Arizona Chemical Co. v. Mohawk Industries, Inc.

RAY, J. This appeal and cross-appeal arise from a $70.1 million jury verdict and resulting judgment for the plaintiffs, Mohawk Industries, Inc., and Aladdin Manufacturing Corporation (collectively “Mohawk”), in a lawsuit about defective Unibond-brand carpet. Mohawk, the manufacturer, became aware of the Unibond defects in 2008, when consumers’ claims against Unibond’s lifetime warranty began to rise sharply. After studying the issue, Mohawk developed a hypothesis that the cause of the claims spike was the use of a particular resin in the Unibond backing system. This resin had been formulated and manufactured by the defendant, Arizona Chemical Company, LLC, for this purpose. As Mohawk later told the jury, Mohawk stopped using Arizona’s resin after developing its hypothesis that the resin was causing failures in Unibond’s backing system. The claims rate for newly manufactured Unibond carpet then returned to a rate consistent with the low rate Unibond had historically enjoyed. Nevertheless, Unibond sales plummeted, and Mohawk discontinued the carpet line in 2011. Mohawk sued Arizona on theories of breach of contract, breach of express warranty, and breach of implied warranty of fitness for a particular purpose, seeking recovery for the cost of the relevant claims and the loss of profits from unrealized sales. The jury trial resulted in a judgment for Mohawk on the contract and express warranty claims and for Arizona on the im — 97 — plied warranty claim. Arizona, the appellant and cross-appellee before this Court, raises three issues. We affirm the appeal on the merits of each issue and write to address two of them. In light of this disposition, and at Mohawk’s invitation, we affirm the cross-appeal as moot. In the first issue, Arizona contends that the trial court erred in excluding evidenée of claims spikes experienced by other Mohawk carpet lines contemporaneously with the Unibond spike, where those carpets were manufactured at the same facility as Unibond. Arizona offered this evidence to show possible alternative causes of the backing defects and resulting claims spike and to rebut Mohawk’s expert testimony on causation. We find no abuse of discretion in the trial court’s ruling that the other products’ claims rates were not legally relevant to causation.1 In the second issue, Arizona argues that Mohawk’s evidence of the amount of lost profits was legally insufficient because Mohawk’s damages expert did not consider important variables affecting the amount of revenue Mohawk would have gained from Unibond sales in the absence of the backing failures. Those variables include, among others, reputational damage to Mohawk from the failure of Mohawk’s Ency-cle carpet line, which contemporaneously generated a claims rate far exceeding the peak Unibond claims rate. In the alternative, Arizona argues that it should have been permitted to cross-examine Mohawk’s lost-profits .expert with Mohawk’s ■business records acknowledging these variables as they affected the company’s market share for. commercial carpet, which included Unibond. We conclude that the expert’s failure to address the variables at issue did not render his testimony legally insufficient to support the lost-profits verdict and that -Arizona has not shown an abuse of discretion in the ■ limitation of cross-examination. I. FACTS AND PROCEDURAL HISTORY A. Motion in Limine Before trial, Mohawk moved in limine to exclude evidence and argument concerning the claims spikes for Mohawk products that were not manufactured using Arizona’s resin. Arizona opposed the motion, arguing that this evidence was relevant to rebut Mohawk’s theories of causation and lost profits. Arizona anticipated that, to show causation, Mohawk would rely on proof that the Unibond claims spike coincided with the use of Arizona’s resin. Arizona intended to introduce, evidence of contemporaneous claims spikes related to three other carpet brands manufactured at the same facility as Unibond, in Glasgow, .Virginia. Those brands were known as PVC, U2, and En-cycle.2 Arizona argued that the similar claims spikes experienced by the other carpet lines indicate a likelihood that the rise and fall in the Unibond claims was caused by a factor shared among the four carpet lines, not a factor exclusive to Uni98 bond, such as the use óf Arizona resin. In conjunction with this rebuttal, Arizona intended to introduce evidence that the Glasgow manufacturing facility had experienced organizational issues, poor performance habits, and a lack- of discipline; among employees. Further, Arizona had evidence that Glasgow’s director of quality assurance had been fired in 2008 for poor job performance. Arizona contended that this evidence would show that all the claims spikes were attributable to poor quality control or failure to adhere to manufacturing protocol at the Glasgow facility. For its part, Mohawk argued that the other products’ claims rates were irrelevant because they were distinct products with different manufacturing processes and uh~ related defects. Unibond, a carpet sold for commercial usé, had three defining characteristics: its longevity in the market, from 1971 until 2011; its broadloom classification, meaning that it was produced in long twelve-foot-wide rolls; and its manufacturing process. Unibond’s manufacturing process was relevant not only to the arguments on the motion in limine but also to an understanding of the- causation evidence at trial. '--This carpet was manufactured using a- two-step “hot melt” process. In the first 'step, a pre-coat made of pure resin was applied to the carpet fibers in a -hot liquid form. In the second step, a main coat, also called a secondary backing and an adhesive, was applied. The' main coat consisted of three ingredients: resin, ethyl-vinyl acetate (“EVA”), and a filler. The resin provided a “sticky” quality to the main coat; the EVA provided strength; and the “filler” was just that. Like the, pre-coat, the main coat was applied hot. The two coats would cool to form a solid, fusing the entire carpet system together. The failure' of this system, in the form of delamination, edge ravel, and poor tuft bind, caused the warranty claims that led to this lawsuit. In contrast to Unibond, .Encycle and PVC were carpet-tile products. U2 was a broadloom product, but U2’s backing was not applied- through a two-step hot-melt process. In fact, none of the other products had this backing system or experienced the backing-failures at issue in this case. Also, unlike Unibond, «both U2 and Encycle were short-lived experimental ’ products. Perhaps most importantly, Arizona proffered no evidence specifically. linking the defects of Encycle, PVC, and U2 to quality control or any particular cause. To support the claim that the other products’ failures were relevant to a determination of how much profit Mohawk lost as a-result of the backing failures, Arizona provided-the deposition testimony of Robert Devlin, Mohawk’s director of commercial quality, expressing, his belief that the claims against Encycle in particular, which at one time accounted for 88.2% of Mohawk’s paid commercial carpet claims, hurt Mohawk’s reputation and affected sales of Mohawk, products. In addition, Arizona presented documents prepared by Mohawk attributing a general loss of commercial market share in 2009 and 2010 to the Encycle issues, at least in part. Arizona contended that this evidence would show that Mohawk’s lost profits were a result of non-Unibond product failures. In-support of its motion in limine, Mohawk asserted that the other products’ claims were irrelevant because Mohawk was not seeking damages for the loss of any market share other than in the broadloom carpet space. Thé trial court granted Mohawk’s motion in limine. Concerning damages, the ruling was without prejudice to Arizona’s raising the issue again if the evidence presented at trial warranted the admission of other-product evidence. — 99 — B. Causation Evidence At the beginning of trial, Mohawk informed the jury that the timing of the Unibond claims spike — and specifically'the fact that the claims rate returned to its historical level when Mohawk discontinued the use of Arizona’s resin — led the company to file the lawsuit. Devlin’s testimony provided the basis for this point. According to Devlin, the vast majority of the defective carpet was produced during .a time when Arizona’s resin was the only resin used in Unibond’s pre-coat and the only resin used in the main coat of at least seventy percent of the Unibond carpet being produced. Mohawk decided to discontinue the use of Arizona’s resin based on the possibility that it was the cause of the backing failures. Devlin testified that Mohawk filed suit a couple of years later after seeing the claims rate on Unibond carpet return to normal. Nevertheless, Devlin, a thirty-year veteran of the carpet industry, described a multi-faceted investigation into the cause of the backing failures. In addition to analyzing the claims rate and Mohawk’s manufacturing process, Devlin reviewed the claim files, personally inspected some of the claim sites, and examined samples of defective carpet. He considered strength and aging tests conducted on relevant carpet samples. These observations and tests led him to conclude that the condition of the defective - carpet indicated degradation of the resin used in the backing' system. Devlin ruled out installation errors and breakdowns in quality control.' Devlin also discovered that Arizona had made changes to its resin formula and determined that those changes coincided with the increase in backing failures. Further,, he concluded that no changes Mohawk had made. to. its main-coat formula would have caused the backing failures, and he reviewed Arizona’s own testing of its resin. Ultimately, Devlin opined as an expert that Arizona’s resin was the “sole cause of the backing failures.” Although not the only basis for Devlin’s conclusion, the claims-rate analysis was an important factor that he considered in reaching, this conclusion. The same conclusion concerning the cause of the backing failures was presented to the jury through the testimony of a chemist, Michael Cronin. He examined the results of scientific tests designed to assess how Arizona’s resin performed as it aged, inspected' samples of defective Uni-bond carpet, studied Mohawk’s Unibond manufacturing process, analyzed documentation of the development of Arizona’s resin, and considered the claims spike. Cronin opined that Arizona’s resin was highly prone to oxidation and degradation. He explained that oxidation caused - degradation of the resin, causing it to lose its adhesive quality. Mohawk buttressed this testimony with a document from a chemist employed by Arizona, which acknowledged that the resin Arizona had produced for Unibond penetrated the carpet fibers less, provided lower adhesion,. and degraded faster than, two other resins that had been used at various times in Unibond carpet. One such resin was the one Mohawk used just before contracting with Arizona. Mohawk had provided this resin to Arizona and received assurances that Arizona’s resin would perform just as well. Arizona offered a multitude of reasons the Unibond backing system may have failed, many of which related to poor quality control and lack of adherence to manufacturing protocol. One of Arizona’s experts on causation testified, for example, that Mohawk.had failed to systematically record the viscosity of the main-coat adhesive being applied to Unibond carpet before 2009. He found this lack of.record-keeping “appalling” and testified that overly thick adhesive , can cause backing fail — 100 — ures. This expert noted that once Mohawk began tracking the viscosity, there was tremendous' variance from batch to batch. He also observed that Mohawk had used recycled material as filler and had changed the ratios of ingredients in the main-coat formula, without providing direction to employees concerning the necessary ratios. Further, Arizona presented evidence that, of-the defective carpet samples Mohawk had produced for Arizona to test, thirty percent contained resin produced by another manufacturer. Finally, Arizona presented evidence that, its resin was of consistent quality, such that if the problem were with the resin, a, higher percentage of backing failures would have occurred. Arizona’s expert opined that it was not possible to say with certainty what caused the backing failures at issue. He suggested, however, that Arizona’s resin could be ruled out as the cause due to- the randomness of the claims, which he opined indicated a failure in the manufacturing process. Ultimately, the jury weighed the conflicting evidence concerning the cause of the backing failures in favor of Mohawk. C. Lost Profits Evidence In support of the damages Mohawk claimed for lost profits, Robert Devlin testified that Mohawk discontinued the Uni-bond carpet line because Unibond sales plummeted by seventy-five percent after the backing failures. Devlin testified that, by 2011, Unibond was no longer the flagship in the commercial-carpet industry given the toll on Unibond’s reputation caused by the backing-failure claims.- Devlin traced the drop in sales back to 2008, when he observed a “big dip.” He testified that sales continued to drop each year, resulting in a significant decline in revenue from 2008 through 2012. He noted specific customer accounts that had dropped off after the claims increased substantially. Mohawk sought recovery for lost profits beginning with the date Unibond sales began .to decline and for ten years forward. Daniel Edelman,. a forensic accountant, testified that Mohawk’s lost profits were $95 million. To calculate Mohawk’s lost profits, Edelman began by determining Mohawk’s average annual profits from the sale of Unibond before 2008. Edelman used that information along with market data for broadloom commercial carpet to project the revenue that Mohawk would have gained from Unibond from 2008 through 2017 but for the backing-failure claims. This procedure resulted in projected revenue below the pre-2008 average. Edel-man testified that he accounted for market trends and the effect of the general economic recession on customer demand for the very narrow category of broadloom commercial carpet. Edelman also took into account mitigating factors, including the costs that Mohawk avoided by not manufacturing Unibond, the increased profits, received on its other commercial broadloom products, and the profit Mohawk actually earned on Unibond during the relevant years. Edelman ended his calculation with the year 2017 because he projected that Mohawk would be able to convert its customers to an alternative product by 2018. When questioned on cross-examination concerning his consideration of economic conditions and consumer confidence in the floor-covering industry as a whole, Edelman testified that he took those matters into consideration but that his focus was on “the broadloom commercial product line, not the whole Mohawk company.” Arizona argued that it should be permitted to cross-examine Edelman based on Mohawk’s business records attributing a loss of commercial market share to factors other than Unibond. These documents — 101 — showed a steady decrease in Mohawk’s commercial market share from 2007 to 2010, followed by a slight rebound. The documents noted, that the market was shifting towards carpet tile. They attributed the decline between 2007 and 2010 primarily to the “Encycle Quality issues” and Mohawk’s “[s]maller footprint in Carpet Tile relative to.the industry,” while noting attractive pricing from competitors as another factor. Arizona’s counsel explained that his intent was simply to ask Edelman if he had considered “other causes for loss of market share to the plaintiffs ... other than Unibond” and then to offer the proffered evidence. The court declined to admit these business records or permit cross-examination concerning their content. ■ The court explained that Mohawk was not seeking damages for a general loss of market share, but for loss of Unibond sales in particular. At the close of Mohawk’s case, Arizona moved for a directed verdict on the lost-profits claim, asserting that Mohawk’s experts did not consider all factors that could have caused , a decrease in Uni-bond sales. Arizona pointed out that Mohawk had specifically excluded external variables from the calculation by choosing to focus on only on Unibond carpet. The motion was denied, and the jury awarded Mohawk $32.2 million for lost profits. II. ANALYSIS Having established this procedural and factual background, we now address the two most substantial issues in this appeal in turn. • First, we analyze the relevance of the other-product evidence to Mohawk’s theory of causation under the abuse-of-discretion standard. See McDuffie v. State, 970 So.2d 312, 326 (Fla.2007); Mitsubishi Motors Corp. v. Laliberte, 52 So.3d 31, 37 (Fla. 4th DCA 2010). Second, we consider de novo Arizona’s argument that the lost-profits evidence was legally insufficient, see New Jerusalem Church of God, Inc. v. Sneads Cmty. Church, Inc., 147 So.3d 25, 28 (Fla. 1st DCA 2013), but review the related question about the limitation on cross-examination for abuse of discretion, see Gosciminski v. State, 132 So.3d 678, 706 (Fla.2013). A. Evidence of Other Claims Spikes on Matters Related to Causation' Part of Arizona’s argument on appeal is that the trial court erred in applying the so-called “substantial similarity” test to its decision over whether to admit or exclude the evidence of other products’ claims spikes. See Ford Motor Co. v. Hall-Edwards, 971 So.2d 854, 858 (Fla.2007). We agree with Arizona that substantial similarity, or lack thereof, was not the disposi-tive question, and we recognize that the trial court considered the fact that the other products were dissimilar carpets with different manufacturing processes and distinct defects. However, that consideration alone does not indicate that the trial court improperly confined its analysis to whether the products were substantially similar. Cf. Laliberte, 52 So.3d at 38 (noting that, although plaintiffs relied heavily on similarity analysis, trial court’s “ultimate evidentiary rulings” were based on relevance and prejudice, among other grounds). The court also noted the general relevance test and took' into account Arizona’s argument that it should have been permitted to present evidence of alternative causation under this Court’s precedent in R.J. Reynolds Tobacco Co. v. Mack, 92 So.3d 244 (Fla. 1st DCA 2012). The court further found that, even if the evidence were relevant, its probative value would be substantially outweighed by the danger of unfair prejudice or misleading the jury. We find no reversible error in the trial court’s consideration of the other products’ similarity, or lack thereof, to Un-ibond. — 102 — The admissibility issue in this case, reflected in the trial court’s ultimate ruling, is whether the evidence was relevant and, if so, whether it was properly excluded for some other reason. Joyner v. State, 4 So.3d 76, 78 (Fla. 1st DCA 2009) (“In general; all relevant evidence is admissible, unless excluded by a specific rule.”). Under section 90.401, Florida Statutes (2013), “[rjelevant evidence is evidence tending to prove or disprove a material fact.” Even though relevant, evidence is inadmissible under, section 90.403, Florida Statutes (2013), “if its probative value is substantially outweighed by the danger of unfair prejudice, confusion of issues, misleading the jury, or needless presentation of.cumulative evidence.” Generally, evidence of “possible explanations” for the plaintiffs harm other than the defendant’s negligence is relevant and must be admitted. See Mack, 92 So.3d at 248-49; Haas v. Zaccaria, 659 So.2d 1130, 1133 (Fla. 4th DCA 1995); Aycock v. R.J. Reynolds Tobacco Co., 769 F.3d 1063, 1069-70 (11th Cir.2014). To establish .the relevance of particular alternative-causation evidence, however, the defendant must provide a competent .eviden-tiary link between the plaintiffs harm and the defendant’s theory. See Sidran v. E.I. Dupont De Nemours &. Co., Inc., 925 So.2d 1040, 1043, 1045-46 (Fla. 3d DCA 2006) (finding abuse of discretion in admission of an expert opinion that contaminated water, rather than the defendant’s fungicide, could have damaged the plaintiffs’ orchids, where there was no competent scientific evidence to support a finding that the level, of contaminants present in the plaintiffs’ water was harmful). This threshold requirement exists as a function of the relevance rule, even though the defendant does not carry a quantifiable burden of proof as to the alternative explanation. See Mack, 92 So.3d at 248 (citing Williams v. Eighth Judicial Dist. Court of Nevada, 127 Nev. 518, 262 P.3d 360, 369 (2011)). Here, Arizona’s basic argument is that the excluded evidence was relevant to rebut Mohawk’s point that thé timing of Unibond’s claims spike indicates that Arizona’s resin was the cause of the Unibond defects. The excluded evidence established .that other products manufactured in the same Mohawk facility without Arizona’s resin experienced claims spikes at approximately the same time as Unibond'. When these two points are considered outside the context of the remaining evidence, they make a compelling case for an abuse of the trial court’s discretion. Of course, context..is key. When the trial court’s decision is viewed in the context in which Mohawk used the evidence of Unibond’s claims spike and the limitations of the evidence concerning other products’ claims spikes, the basis for the manner in which the trial court exercised its discretion in this case is understandable. Mohawk did not argué that the claims rate alone established causation. Mohawk pointed to evidence of actual resin degradation, provided expert testimony that the resin formula caused the degradation, and analyzed the claims spike im the context of Unibond’s specific history with warranty claims and changes Arizona made to its resin. Mohawk did not argue simply that the coincidence of a claims spike with the use of Arizona’s resin -established causation; Mohawk argued that other factors, such as the history of Unibond claims and the lack of any significant change to the product other than the manufacturer of the resin during the relevant time period, combined with the claims-spike to suggest that the resin was to blame. Thus, the claims-spike analysis was specific to this product and was argued only in conjunction with other points. — 103 — Arizona argues that the evidence of other products’ claims spikes rebuts Mohawk’s causation theory because this evidence bears on the likelihood that other factors caused the Unibond claims spike. The specific factors Arizona suggests are poor quality control and lack of adherence to manufacturing protocol. However, there is no evidentiary basis in the record for supplying the connection between contemporaneous claims spikes of the four product lines and these factors. The record supports a theory that there may have been some general quality-control failings at the facility, but Arizona has not identified any evidence to substantiate its corn elusion that problems with' quality control explained the defects in the non-Unibond products. In fact, evidénce Arizona submitted in opposition to Mohawk’s motion indicates that Mohawk traced the causes of the defects in the other products to design flaws and choices of raw materials, not quality-control or procedural failures. Without a more direct connection between the other products’ failures and Uni-bond’s failures, the evidence of the other products’ failures showed causation, or rebutted Mohawk’s causation theory, only to the extent, that it showed. Mohawk has a propensity to produce bad carpet. Introducing evidence for this purpose is improper. § 90.404(2)(a), Fla. Stat. (2013). In light of these considerations, we find no abuse of discretion in the trial court’s decision to -exclude the other-product evidence as legally irrelevant to the issue of liability. B. Sufficiency of the Lost-Profits Evidence and Limitation on Cross-Examination Arizona next argues that Mohawk’s evidence of lost profits was insufficient ás a matter of law because, when calculating the amount of lost profits, Mohawk’s damages expert did not take into account relevant factors such as competition in the marketplace; an overall shift in the market away from broadloom carpet and toward carpet tile, and damage to Mohawk’s reputation from the failure-of Ericycle. In conjunction with this argument, Arizona claims that, at a minimum, the court erred in precluding it from cross-examining Mohawk’s expert with these business records. We find no reversible error. A plaintiff can recover lost profits as damages if “the defendant’s actions caused the. damage” and “there is some standard by which the amount of damages may be determined.” Cedar Hills Props. Corp. v. Eastern Fed. Corp., 575 So.2d 673, 677 (Fla. 1st DCA 1991) (citing W.W. Gay Mech. Contractor, Inc. v. Wharfside Two, Ltd., 545 Sb.2d 1348, 1351 (Fla.1989)). The plaintiff need not show that the defendant’s action was the sole cause of the damages sought; ’instead, the plaintiffs burden is to show that the defendant’s action was a “substantial factor” in causing the lost profits and establish the amount with reasonable certainty. Id. at 678; Whitby v. Infinity Radio, Inc., 951 So.2d 890, 898 (Fla. 4th DCA 2007) (quoting Forest’s Mens Shop v. Schmidt, 536 So.2d 334, 336 (Fla. 4th DCA 1988)). However, where the plaintiffs evidence reflects a mere assumption that the defendant’s action caused its lost profits without consideration of other factors shown by the record to be significant, the evidence is legally insufficient to support á claim for lost profits. See A.R. Holland, Inc. v. Wendco Corp., 884 So.2d 1006, 1008 n. 2 (Fla. 1st DCA 2004) (affirming judgment for defendant where plaintiff sought lost profits due to defendant’s sale of neighboring property to competing business in breach of parties’ contract but failed to present evidence taking into consideration any contributing factors such as road construction, national marketing problems; and internal-operating issues); Whitby, 951 So.2d at 898 (find — 104 — ing lost-profits evidence speculative where expert assumed radio station’s advertising revenue for the morning would increase- if revenue increased for the afternoon, without regard to numerous variables that must have had an effect “by their very nature”); Rooney v. Skeet’r Beat’r of Sw. Fla., Inc., 898 So.2d 968, 969 (Fla. 2d DCA 2005) (reversing award of lost .profits as speculative where it assumed lost profits were entirely attributable to defendant even though there was evidence that legitimate competition or a reduction in sales force may have contributed). Edelman’s testimony in this case shows, that he considered a variety of factors and rendered a reasonably .certain expert opinion on the amount of lost profits attributable to the loss of Unibond sales. Although Edelman did not expressly state that the market had shifted towards carpet tile and away from broadloom carpet, his testimony concerning customer demand and the general state of the broadloom carpet market necessarily encompassed this shift. We realize that Edelman’s trial testimony did not overtly address the damage to Mohawk’s reputation from the Encycle problems or the competition issue. However, we still cannot say the evidence of lost profits was legally insufficient to present the issue to the jury, because the record does not establish that the factors that Edelman did not discuss had a substantial effect on the sales of Unibond. •The evidence presented at trial indicated that Unibond was a product in a league of its own. Unibond was an established brand with a thirty-year history, while En-cycle was an experimental product.- Both were manufactured by Mohawk, but they were sold under different brand names and were much different products. No evidence proffered at trial established a link between the Encycle claims or any change in the competition Mohawk was facing and the decline in Unibond sales. Compared to the evidence of Unibond’s strength as a stand-alone product with a loyal customer base, Mohawk’s business record attributing loss of commercial market share in general to other factors was not specific enough to Unibond to render Edelman’s opinion unreliable as a matter of law. That record suggested only that the Encycle claims had an effect on Mohawk’s overall profits from commercial carpet, not that those, claims affected the sales of Unibond carpet in particular. The reference in that record to competition is similarly insubstantial. The record did not provide details concerning what product sales were suffering due to the effects of competition. In sum, the record does not show that consideration of these two factors by Edelman should have influenced his conclusion. Arizona’s alternative argument, that the trial court should have allowed cross-examination with the business records identifying these three factors, presents a closer question, the answer to which rests on the deférential standard of review applicable to the trial court’s control of the admission of evidence and cross-examination. Because the business records Arizona wanted to introduce and use to cross-examine Edelman did not purport to analyze the reason for a decline in Unibond sales in particular, reasonable minds could differ as to their relevance and probative value. If these general documents had a bearing on the lost-profits issue, that point was not made sufficiently clear through the proffer.. Consequently, we defer to the trial court’s judgment on this matter. Moreover, we do not find a sufficient basis to disturb the trial court’s ruling that these records were substantially more unfairly prejudicial or misleading than probative — 105 — due to their reference to the Encyele claims. AFFIRMED. OSTERHAUS and WINOKUR, JJ., concur. Footnotes [1] Arizona also argues that, even if this evidence were irrelevant, it should have been admitted under the opening-the-door doctrine after certain testimony by one of Mohawk’s primary witnesses. We find no abuse of discretion in the trial court’s ruling to the contrary. [2] Arizona also argued that a spike in claims for a line branded "Ecoflex” was relevant. However, Mohawk explains on appeal that Arizona was mistaken in its belief that Ecoflex was manufactured in Glasgow, and Arizona does not refute that point. Because Arizona’s theory of admissibility depends on manufacture of the other carpet-products at Glasgow, we affirm the ruling as to Ecoflex without further discussion.