drew attention to the number of saw cases plaintiffs counsel had brought, referring to plaintiffs counsel by name and even pointing at them. In closing, Ryobi returned to the theme, reminding the jury of the Oregonian newspaper article and suggesting that the article explained what was really going on in the case. This sustained focus on the motives of plaintiffs counsel in all phases of the trial was substantial.

Next, the attack on counsel was not invited. The district judge originally allowed Ryobi to make the joint venture argument because it understood the argument to be limited to Gass’s credibility and motives. The judge said that “the argument that the defendant is making is certainly unusual and not something that I in my limited experience have seen,” but found that it “goes to Dr. Gass’s motive for testifying and his bias.” Tr. at 1088-89. Because the plaintiff chose to use Gass as an expert, the judge reasoned, Stollings’s counsel had to expect Ryobi to argue that Gass was not credible. Tr. at 1091-92. If the argument had been limited to Gass’s motive for testifying and his credibility, it would have been fair game and we would see no error. The joint venture argument that Ryobi actually made went well beyond a challenge to Gass’s credibility. Nothing in the plaintiffs case invited an attack on the motive of plaintiffs counsel.

Ryobi also argues that Stollings had an adequate opportunity to rebut the attack and chose not to. Because Ryobi first made the comments in its opening statement, Stollings had an opportunity to rebut the charge during his examination of Gass and during closing argument. By declining not to, Ryobi suggests, Stollings concluded that the comments were not a big deal. This argument misses the point. We do not look to whether there was simply an opportunity to challenge the statement to determine whether the opportunity was adequate. Rather the opposing party must have an effective opportunity to rebut the charge. It is hard to imagine what Stollings could have done in this case to dispel the charge that his lawyers were bringing an intellectual property case masquerading as a products liability case. Once the argument was raised, Stollings had to choose between two bad alternatives. He could spend time addressing the argument and risk suggesting to the jury that it was important, or he could do his best to ignore the argument and hope the jury focused on the elements of the case. The fact that Stollings faced this dilemma does not mean that he had an adequate opportunity to rebut the argument.

Nor did the court’s instruction cure the prejudice. After realizing that Ryobi had attacked the motives of Stollings’s lawyers, the judge concluded that a curative instruction was needed. The judge’s solution was to instruct the jury that the only other similar case that was tried to a jury resulted in a plaintiffs verdict, but taken as a whole, the instruction was not effective and may well have made the situation worse. The judge told the jury that the plaintiffs verdict could be used for purposes of considering “whether the other table saw cases brought by Mr. Carpinello and Mr. Sullivan in which Dr. Gass served as an expert witness were brought to force table saw manufacturers to license table saw technology.’ ” This instruction thus incorrectly told the jury that it could consider the motives of Stollings’s lawyers in bringing this case, and it did so after preventing Stollings’s lawyers from offering additional evidence about those cases that could have taken some of the sting out of the attack and the instruction. Under no circumstances was counsel’s motive a proper argument for the jury to consider. The motives of Stollings’s lawyers had no bearing on any element of the case.

Finally, the evidence did not so strongly favor Ryobi that the error was harmless. Ryobi had some strong evidence: the saw had safety features that complied with federal and industry standards; the automatic braking technology was expensive; and Stollings admitted that he failed to use the safety guard, which could have prevented the accident. Yet there was also evidence in Stollings’s favor. Ryobi’s former chief engineer testified that he did not use the 3-in-l guard system on his own saw because he believed it was dangerous when it fogged up. He also testified that he installed a riving knife on his own saw because he believed it was safer than the Ryobi guard. Unlike the automatic braking technology, riving knives are both cheap and proven, and they have been widely accepted by European regulators. And some of Stollings’s evidence in favor of the automatic braking system had been excluded erroneously by the court.

On balance, the evidence supporting the verdict was not overwhelming, so the joint venture attack on counsel may well have influenced the outcome of the trial. Accordingly, we find that Ryobi’s statements about Stollings’s lawyers’ motives deprived Stollings of a fair trial. He is entitled to a new one. We next address the remaining issues in the appeal in the interest of judicial economy. These issues are almost certain to arise in a new trial. If we deferred their resolution, we would risk upsetting a second verdict in a subsequent appeal. We turn now to the exclusion of plaintiffs proposed expert testimony.

II. Exclusion of Expert Testimony

A. Factual Background

Stollings planned to offer expert testimony from John Graham, a scholar who served from 2001 to 2006 as the director of the Office of Information and Regulatory Affairs in the federal Office of Management and Budget and is now the dean of the Indiana University School of Public and Environmental Affairs. Graham was to testify that including automatic braking technology on all power saws would be socially beneficial because the average cost of accidents per saw that would be prevented by the technology exceeded the cost of the braking system. Graham estimated that saws that lack the automatic braking technology cost society an average of $753 in accident costs over the lifetime of the average saw. He concluded that it would therefore make economic sense to install the technology on all saws if the cost of doing so was less than $753 per saw. The district judge held before trial that Graham’s testimony should be excluded under Federal Rule of Evidence 702 as not reliable or relevant for the trier of fact. We conclude that this ruling was an abuse of discretion.

Graham submitted an expert report two years before trial laying out his opinion, methodology, supporting data, and qualifications as required by Federal Rule of Civil Procedure 26(a)(2)(B). Graham described how he reached his conclusion by calculating the average cost of a table saw injury (medical costs, lost wages, pain and suffering, and litigation costs). He then multiplied this figure by the likelihood that a saw user would suffer an injury over the lifetime of a saw, yielding an estimate of the societal costs of injury per table saw. He then discounted this number by the effectiveness rate of the automatic braking technology, which he estimated to be 90 percent. The result was an estimate of the average societal costs of injuries per saw that occur because the technology is not installed. The report provided the sources for all of these inputs. Of particular relevance, Graham estimated the effectiveness of the braking technology to be 90 percent based on Gass’s testimony that the technology worked in the “vast majority” of instances.

Ryobi moved before trial to exclude his testimony for a number of reasons, including, in passing, the reliability of the 90 percent estimate. In response to Ryobi’s motion, the district court focused on the basis for Graham’s assumption that the automatic braking technology was 90 percent effective at preventing injuries. At the end of the hearing, Stollings’s counsel received the court’s leave to make a submission “to answer [the court’s] specific fact questions, the 90 percent and the source of that....”

Graham submitted a supplemental report ten days later that provided a more detailed justification for the 90 percent assumption. The new justification estimated the rates of three common forms of failure (injuries when the saw is turned off, injuries when the detection technology is disabled, and injuries that occur when the operator’s hand is moving too rapidly into the blade for the technology to activate in time). Again Graham estimated that the technology would prevent 90 percent of accidents.

The court then held another hearing and decided to exclude Graham’s testimony. This ruling was explained in an opinion issued after trial. The court struck Graham’s supplemental report as untimely because it was submitted one month before trial and it was based on information that was available at the time Graham prepared his initial report. The court noted that when it allowed Stollings to make a submission, it should have been limited to an explanation of how Gass’s testimony supported the 90 percent assumption and not a new justification for the assumption. The court then concluded that Graham’s testimony had to be excluded under Federal Rule of Evidence 702 because the fact that the 90 percent effectiveness input was not reliable rendered Graham’s entire opinion unreliable. The court also agreed with a magistrate judge’s opinion in a similar case finding that Graham’s testimony was not relevant because the cost to society of saw accidents did not speak to the utility of Ryobi’s specific saw design. Accordingly, the court prohibited Graham from testifying at trial. Stollings argues that the exclusion of the expert report was erroneous.

B. Whether Graham’s Testimony Satisfied Rule 702

Expert testimony is admissible at trial under Federal Rule of Evidence 702 if the testimony is relevant to a fact in issue, is based on sufficient facts or data, and is the product of reliable scientific or other expert methods that are properly applied. 3 The district court is responsible for acting as a gatekeeper to ensure that all admitted expert testimony satisfies the Rule’s reliability and relevance requirements. See Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 592-93, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993). But the district court’s role as gatekeeper does not render the district court the trier of all facts relating to expert testimony. See Daubert, 509 U.S. at 595, 113 S.Ct. 2786 (“The focus, of course, must be solely on principles and methodology, not on the conclusions that they generate.”); Smith v. Ford Motor Co., 215 F.3d 713, 718 (7th Cir.2000) (“soundness of the factual underpinnings of the expert’s analysis and the correctness of the expert’s conclusions based on that analysis are factual matters to be determined by the trier of fact”). The jury must still be allowed to play its essential role as the arbiter of the weight and credibility of expert testimony.

We review a district court’s decision to exclude expert testimony on the grounds that it is unreliable or irrelevant for an abuse of discretion. Jenkins v. Bartlett, 487 F.3d 482, 489 (7th Cir.2007). We conclude here that the district judge’s decision to exclude Graham’s testimony in this case was too great an intrusion into the role of the jury. The testimony was both reliable and relevant and should not have been excluded under Rule 702.

1. Reliability

Expert testimony is permitted to assist the trier of fact with technical issues that laypeople would have difficulty resolving on their own. Expert testimony furthers this purpose only if the expert is in fact providing the jury with genuine expertise. The role of the judge is to ensure that the testimony the jury hears satisfies Rule 702’s reliability requirements: that the expert is using a valid methodology (scientific or otherwise), that there is sufficient data to justify the use of the methodology in the particular case, and that the expert applied the methodology appropriately. See Fed.R.Evid. 702(b)-(d).

Rule 702’s requirement that the district judge determine that the expert used reliable methods does not ordinarily extend to the reliability of the conclusions those methods produce—that is, whether the conclusions are unimpeachable. See Dau bert, 509 U.S. at 595, 113 S.Ct. 2786. An expert may provide expert testimony based on a valid and properly applied methodology and still offer a conclusion that is subject to doubt. It is the role of the jury to weigh these sources of doubt. In Dauberb the Supreme Court expressly envisioned this continued role for the jury when it reminded all that “[vigorous cross-examination, presentation of contrary evidence, and careful instruction on the burden of proof are the traditional and appropriate means of attacking shaky but admissible evidence.” Id. at 596, 113 S.Ct. 2786; see also Walker v. Soo Line R.R. Co., 208 F.3d 581, 586-87 (7th Cir.2000); 29 Charles Alan Wright et. al, Federal Practice and Procedure § 6266 (1st ed.) (observing that “where expert testimony is based on well-established science, the courts generally have concluded that reliability problems go to weight, not admissibility”). As the Second Circuit has explained, trial judges acting as gatekeepers do not assume “the role of St. Peter at the gates of heaven, performing a searching inquiry into the depth of an expert witness’s soul” that would “inexorably lead to evaluating witness credibility and weight of the evidence, the ageless role of the jury.” McCullock v. H.B. Fuller Co., 61 F.3d 1038, 1045 (2d Cir.1995).

Rule 702’s reliability elements require the district judge to determine only that the expert is providing testimony that is based on a correct application of a reliable methodology and that the expert considered sufficient data to employ the methodology. For example, if an expert seeks to testify about an average gross sales price but is going to base the testimony on sales to only a single customer, a court would appropriately exclude the testimony because a single observation does not provide a sufficient basis for calculating an average. See Wasson v. Peabody Coal Co., 542 F.3d 1172, 1176 (7th Cir.2008) (affirming exclusion of testimony on these facts); see also General Electric Co. v. Joiner, 522 U.S. 136, 143-47, 118 S.Ct. 512, 139 L.Ed.2d 508 (1997) (affirming exclusion of expert testimony where expert did not provide basis for claim that studies of cancer incidence in mice supported testimony as to cancer incidence in humans).

The judge should permit the jury to weigh the strength of the expert’s conclusions, provided such shortcomings are within the realm of a lay juror’s understanding. See Metavante Corp. v. Emigrant Sav. Bank, 619 F.3d 748, 762 (7th Cir.2010) (criticism of quality of testimony goes to weight of expert testimony, not admissibility); Smith, 215 F.3d at 718 (soundness of factual basis and correctness of expert’s conclusions are questions for jury). If the judge believes expert testimony is too complex for the jury to appreciate important issues of reliability, such that admitting the testimony would prejudice the opposing party, the judge remains free to exclude such evidence under Rule 403. See ATA Airlines, Inc. v. Fed. Express Corp., 665 F.3d 882, 896 (7th Cir.2011).

In this case, the judge’s exclusion of Graham’s expert testimony on reliability grounds intruded too far into the province of the jury. The judge agreed that Graham correctly employed a valid methodology, but the judge concluded that Graham’s testimony was unreliable because his input for the effectiveness of the automatic braking technology was not sufficiently reliable. Although the 90 percent figure was undoubtedly a rough estimate, it is also clear that Graham’s bottom-line estimate of societal costs of saw accidents was so high that his opinion would have remained essentially the same even if the effectiveness rate were actually quite a bit lower. If the effectiveness rate were 50 percent, for example, Graham’s model would estimate the saved costs of avoided saw accidents to society per saw to be $417—still much higher than Ryobi’s conservative estimate of $150 to add the technology to a saw.

The judge should have let the jury determine how the uncertainty about the effectiveness rate affected the weight of Graham’s testimony. Ryobi was free to use cross-examination to attack the assumption and to ask Graham how altering the assumption would affect his analysis. Graham could then have explained that his bottom-line opinion did not depend on the precise value for the effectiveness of the technology. Lowering the effectiveness rate would simply lower the benefit to society from including automatic braking on every saw, and Graham would likely have been able to provide the jury with the full range of alternative values. A jury should be capable of understanding how the value of the estimate affected Graham’s conclusions. The judge should not have excluded Graham’s testimony on these grounds.

2. Relevance

Graham’s testimony also satisfied Rule 702’s relevance requirement. Whether an issue is relevant in a case is a question of substantive state law; whether the specific evidence offered is relevant to resolving the issue is a procedural question governed by the Federal Rules of Evidence. See In re Air Crash Disaster Near Chicago, Ill. on May 25, 1979, 701 F.2d 1189, 1193 (7th Cir.1983). As we read Illinois law, Graham’s expert opinion is relevant to whether the Ryobi saw was unreasonably dangerous.

Illinois applies both the consumer expectations test and the risk-utility test in design defect cases to determine whether a product is unreasonably dangerous. See Mikolajczyk v. Ford Motor Co., 231 Ill.2d 516, 327 Ill.Dec. 1, 901 N.E.2d 329 (2008). If a product is unreasonably dangerous, then the manufacturer is strictly liable for any injuries caused by the product. See Murphy v. Mancari’s Chrysler Plymouth, Inc., 381 Ill.App.3d 768, 320 Ill.Dec. 425, 887 N.E.2d 569, 574 (2008). Under the risk-utility test, a product is unreasonably dangerous if the risks associated with the product design outweigh the utility of the design. In Calles v. Scripto-Tokai Corp., 224 Ill.2d 247, 309 Ill.Dec. 383, 864 N.E.2d 249 (2007), the Illinois Supreme Court made clear that the inquiry into the risk and utility of a given product is broad. After listing at least fourteen relevant factors for courts to consider, including the “usefulness and desirability of the product—its utility to the user and to the public as a whole,” the court went on to make clear that it did not intend the list to be exhaustive. Id., 309 Ill.Dec. 383, 864 N.E.2d at 260-61. The relevant considerations are many, and it is up to the fact finder to “determine the importance of any particular factor” in each case. Id., 309 Ill.Dec. 383, 864 N.E.2d at 261.

Given their approval of a broad inquiry into all aspects of a product’s risk and utility, we believe Illinois courts would consider the costs of a category of accidents to society a relevant consideration in a product liability suit. This is implicit in the court’s adoption in Calles of the factor that focuses on the “utility to the user and to the public as a whole.” Id., 309 Ill.Dec. 383, 864 N.E.2d at 260. Determining utility to the public as a whole requires a consideration of the costs the product imposes on society as well as the benefits to society. Because the Illinois Supreme Court has taken an inclusive view of the factors a jury may consider, it would likely find a product’s costs to society a relevant consideration. Graham’s testimony is relevant if it would help the jury weigh the Ryobi saw’s costs to society.

Under the Federal Rules of Evidence, testimony is relevant as long as it “has any tendency to make a fact more or less probable” than it would otherwise be. Fed.R.Evid. 401; see also Fed.R.Evid. 702. Graham’s testimony satisfied this liberal relevance standard because it would have helped the jury weigh the saw’s utility by providing the jury with a basis to appreciate the saw’s cost to society, which is relevant under Illinois law. See Daubert, 509 U.S. at 587, 113 S.Ct. 2786 (noting liberal relevance standard).

Ryobi challenges this conclusion by arguing that the cost to society is only one component of the utility of the product to the public and is therefore too attenuated from the question of whether the Ryobi saw’s design benefits outweigh its costs to be relevant to an issue in dispute. Graham’s testimony, taken alone, would not conclusively answer this ultimate question. But expert testimony does not need to be conclusive to be relevant. Smith, 215 F.3d at 718 (“When analyzing the relevance of proposed testimony, the district court must consider whether the testimony will assist the trier of fact with its analysis of any of the issues involved in the case. The expert need not have an opinion on the ultimate question to be resolved by the trier of fact in order to satisfy this requirement.”).

Expert testimony about the average social cost of table saw injuries over the lifetime of a saw is relevant to the jury’s consideration of the social utility of the Ryobi saw. Graham’s testimony would have provided the jury with a probative piece of the evidentiary puzzle. Ryobi was free to offer other pieces of relevant information, such as the diminished utility of the saw caused by the addition of automatic braking technology. It also would have been free to expose the weaknesses in Graham’s testimony to the jury through cross-examination. These objections go to the testimony’s weight, not its relevance or admissibility.

Ryobi also objects to the relevance of Graham’s testimony because it focused on the social cost of table saws generally, not the Ryobi saw specifically. Graham’s testimony rests on an assumption that the Ryobi saw is similar enough to other table saws that the average injury rate and average cost per injury of all saws will provide an accurate estimate for the Ryobi saw. Perhaps this assumption makes Graham’s testimony less powerful than if he had relied solely on an analysis of the specific Ryobi saw model that Stollings used, but this arguable limitation can also be addressed through cross-examination. The fact that an expert’s testimony contains some vulnerable assumptions does not make the testimony irrelevant or inadmissible. Accordingly, the district court erred in preventing Graham from testifying. The cost to society is a relevant consideration under Illinois law, and Graham’s expert testimony was relevant to this issue.

III. Jury Instructions

We turn finally to the jury instructions on unreasonably dangerous products and proximate cause. We review de novo whether the jury instructions stated the law correctly, but we “afford the district court ‘substantial discretion with respect to the precise wording of instructions so long as the final result, read as a whole, completely and correctly states the law.’ ” Lewis v. City of Chicago Police Dep't, 590 F.3d 427, 433 (7th Cir.2009), quoting United States v. Gibson, 530 F.3d 606, 609 (7th Cir.2008), and conveys the correct law “to the jury reasonably well.” Dawson v. New York Life Ins. Co., 135 F.3d 1158, 1165 (7th Cir.1998). Because this is a diversity action, we look to state law to determine whether the instruction properly stated the substantive law; however, federal law governs whether the instruction was sufficiently clear. See Anderson v. Griffin, 397 F.3d 515, 520 (7th Cir.2005). We need not parse here the distinction between substantive law and clarity because federal and state law are in accord. As we explain below, the unreasonably dangerous product instruction was appropriate, but the “sole proximate cause” instruction was likely to confuse the jury in view of Ryobi’s choice not to assert a comparative fault defense.

A. Unreasonably Dangerous Product Instruction

Stollings asked the judge to give the Illinois pattern instruction on what constitutes an unreasonably dangerous product. The pattern instruction reads:

When I use the expression “unreasonably dangerous” in these instructions, I mean unsafe when put to a use that is reasonably foreseeable considering the nature and function of the [product].

Illinois Pattern Instruction 400.06. The judge chose to supplement the pattern instruction with a list of seven factors drawn from the Calles ease, instructing the jury that they “may consider, but are not limited to” the factors. Notably for Stollings, the list did not include a utility or cost-to-society factor. Stollings maintains that it was error to provide the jury with a non-exhaustive list of factors because the list drew the jurors’ attention to some factors at the expense of others.

The instruction properly stated the law and was not unreasonably confusing. The district judge employed the Illinois pattern instruction and supplemented the instruction with several factors that the Illinois Supreme Court approved in its Calles decision. The judge made clear to the jury that the factors were not exclusive and that the jurors were free to consider other factors. This was a correct statement of the law. It was also not confusing because we presume that jurors are able to follow instructions, at least absent reason to believe the instructions are ineffective. See United States v. Lee, 558 F.3d 638, 649 (7th Cir.2009) (“Absent any showing that the jury could not follow the court’s limiting instruction, we presume that the jury limited its consideration of the testimony in accordance with the court’s instruction.”) (quotation omitted). Here the judge directed the jury to several relevant factors and made clear that these were not the only factors for the jury to consider. We think this statement was clear enough both for jurors to understand that they were free to consider all of the relevant evidence and to allow counsel to argue additional factors. This instruction was not error.

B. Sole Proximate Cause Instruction

Stollings also objected to the proximate cause instruction because it contained “sole proximate cause” language. The proximate cause instruction combined two different Illinois pattern instructions— the pattern instruction on proximate cause and the instruction for situations in which someone other than the defendant may have been a cause of the injury. That instruction said: “If you decide that the sole proximate cause of the plaintiffs injury was his own conduct, then your verdict should be for the defendant.” Tr. at 2806. Stollings objected to this instruction, arguing that it was confusing because it could be read to invite the jury to weigh the comparative fault of Ryobi and Stollings. Stollings favored an instruction that made clear that Ryobi was a proximate cause if its negligence was at all responsible for Stollings’s injury.

The issue with the proximate cause instruction requires a bit of background. Illinois is a modified comparative fault juris diction. This means that if a plaintiff is partially responsible for his injury, damages are reduced according to the amount he was at fault as long as he was not more than 50 percent at fault. See 735 111. Comp. Stat. 5/2-1116. For example, if the plaintiff is 30 percent responsible for his injury and the defendant 70 percent responsible, the defendant would be liable to the plaintiff for 70 percent of the damages. But if the plaintiff was more than 50 percent responsible for the injury, then the plaintiff is barred from any recovery.

Prior to trial, Ryobi made a strategic decision to abandon the comparative fault defense. This meant that the comparative fault rule just described would not apply, and Stollings would be entitled to recover all of his damages if Ryobi’s negligence was just one proximate cause of his injury. The effect of this decision was to force the jury to make an all-or-nothing choice: either find Ryobi liable for all of Stollings’s damages or find Ryobi not liable.

The court gave the jury the following instruction on proximate cause:

When I use the expression 'proximate cause,’ I mean a cause that, in the natural or ordinary course of events, produced the plaintiffs injury. It need not be the only cause nor the last or nearest cause. It is sufficient if it combines with another cause resulting in the injury. If you decide that the sole proximate cause of the plaintiff’s injury was his own conduct, then your verdict should be for the defendant. On the other hand, if you decide that the defendants’ conduct was a proximate cause of the plaintiffs injury and if the plaintiff has proved the remaining propositions for his negligence claim or strict liability claim, then your verdict should be for the plaintiff as to that claim.

Tr. at 2806-07 (emphasis added).

The non-italicized text reflects the Illinois pattern instruction on proximate cause. See IPI 15.01. The italicized text is a variation on the pattern instruction for cases in which the defendant argues that a third party is responsible for the injury. See IPI 12.04. The purpose of pattern instruction 12.04 is to make clear to the jury that a negligent defendant is still liable to the plaintiff even if a third party was also at fault for the injury. The comments to the model instruction explain that the “instruction should be used only where negligence of a person who is not a party to the suit may have concurred or contributed” to the plaintiffs injury and “there is evidence tending to show that the sole proximate cause of the occurrence was the conduct of a third person.” Id. (emphasis added). This obviously does not describe this case, in which Ryobi maintained that Stollings—a party to the suit—was entirely responsible for his injury. Stollings argues that the inclusion of the following “sole proximate cause” language made the instruction unreasonably confusing: “If you decide that the sole proximate cause of the plaintiffs injury was his own conduct, then your verdict should be for the defendant.”

We agree with Stollings that the sole proximate cause language was likely to cause undue confusion where no other person’s conduct was at issue and the defendant had abandoned a comparative fault defense. The first three sentences, taken from the Illinois pattern instruction, correctly explain the law on proximate cause. The objectionable “sole proximate cause” sentence is also consistent with the law, in the abstract, at least, if Ryobi had been trying to show that a third party was at least partially responsible for the accident. It is also logically correct that if Stollings was the sole proximate cause, Ryobi would not be a proximate cause. (If Stollings was the only proximate cause, by definition no one else could have been a proximate cause.)

Nevertheless, the instruction was confusing for two reasons. First, the sole proximate cause language in the fourth sentence would draw the jury’s attention away from the conduct of the defendant toward the conduct of the plaintiff. Yet there was no reason for the jury to focus on the plaintiffs conduct in this case because Ryobi abandoned the comparative fault defense. Directing the jury to focus on whether the plaintiff was the sole cause invited the jury to weigh the respective fault of the plaintiff and the defendant and to attempt to determine a sole cause when one may very well not have existed. An Illinois court has warned that the sole proximate cause language is risky because “of its tendency to distract the jury’s attention from the simple issues of whether the [defendant] was negligent and whether that negligence was the cause, in whole or in part, of the plaintiffs injury,” Baker v. CSX Transp., Inc., 221 Ill.App.3d 121, 163 Ill.Dec. 616, 581 N.E.2d 770, 777 (1991) (quotations omitted), and we have similarly found it error to give the sole proximate cause instruction because it “was bound to confuse” even though it was correct as “an abstract proposition of law,” McCarthy v. Pennsylvania R. Co., 156 F.2d 877, 881-82 (7th Cir.1946) (Minton, J.). 4

The second problem with the instruction is that the evidence in the case did not suggest that there was a sole proximate cause of Stollings’s injury. A sole proximate cause instruction was thus likely to be particularly confusing to the jury. See id. at 882 (cautioning against giving sole proximate cause instruction when evidence does not suggest there was a single proximate cause). Stollings’s legal theory was that Ryobi was negligent (or strictly liable) for failing to include alternative safety features on the saw because Ryobi knew that many users did not use the 3-in-l system. Ryobi’s response was that it was not a cause of the injury because Stollings failed to use the safety guard that would have prevented the injury. If we assume for purposes of argument that Stollings’s negligence or strict liability theory is correct, then Ryobi would almost certainly also be at least a proximate cause of the injury. It would be odd to conclude that a saw was designed defectively due to inadequate safety features and also to conclude that the inadequate safety features were not a proximate cause of an injury that was made possible by the inadequate safety features. Because it was unlikely that a jury would find that Ryobi created a defective product but was not a proximate cause of the injury, the sole proximate cause instruction invited confusion here.

Finally, if more is needed to understand that the instruction was confusing, we need only look to Ryobi’s decision to abandon the comparative fault defense. If Ryobi had raised the defense of comparative fault, it would have been liable for only the portion of Stollings’s injury for which it was responsible. Even better for Ryobi, if Stollings was more than 50 percent responsible for his injury, Ryobi would not be liable at all. By forgoing the defense, Ryobi would be liable for all of Stollings’s injury if it was merely one percent responsible for Stollings’s injury. Why then would Ryobi forgo this defense? If the jury correctly applied the law, Ryobi would always be better off with the com parative fault defense. Ryobi’s strategy seems to have depended on confusing the jury.

The district court should not have given the sole proximate cause instruction. Illinois Pattern Instruction 15.01 provided an adequate statement of the law that was unlikely to confuse the jury.

IV. Conclusion

Ryobi’s improper “joint venture” theory argument aimed at Stollings’s counsel prejudiced Stollings and deprived him of a fair trial. Accordingly, the judgment is VACATED and the case is REMANDED to the district court for a trial. We believe Ryobi’s joint venture argument surprised the district judge as much as it surprised Stollings. With hindsight the judge recognized that the argument was improper, but at that point it was too late to remedy the problem. We are confident, though, that Judge Feinerman will handle this issue deftly if it arises again, so Circuit Rule 36 shall not apply on remand.

1 . These factors have been developed in criminal cases like Kiebig to determine whether prosecutorial misconduct prejudiced a defendant. The factors are equally valuable in evaluating civil trials, at least when timely objections are made, as they were here. If there is any difference in the harmless error determination between criminal and civil trials, it is in the strength of the showing required to demonstrate that an error is harmless, not in the relevant considerations.

2 . Ryobi then exacerbated this problem by giving an inaccurate summation of the evidence in its closing argument. In closing, Ryobi’s counsel reminded the jury of the newspaper article and asked rhetorically, “Why is that significant?” To which counsel answered, “Because these are quotes from Mr. Gass.... Mr. Gass tells us in this newspaper article— and he doesn’t deny these quotes.” This statement was incorrect. Gass explicitly denied that the reporter quoted him in the article. Stollings did not object to this point during the closing argument, but by this time damage had already been done and could not have been cured.

3 . To be admissible, the party seeking to admit expert testimony must also disclose the expert witness and submit a written report prepared by the witness that contains among other things "a complete statement of all opinions the witness will express and the basis and reasons for them” and "the facts or data considered by the witness in forming them----” Fed.R.Civ.P. 26(a)(2)(B)(i)-(ii). This requirement ensures that the opposing party has an adequate basis to examine the expert. Fed.R.Civ.P. 26 advisory committee’s note. A party is barred from introducing evidence or testimony that it failed to disclose without substantial justification unless the failure was harmless. Fed.R.Civ.P. 37(c)(1). Ryobi does not argue on appeal that the disclosure was insufficient.

4 . McCarthy addressed the appropriateness of a sole proximate cause instruction under the Federal Employers Liability Act, which holds railroads liable to injured employees if the injury was caused in whole or in part by the railroad’s negligence. See 45 U.S.C. § 51. Under the FELA, the railroad is liable if its negligence is a proximate cause of the employee’s injury regardless of whether the employee was also negligent. This case presents a situation analogous to McCarthy, so our words of caution in that case are equally applicable here.