The key point is the first one. Which is the court recognized that there was evidence for the $145M award. Which is worth $200M USD including prejudgment interest. And over $400M USD for units including after the trial. I don't know how much the Intel iPhones are worth. But that is in addition. You are talking potentially $5 Canadian a share in judgment from just this one trial. For a stock that is trading at only just about 1.50 share excluding cash.
The judge ruled against himself. He let evidence in and then after the jury award, he decided that he made a mistake and said that evidence shouldn't have been admitted. Wow!
ARGUMENT
I. THE ORIGINAL DAMAGES VERDICT SHOULD BE REINSTATED
The district court recognized the evidence amply supported the jury’s $145 million award. Appx31 n.1. It vacated that award only because it thought three pieces of evidence—Wi-LAN’s rate sheets, Samsung license, and direct-valuation evidence—were improperly admitted. But that evidence was admissible. Nothing so “‘substantially prejudiced’” Apple as to justify discarding the jury verdict after a full trial. Harper v. City of Los Angeles, 533 F.3d 1010, 1030 (9th Cir. 2008). Overturning a verdict based on legal or clear factual error constitutes an abuse of discretion. Medsim Ltd. v. BestMed, LLC, 758 F.3d 1352, 1355 (Fed. Cir. 2014).
A. The rate sheets were plainly admissible. As Skippen explained, Wi-LAN prepared those rate sheets based on market analysis, and used them in actual negotiations. Appx10462-10464. This Court has repeatedly upheld use of licensing practices—including rate sheets—when determining royalties. CSIRO, 809 F.3d at 1303 (upholding reliance on “Rate Card”); VirnetX, 767 F.3d at 1330.
The district court suggested the rate sheets were “outrageous offers” like those in Whitserve, LLC v. Computer Packages, Inc., 694 F.3d 10, 30 (Fed. Cir. 2012). Appx42. Whitserve, however, acknowledged that “proposed licenses”— like rate sheets—“may have some value for determining a reasonable royalty.” 694 F.3d at 29-30. It rejected a proposed 31.8% royalty rate—offered to justify a
76
Case: 20-2011 Document: 23 Page: 94 Filed: 02/12/2021
CONFIDENTIAL MATERIAL FILED UNDER SEAL REDACTED
19% royalty award—not merely because the rate was “unaccepted,” but because it was “based on fiction and contradict[ed] [other expert] testimony.” Id. at 29-30. But here, Apple’s expert testified that Wi-LAN licensees had actually paid up to $1.25/unit, Appx11161—well over the rate sheets’ $0.65-$1.10 range. Rate sheets proposing less than Wi-LAN previously negotiated could not “artificially inflate the royalty rate” or “‘ske[w] the damages horizon.’” Appx42. They were plainly admissible and caused no conceivable prejudice. It was an abuse of discretion to rule otherwise.
B. The Samsung license was properly admitted. It bore on a salient difference between Wi-LAN’s prior negotiations and its hypothetical negotiation with Apple. The licenses Wi-LAN negotiated with Doro, Unnecto, and Vertu would not affect existing license rates. But Apple’s would: Samsung’s “adjust- ment right” would reduce payments from Samsung if Wi-LAN granted Apple— Samsung’s closest competitor—a cheaper license. Appx10665. Given that, Kennedy explained, it would not “make economic sense” to grant Apple a license below a certain amount, because Wi-LAN would lose more from Samsung than it would gain from Apple. Appx10657. That evidence about Wi-LAN’s “economic circumstances” was plainly relevant. Finjan, 626 F.3d at 1211. Nor would the license inflate damages. While substantial, $ Royalty was less than Kennedy’s
77
Case: 20-2011 Document: 23 Page: 95 Filed: 02/12/2021
$145 million proposed royalty for Apple (at $0.85/unit), which fell squarely within Wi-LAN’s rate sheets.
The district court deemed the Samsung license “‘not relevant’” because Kennedy supposedly had not testified it was “comparable” to Apple’s hypothetical license. Appx42. But Kennedy testified that Samsung was “certainly the most comparable.” Appx10640. He accounted for pertinent differences, such as the Samsung license’s greater scope and Apple’s greater sales—differences going to “weight,” not “admissibility.” Ericsson, 773 F.3d at 1227. Regardless, Samsung’s license was relevant because its adjustment right was an “economic circumstance” bearing on the hypothetical negotiation. Declaring it “ ‘not relevant,’ ” Appx42, was legal error.
C. Wi-LAN’s direct-valuation evidence was also properly admitted. Damages experts may consider the value to the infringer “of use of the patented technology . . . compared to using non-infringing alternatives.” Aqua Shield v. Inter Pool Cover Team, 774 F.3d 766, 770-71 (Fed. Cir. 2014). Here, the ’145 patent’s bandwidth-allocation invention—which VoLTE implements—prevents “degraded voice quality” by creating an “express lane” for voice packets. Appx10094; see Appx324. To isolate that technology’s benefits, Madisetti relied on tests comparing VoLTE to the best non-infringing alternative, Skype—which also sends voice packets over LTE, but without their own “express lane.” See pp.
78
Case: 20-2011 Document: 23 Page: 96 Filed: 02/12/2021
72-73, supra. Specifically, Madisetti compared call quality with “a lot of background applications” running and on a “loaded network.” Appx10265. The VoLTE call-quality advantage he described thus reflects the invention’s incre- mental benefit in conditions where its “express lane” matters most. Prince and Kennedy then estimated the value of that specific benefit. See p. 73, supra.
The district court held that evidence inadmissible based on a clearly erron- eous “premise”: that Madisetti’s opinions were “not confined” to “that aspect of VoLTE associated with improved voice quality ‘during loading.’” Appx41. The record refutes that characterization. Madisetti specifically compared VoLTE and Skype call quality during loading, with “a lot of background applications that are running on the phone” and on “a loaded network.” Appx10065. The “2.3 unit increase to the MOS score” on which Madisetti based his opinion, Appx36, explicitly compared the scores for VoLTE (3.7) and Skype (1.4) under those conditions, Appx10265-10266. Indeed, Skype scored better (3.1) under other conditions. Appx10265. It could not be clearer that Madisetti was not measuring the benefit of VoLTE “generally,” Appx41, but the specific benefits of Wi-LAN’s technology during multitasking.
The district court’s record citations are not to the contrary. While Madisetti testified (correctly) that VoLTE is “‘enabled’” by the ’145 patent, Appx35-36; Appx41, he did not measure VoLTE’s benefits as a whole. He compared VoLTE
79
Case: 20-2011 Document: 23 Page: 97 Filed: 02/12/2021
and Skype based on call quality during multitasking, the task the patented technology improves. Nor was the link between VoLTE’s better call quality and the ’145 patent “without factual basis.” Appx37. Madisetti explained how VoLTE uses the patented technology to “allocate bandwidth across two connections” and avoid the quality-destroying delay that Skype’s “single data connection” experi- ences during multitasking. Appx10261. Apple did not identify any other VoLTE feature that would improve voice quality under such conditions.
The district court’s comment that “there would have been no need” for
Kennedy to “further apportion” the described benefits if Madisetti’s testimony
were directed to the patents, Appx37, reveals a misconception. As Kennedy
explained, licensees do not pay patentees the entire value added by a patented
invention. They retain a portion, based on their respective “negotiating leverage.”
Appx10638. Kennedy’s opinion that Apple would pay Wi-LAN just 1% of the
invention’s marginal benefit, Appx10628, reflected that reality. There being no
proper basis for excluding Madisetti’s testimony, it was error to exclude Prince’s
and Kennedy’s reliance on him. Appx37. The verdict should be reinstated