NOT RECOMMENDED FOR PUBLICATION

File Name: 25a0514n.06

Case No. 24-5852

UNITED STATES COURT OF APPEALS

FOR THE SIXTH CIRCUIT FILED

Nov 04, 2025

KELLY L. STEPHENS, Clerk

) ZACHARY TYLER MARTIN,

)

Plaintiff-Appellant, ) ON APPEAL FROM THE UNITED

) STATES DISTRICT COURT FOR v. ) THE EASTERN DISTRICT OF

) TENNESSEE POLARIS, INC., POLARIS INDUSTRIES, ) INC., and POLARIS SALES, INC.,

) OPINION

Defendants-Appellees. ) Before: GRIFFIN, THAPAR, and MATHIS, Circuit Judges.

THAPAR, Circuit Judge. Zachary Martin was injured in an accident while riding in a Polaris off-road vehicle. He sued Polaris in federal court and lost after a jury trial. He now appeals three of the trial court’s evidentiary rulings. Finding no abuse of discretion, we affirm.

I.

Utility terrain vehicles (UTVs) are motorized off-road vehicles initially designed for farm work. But where most saw farm equipment, others saw a chance for adventure. Manufacturers like Polaris later designed a new kind of UTV “intended solely for thrills,” built to handle rugged terrain at high speeds. R. 14, Pg. ID 46. These vehicles—known as recreational off-highway vehicles (ROVs)—include Polaris’s RZR line, which boasts high-horsepower engines that can reach speeds well above 60 miles per hour. No. 24-5852, Martin v. Polaris, Inc.

That rush comes with risk, however. By design, ROVs are narrow and top-heavy. So they are prone to “rollover” accidents—especially at high speed. Id. at 46–47. And rollover accidents may result in passengers being partially or fully ejected from the open cab, which often leads to severe injuries or fatalities.

Polaris ROVs include safety features intended to lessen these risks. Perhaps the most important of these is the “roll cage”: a steel cage-like structure that partially encloses the passenger compartment. Roll cages are a double-edged sword. Although they prevent passengers from being ejected, in the event of a rollover, the cages can crush passengers’ limbs, leading to amputation and other injuries. To avoid that, Polaris instructs passengers to keep both hands on a “passenger hand hold,” a T-shaped grip located in front of the passenger seat. Id. at 48. The hand hold mitigates the risk that a passenger—or his limbs—will be ejected in the event of a rollover. But unfortunately, that risk became reality here.

On September 17, 2021, Zachary Martin went to Windrock Park in Oliver Springs, Tennessee to ride ROVs with two friends. One of his friends, Houston Adams, owned the vehicle Martin rode in—a two-seater Polaris RZR XP 1000 EPS with significant aftermarket modifications.

Adams’s RZR was altered significantly from factory conditions before he bought it. In fact, he purchased it because of its modifications, specifically because he thought Polaris’s standard roll cage was “ugly.” R. 71-4, Pg. ID 1470–71. So Adams found an RZR with an aftermarket roll cage, several inches lower than factory height. And that wasn’t the only change. The seats were lowered to compensate for the aftermarket cage so that riders’ heads wouldn’t hit the bars of the lower cage. Plus, Adams opted for a model with bigger wheels and tires. All told, these changes seated Adams’s riders lower relative to the doors and roll cage.

-2- No. 24-5852, Martin v. Polaris, Inc.

Martin was riding in that modified RZR when the accident happened. Martin and Adams left for an early-morning ride and met a third friend along the way. The group arrived at Windrock and hit the trails. Adams drove, with Martin in the passenger seat. Near the top of Windrock Mountain, the trio hit a dead end and decided to turn around, so Adams attempted a “donut.” R. 71- 4, Pg. ID 1443. But in doing so, he “turned around too fast and tipped the [Polaris] over.” R. 181, Pg. ID 8870. Martin dropped the passenger hand hold and was partially thrown out of the vehicle. His right arm was pinned under the roll cage and later required amputation below the elbow.

Because of his injuries, Martin sued Polaris Inc., Polaris Industries Inc., and Polaris Sales Inc. in the Eastern District of Tennessee. He brought state-law claims seeking compensatory and punitive damages for strict products liability, breach of express and implied warranties, and negligent misrepresentations. The case proceeded to a six-day jury trial. The jury found for Polaris on all of Martin’s claims. Martin timely appealed.

On appeal, Martin challenges three of the district court’s evidentiary rulings. We review all three for abuse of discretion. Burley v. Gagacki, 834 F.3d 606, 617 (6th Cir. 2016). “[A]n abuse of discretion occurs when the district court relies on clearly erroneous findings of fact, improperly applies the law, or employs an erroneous legal standard, or when we are firmly convinced that a mistake has been made.” Id. (citation omitted). But erroneous rulings don’t always require reversal. We reverse a jury’s verdict only if the error “was not harmless; that is, only if it affected the outcome of the trial.” Cummins v. BIC USA, Inc., 727 F.3d 506, 510 (6th Cir. 2013).

II.

Martin alleges that the district court committed three evidentiary errors. First, he challenges the district court’s exclusion of information Martin offered about failed government

-3- No. 24-5852, Martin v. Polaris, Inc. efforts to regulate ROVs as unduly prejudicial. Second, Martin claims the district court improperly excluded as unreliable certain opinions about ROV design from Martin’s expert witness. Finally, he argues the district court incorrectly allowed Polaris’s expert to provide opinions at trial that he omitted from pretrial expert disclosures. But the district court properly excluded the first two categories of evidence, and its admission of the third was harmless. As a result, we affirm.

A.

The district court didn’t abuse its discretion when it granted Polaris’s pre-trial motion to exclude evidence about the U.S. Consumer Product Safety Commission’s (CPSC) failed efforts to regulate ROVs. Even if relevant, the prejudicial effect of Martin’s proposed evidence substantially outweighed its probative value.

In 2009, the CPSC became concerned about the “significant” injuries and fatalities caused by ROV accidents and issued a Notice of Proposed Rulemaking to consider regulatory intervention. Standard for Recreational Off-Highway Vehicles, 74 Fed. Reg. 55495, 55496 (proposed Oct. 28, 2009). The proposal sparked public debate, with participation from the Recreational Off-Highway Vehicle Association (ROHVA), a trade association that includes Polaris. No rule resulted.

Instead, ROHVA developed voluntary safety standards in collaboration with the CPSC. In 2014, the CPSC began to question whether those voluntary standards adequately addressed the risk of rollover accidents. Steering and stability problems made rollovers likely, and existing seatbelt requirements didn’t sufficiently protect passengers if a rollover occurred. See Safety Standard for Recreational Off-Highway Vehicles (ROVs), 79 Fed. Reg. 68964, 68986–87 (proposed Nov. 19, 2014). Among the CPSC’s proposed solutions were passive restraints—some kind of barrier, net, or other physical feature that would keep occupants from being ejected from

-4- No. 24-5852, Martin v. Polaris, Inc. the ROV in an accident. See id. at 68964, 68995. But after the comment period, the CPSC closed the rulemaking without adopting any mandatory safety standards. Indeed, at all times during the design, manufacture, and sale of the Polaris RZR, no CPSC regulation ever applied.

Martin sought to introduce evidence about the attempted rulemakings, including the proposed rules, responsive comments, ROHVA documents, and statements from CPSC members (together, the “CPSC evidence”). Before trial, Polaris moved to exclude all that evidence, arguing it was irrelevant and inadmissible under Federal Rule of Evidence 402 or, in the alternative, that its prejudicial effect outweighed its probative value under Rule 403.

The district court didn’t abuse its discretion by excluding the CPSC evidence. Rule 402 creates a presumption foundational to evidence law: Relevant evidence is admissible; irrelevant evidence isn’t. Fed. R. Evid. 402. Relevant evidence “has any tendency to make a fact more or less probable” if “the fact is of consequence in determining the action.” Fed. R. Evid. 401. “This standard is extremely liberal.” United States v. Householder, 137 F.4th 454, 481 (6th Cir. 2025) (per curiam) (cleaned up).

Martin intended to use the CPSC evidence to show that “not only was the Polaris RZR in this case defective, but Polaris was also on notice of it.” R. 75, Pg. ID 1553. Manufacturer notice could help satisfy one element of Martin’s claim for strict products liability under Tennessee law. See Tenn. Code Ann. §§ 29-28-104, 29-28-105(b). So the evidence may meet Rule 402’s low bar—as the district court acknowledged. See R. 128, Pg. ID 6041 (“[T]here may arguably be some relevance to CPSC or other proposed rules . . . .”).

But the district court didn’t abuse its discretion in finding that the evidence’s “probative value” was “substantially outweighed by a danger of . . . unfair prejudice.” Fed. R. Evid. 403. “[U]nfair prejudice refers to evidence that has an ‘undue tendency to suggest decision on an

-5- No. 24-5852, Martin v. Polaris, Inc. improper basis.’” United States v. Smith, 70 F.4th 348, 353 (6th Cir. 2023) (quoting Old Chief v. United States, 519 U.S. 172, 180 (1997)). And we give district courts, who are closest to the evidence, “wide latitude” in making prejudice determinations. United States v. Gibbs, 797 F.3d 416, 423 (6th Cir. 2015).

All three of the district court’s reasons for exclusion aimed to minimize the risk that the jury would render a decision for an improper reason.

First, the district court suspected the CPSC evidence would confuse the jury about what regulations were in force at the time Adams’s RZR was manufactured. That concern is well- founded. Under Tennessee law, notice of a defect depends on a manufacturer’s knowledge only “at the time the product was placed on the market,” Tenn. Code Ann. § 29-28-105(b), and products liability law may account for regulations in effect at that time, id. § 29-28-104(a). But the CPSC’s proposed rules never went into effect, so they weren’t among those existing regulations.

Second, the district court thought the CPSC evidence would confuse the issue of notice. It believed the evidence would suggest that Polaris knew about general issues with ROVs. And the district court correctly saw a problem with that: Knowing that ROVs are generally dangerous is different from knowing Polaris’s own product—the RZR, with all its included safety features— was unreasonably dangerous in some way. Only the latter knowledge would make Polaris liable under Tennessee law. See Tenn. Code Ann. § 29-28-105. So the district court’s concern was well- founded.1 1 Additionally, the district court allowed Martin to introduce the CPSC evidence “to the extent the evidence demonstrates actual notice to Defendants of a defect in [Adams’s] RZR.” R. 128, Pg. ID 6042. Martin attempted to introduce the evidence for that purpose at trial. The district court carefully considered those attempts and found each one lacked a sufficient foundation indicating the evidence would show actual notice.

-6- No. 24-5852, Martin v. Polaris, Inc.

And finally, the district court thought the CPSC evidence would unfairly prejudice Polaris by suggesting the company ignored legal standards that didn’t even apply to them—or anyone else. This court has recognized “the difficulty (. . . if not the impossibility) of eliminating from the minds of the jurors the persuasive weight of” regulations not in effect at the time of the injury. Vroman v. Sears, Roebuck & Co., 387 F.2d 732, 738 (6th Cir. 1967). All three reasons for exclusion were eminently reasonable.

Balanced against these risks, the CPSC evidence offered little probative value. Information about general (and unadopted) regulations applicable to the entire class of ROVs sheds little light on whether RZRs are defective or unreasonably dangerous. Plus, at best, they could shed light on the condition of standard, factory-issue RZRs. But that’s not what Martin rode in—Adams’s model had significant aftermarket alterations. See supra Part I. So the CPSC evidence risked confusing jurors.

The CPSC evidence logically supports all three inferences identified by the district court. Thus, the district court didn’t abuse its discretion by excluding the CPSC evidence.2 Gibbs, 797 F.3d at 423.

B.

Martin next argues the district court erred by excluding certain testimony from his expert, Dr. Tyler Kress, as unreliable under Rule 702. It did not.

Federal Rule of Evidence 702 governs the admission of expert testimony. District courts play a “gatekeeping role” in applying this rule to potential expert witnesses. Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579, 597 (1993). The “central point” of the Rule 702 inquiry is that 2 Because we find the district court’s exclusion proper, we don’t address Martin’s arguments that the CPSC evidence is self-authenticating and falls under an exception to the rules prohibiting hearsay.

-7- No. 24-5852, Martin v. Polaris, Inc. expert testimony “must be the ‘product of reliable principles and methods’ and must have been ‘reliably applied’ in the case.” United States v. Gissantaner, 990 F.3d 457, 463 (6th Cir. 2021) (quoting Fed. R. Evid. 702). Reliability is a “flexible” question, and the district court has “broad latitude” in both its method of inquiry and its “ultimate reliability determination.” Kumho Tire Co. v. Carmichael, 526 U.S. 137, 141–42 (1999).

The district court specifically excluded portions of Kress’s testimony related to (1) any defects in Adams’s RZR and whether they caused Martin’s injury, and (2) the existence of a safer alternative design.3 Both rulings properly applied the law.

1.

The district court correctly excluded Kress’s testimony about defects in Adams’s RZR as unreliable. The critical question under Rule 702 is not whether an expert’s testimony is “correct,” but rather whether it “rests upon a reliable foundation.” In re Scrap Metal Antitrust Litig., 527 F.3d 517, 529–30 (6th Cir. 2008). The reliability inquiry reflects that “Rule 702 requires more than simply taking the expert’s word for it.” Gales ex rel. Ranson v. Allenbrooke Nursing & Rehab. Ctr., LLC, 91 F.4th 433, 437 (6th Cir. 2024) (cleaned up). Kress’s proposed testimony about Adams’s RZR—the one involved in Martin’s accident—didn’t provide any reliable foundation beyond taking his word for it.

Kress concluded that Adams’s RZR was defective when Polaris sold it and thus caused Martin’s injury. But that conclusion suffers from multiple Rule 702 defects. For one, Kress’s description of his methodology boils down to “Trust me, I analyzed it.” His best explanation is 3 The district court also excluded Kress’s legal conclusion that the RZR was unreasonably dangerous or defective and his testimony about the nature of a manufacturer’s duty to consumers. Martin’s brief discusses only the excluded testimony about defects and alternative designs. Appellant’s Br. at 12, 35–45. So he has forfeited any arguments about the exclusion of other parts of Kress’s testimony. See Island Creek Coal Co. v. Wilkerson, 910 F.3d 254, 256 (6th Cir. 2018).

-8- No. 24-5852, Martin v. Polaris, Inc. that he “review[ed] and rel[ied] on the case-specific materials” and considered how people react to rollover accidents. R. 56-7, Pg. ID 926, 928. But mere “review and analysis,” id. at 928, falls short of the testable, peer-reviewed, widely accepted scientific methodology we usually expect from experts. See, e.g., United States v. Sammons, 55 F.4th 1062, 1072 (6th Cir. 2022) (citing Daubert, 509 U.S. at 593–94). Absent further explanation, Kress asks us to accept his conclusions without insight into any reliable methodology—which Rule 702 forbids. Gales, 91 F.4th at 437.

Even if Kress used reliable methods, he didn’t reliably apply them. Kress opined that “the subject ROV was defective and unreasonably dangerous, and as a result was a cause of Mr. Martin’s injury.” R. 56-7, Pg. ID 929. But he hadn’t inspected the RZR before saying so, even though it had undergone significant aftermarket modifications.4 The “sight unseen” report makes his opinion about causation unreliable. A plaintiff alleging design defects “must trace his or her injury to the defect.” Brown v. Crown Equip. Corp., 181 S.W.3d 268, 282 (Tenn. 2005). And that defect must be present “at the time [the product] leaves the control of the manufacturer or seller.” Tenn. Code Ann. § 29-28-108. In other words, plaintiffs must show causation—and show the product wasn’t altered in a way that would break the causal chain. See id. How could Kress possibly draw that causal conclusion without inspecting Adams’s heavily modified RZR?

The district court also found that Kress’s failure to “specifically and clearly identify” any defect in the vehicle at issue made his report unreliable. R. 137, Pg. ID 6140. The closest Kress’s report got was stating that “Polaris should have designed the subject ROV with passive . . . occupant retention” features. R. 56-7, Pg. ID 930. But he doesn’t say which ones. And his deposition didn’t provide much clarity. Kress identified some plausible design features, including 4 Kress finally compared the altered RZR to a factory model the day before he testified at trial, but that was long after the district court reviewed his written reports and ruled on their reliability.

-9- No. 24-5852, Martin v. Polaris, Inc. nets, windows, wrist straps, or different roll cage designs. He elaborated about none of them— nor did he physically test a single one. Rule 702 demands more than such “unsupported speculation.” In re Scrap Metal, 527 F.3d at 529–30. Given Kress’s bare-bones list of possible alternatives and failure to test any of them, the district court reasonably found his opinion about alternative designs unreliable.

Even if his suggested alternatives were specific enough, Kress failed to inspect Adams’s altered model before coming to this conclusion. So he couldn’t possibly address whether any features would have prevented the accident given the RZR’s aftermarket roll cage, lowered seats, and larger wheels and tires. And overall, he provided only general recommendations for ROV design, hardly mentioning Polaris ROVs or its specific RZR line. Given his sparse description of his methodology and the glaring omissions from his “application” of those principles to this case, Kress’s testimony can’t meet Rule 702’s reliability requirement.

Martin’s counterarguments are unavailing. He argues that Kress was “clearly qualified to opine” about defects and safety features based on his general qualifications and experience in relevant fields. Appellant’s Br. at 38. That may be true. But experience alone can’t satisfy Rule 702’s requirement of “reliable methods, reliably applied.” Sammons, 55 F.4th at 1072. Even the most qualified experts “must explain how that experience leads to the conclusion reached, why that experience is a sufficient basis for the opinion, and how that experience is reliably applied to the facts.” Fed. R. Evid. 702 advisory committee’s note to 2000 amendment. Kress’s vague analysis did none of that.5 5 The district court is not the only court to exclude Kress’s expert testimony as unreliable. See, e.g., Lyons v. Leatt Corp., 322 F.R.D. 327, 343–44 (N.D. Ind. 2017); Kough v. Wing Enters., Inc., No. 3:12-cv-250, 2015 WL 164609, at *8 (E.D. Tenn. Jan. 8, 2015); Myers v. Ill. Cent. R.R. Co., 679 F. Supp. 2d 903, 916 (C.D. Ill. 2010).

- 10 - No. 24-5852, Martin v. Polaris, Inc.

2.

The district court was also right to exclude Kress’s testimony about safer alternative designs. Of course, evidence about feasible alternative designs is relevant to determining whether a product is unreasonably dangerous. See, e.g., Potter v. Ford Motor Co., 213 S.W.3d 264, 269 (Tenn. Ct. App. 2006); Tenn. Code Ann. § 29-28-105(b). But the district court found Kress’s testimony unreliable because he “d[id] not specify what safer alternatives he propose[d]” or “how the specific model at issue here could have been more safely designed except to generally suggest features.” R. 137, Pg. ID 6141. The district court made no factual or legal errors in concluding Kress’s vague and general opinions about alternative designs weren’t reliable.

On appeal, Martin argues that Kress specifically identified not just one, but two safer alternative designs. He claims that “side window nets are at the heart of all Kress’s reports” and that Kress “opines in no uncertain words that a passive side barrier would have prevented Martin’s injury.” Appellant’s Br. at 40.

Start with window nets. Kress’s first report suggested that Polaris should make its optional window net attachments a standard feature—if Polaris believed them necessary to make its vehicles safer. R. 56-7, Pg. ID 930–31. That’s different from Kress stating they were necessary. Other than that bare mention, Kress merely recommends additional unnamed safety measures. Although he repeatedly states Polaris should “incorporate[] sufficient design features to adequately protect riders,” he doesn’t specifically identify window nets in those recommendations. Id. at 931. And despite nets being “at the heart” of his testimony, Kress didn’t test them on Adams’s altered RZR or any standard model. Kress also failed to address potential issues of “feasibility or cost,” like the nets’ possible impact on visibility. R. 137, Pg. ID 6141. Instead, he put any visibility issues with the nets on Polaris. That’s fatal. In Tennessee products-liability cases, the jury must

- 11 - No. 24-5852, Martin v. Polaris, Inc. eventually determine “whether [an expert’s proposed] alteration would negatively have affected” the product’s “safety or performance.” Johnson v. Manitowoc Boom Trucks, Inc., 484 F.3d 426, 432 (6th Cir. 2007). So to the extent Kress believed the nets were necessary, he needed to address the concerns that were raised about visibility. He didn’t. We’ve similarly upheld exclusion when a purported expert witness conducted no testing and failed to consider possible costs. See, e.g., id. at 433; Brown v. Raymond Corp., 432 F.3d 640, 647 (6th Cir. 2005). The district court didn’t err when it likewise found Kress’s testimony unreliable because of these failures.

Kress’s opinions about side retention barriers were also deficient. It’s true that he stated a different design would have prevented Martin’s injury “in no uncertain words.” Appellant’s Br. at 40. But that’s the problem—while he was certain of his conclusion, he provided no explanation for it. And that explanation is precisely what we need experts for. It may be true, as Kress wrote, that “[i]t would have been feasible to design and/or guard against” Martin’s injury with some kind of side barrier. R. 56-7, Pg. ID 929. But Kress gave the district court little reason, beyond his education and credentials, to think that conclusion was reliable or based on any methodology. On appeal, Martin fares no better, arguing that the district court should have admitted Kress’s testimony because it “is based on common sense.” Appellant’s Br. at 43 (emphasis in original). While common sense is always important, standing alone, it doesn’t equate to “reliable methods, reliably applied.” Sammons, 55 F.4th at 1072.

Finally, Martin argues that “[r]ejection of expert testimony is the exception, rather than the rule.” Appellant’s Br. at 36 (quoting In re Scrap Metal, 527 F.3d at 530). That’s correct. But favoring admission doesn’t mean judges can shirk our responsibility to find that expert testimony “is properly grounded, well-reasoned, and not speculative before it can be admitted.” Fed. R. Evid. 702 advisory committee’s note to 2000 amendments (emphasis added). And that threshold

- 12 - No. 24-5852, Martin v. Polaris, Inc. question includes ensuring the testimony is based on reliable methods. Martin argues that Kress’s testimony is the sort of “shaky but admissible” evidence a district court should admit for the “traditional and appropriate means of attacking” it: “[v]igorous cross-examination, presentation of contrary evidence, and careful instruction on the burden of proof.” Appellant’s Br. at 36 (quoting Daubert, 509 U.S. at 596). However, Kress’s testimony lacked a sufficient foundation to be admitted in the first place. So Daubert’s presumption in favor of subjecting “shaky but admissible evidence” to the traditional adversarial process doesn’t help Martin.

Affording “broad latitude” to the district court’s reliability determinations, we find no abuse of discretion. Kumho Tire, 526 U.S. at 142.

C.

Finally, Martin objects to the district court’s rulings at trial that permitted Polaris’s expert, Gary Rogers, to testify about the aftermarket alterations of Adams’s RZR. Martin contends that this testimony violated Federal Rule of Civil Procedure 26(a)(2)(B), which requires parties to disclose all expert witnesses and provide their written reports before trial. Martin argues that Rogers’s testimony “ambushed” him at trial principally because he spoke about scratch marks on Adams’s RZR, a topic that didn’t appear in his report. Appellant’s Br. at 22, 47–49. We agree that Rogers’s testimony violated Rule 26, but its admission was harmless.

Rule 26 requires parties to disclose expert testimony at least 90 days before trial, including “a complete statement of all opinions the witness will express and the basis and reasons for them.” Fed. R. Civ. P. 26(a)(2)(B)(i), (D)(i). The rule also requires rebuttal testimony to be disclosed no more than 30 days after disclosure of the testimony it addresses. Fed. R. Civ. P. 26(a)(2)(D)(ii). And experts must supplement their reports with material changes and additions, including opinions they offer in depositions. Fed. R. Civ. P. 26(e)(2). Absent court permission, Rule 26 disclosures

- 13 - No. 24-5852, Martin v. Polaris, Inc. must be in writing and served on the other party. Fed. R. Civ. P. 26(a)(4). Even then, a party’s failure to comply with Rule 26 may be excused if the failure “is substantially justified or harmless.” Howe v. City of Akron, 801 F.3d 718, 748 (6th Cir. 2015) (cleaned up).

Rogers’s initial report omitted a key insight that he later mentioned in his deposition and at trial. His testimony included highly technical scratch-mark analysis of Adams’s modified RZR. The testimony centered on extensive testing Rogers performed on both Adams’s RZR and a factory model, with accompanying photos. Rogers simulated the “roll angle” and post-crash positions of a factory model RZR and Adams’s altered RZR. R. 180, Pg. ID 8698. His first key takeaway was that the factory RZR roll cage left a gap between the cage bar and the ground that wasn’t present in the altered version. His second key takeaway was that scratch marks on the altered RZR’s roll cage confirmed that it caused Martin’s injury. Because the aftermarket cage rolled on gravel, the metal was covered in scratch marks, except for one gap: the spot where Martin’s arm hit the gravel instead. Together, his testimony suggested that the aftermarket roll cage hit the ground—when Polaris’s factory model wouldn’t have—except where Martin’s arm was crushed beneath it. Rogers offered this conclusion for the first time in his deposition testimony, not in his written expert report.

Rogers’s failure to disclose his scratch-mark evidence in his report violated Rule 26’s requirement that expert testimony be timely disclosed in writing. The scratch-mark testimony didn’t appear in his initial report, and he failed to provide a written supplement.6 But the violation was harmless. 6 Polaris claims this trial testimony didn’t violate Rule 26 because it appeared in Rogers’s pretrial deposition. Appellee’s Br. at 46. That’s irrelevant to Rule 26, which requires disclosures in writing. See Fed. R. Civ. P. 26(a)(4).

- 14 - No. 24-5852, Martin v. Polaris, Inc.

Five factors determine whether a Rule 26 violation was harmless: “(1) the surprise to the party against whom the evidence would be offered; (2) the ability of that party to cure the surprise; (3) the extent to which allowing the evidence would disrupt the trial; (4) the importance of the evidence; and (5) the nondisclosing party’s explanation for its failure to disclose the evidence.” Howe, 801 F.3d at 748 (quotation omitted). These factors favor Polaris.

Rogers’s deposition reveals that none of the objected-to testimony surprised Martin. Dr. Kress, Martin’s expert, submitted a supplemental report containing previously undisclosed opinions and served it to Polaris on July 1, 2024. So, in Rogers’s July 24 deposition, he responded to Kress’s newest report. More importantly, he fully discussed the opinions Martin argues he didn’t disclose. See Ex. 1, 31:8–16. Rogers previewed his entire argument: The original Polaris roll cage “leaves a gap between the ground and” cage bars, see id. at 31:11–12, and a small “interrupt[ion]” in the post-crash scratch pattern on Adams’s RZR “is potentially where Mr. Martin’s arm was” crushed, id. at 32:13–15. His trial testimony mirrored the deposition, albeit in more depth. But it’s natural to expect that an “expert will supplement, elaborate upon, explain and subject himself to cross-examination upon his report.” Thompson v. Doane Pet Care Co., 470 F.3d 1201, 1203 (6th Cir. 2006). Neither the content nor the extra detail could come as a surprise to Martin, so the first Howe factor weighs in Polaris’s favor.

Because the testimony was disclosed in fact—even if in the wrong form—Martin had the ability to cure any surprise before trial even started. But he chose not to. Martin’s counsel didn’t object at the deposition. On the contrary, his counsel asked Rogers multiple questions about the scratch-mark testimony. So his claim of a “bait-and-switch” as soon as Rogers began discussing the scratch-mark evidence at trial isn’t supported by the record. R. 180, Pg. ID 8696; see also id. at 8646 (objecting because “the word ‘scratch’ isn’t even in his report”). In fact, Martin’s counsel

- 15 - No. 24-5852, Martin v. Polaris, Inc. objected only on the second day Rogers testified about scratch marks at trial. Nor did Martin file a pretrial motion to exclude Rogers’s deposition or the testimony it contained in the three weeks between the deposition and trial. Martin also had the ability to cross-examine Rogers at trial, which helps remedy any surprise. Howe, 801 F.3d at 749; cf. RJ Control Consultants, Inc. v. Multiject, LLC, 100 F.4th 659, 670 (6th Cir. 2024). Thus, the ability to cure also weighs in Polaris’s favor.

The untimely disclosure didn’t disrupt trial because Rogers’s deposition occurred before trial, giving Martin the opportunity to cure it through his decisions at trial. Howe teaches that untimely disclosures that occur before a trial commences don’t necessarily disrupt it, as the affected party can still cross-examine the witness or introduce competing evidence. 801 F.3d at 749. Because the late disclosure didn’t disrupt trial, this Howe factor favors Polaris too.

The importance of the evidence can favor either party. After all, “[t]he more important the proof, the greater the effect of preclusion, but also the greater the harm in tardy disclosure.” Bisig v. Time Warner Cable, Inc., 940 F.3d 205, 220 (6th Cir. 2019) (cleaned up). But where, as here, the affected party claims surprise but takes no steps to cure it—like objecting at Rogers’s deposition or seeking clarification before trial—this factor weighs against the affected party. Id. So the fourth Howe factor favors Polaris.

Finally, the reason for the untimely disclosure favors Martin, but only slightly. At trial, Polaris explained that the delay occurred because the testimony rebutted Kress’s July 1 supplemental report. It’s natural that a new, late-breaking report would garner new responsive testimony. But Rule 26 accounts for such cases by allowing written expert supplements. Fed. R. Civ. P. 26(a)(4), (e)(2). On appeal, Polaris also contends that because Martin’s counsel didn’t object, it believed no supplement was necessary. However, an “honest mistake” doesn’t excuse

- 16 - No. 24-5852, Martin v. Polaris, Inc. noncompliance. See RJ Control Consultants, 100 F.4th at 671. This factor overall favors Martin. But it doesn’t carry much weight. Polaris didn’t engage in the kind of “underhanded gamesmanship” Rule 26 seeks to prevent. Howe, 801 F.3d at 749; see also McHugh v. Olympia Ent., Inc., 37 Fed. App’x 730, 735 (6th Cir. 2002) (per curiam) (“Rule 26 must be read in light of its dual purposes of narrowing the issues and eliminating surprise.”). Polaris’s cards were on the table at Rogers’s deposition, so this factor doesn’t strongly favor Martin.

In sum, four of the five Howe factors favor Polaris. Critically, because Rogers’s trial testimony mirrored his pretrial deposition, Martin wasn’t surprised by the Rule 26 violation. So Polaris’s violation was harmless, which means the district court didn’t abuse its discretion by permitting Rogers to testify.

* * *

All three of Martin’s evidentiary challenges fail. The district court didn’t rely on clearly erroneous factual findings, commit a legal error, or do anything to “firmly convince[]” us “that a mistake has been made.” Burley, 834 F.3d at 617 (cleaned up). Seeing no abuse of discretion, we affirm.

- 17 -