In recent years, the Patent Trial and Appeal Board has frequently declined to institute IPRs for procedural reasons unrelated to a petition’s substantive strength. In particular, the Board has increasingly denied petitions in view of related, parallel litigation that it perceives as so far advanced that it would be most efficient to deny institution and leave patentability issues to be resolved in the other forum. Apple Inc. v. Fintiv, Inc., IPR2020-00019, Paper 11 (PTAB Mar. 20, 2020) (Precedential). Key among the factors guiding those Fintiv denials is whether and to what extent the other proceeding’s trial date is scheduled to precede the Board’s deadline for issuing a final written decision, i.e., Fintiv factor two. Id. at 9.
But how reliable are those trial dates?
The Board “generally take[s] trial courts’ trial schedules at face value absent some strong evidence to the contrary.” Apple Inc. v. Fintiv, Inc., IPR2020-00019, Paper 15, at 12-13 (PTAB May 13, 2020) (Informative). Some have questioned that approach, citing limited data sets that suggested such trial dates often change and therefore present an unreliable basis for denying institution. We took a more comprehensive look at this question by identifying all discretionary denials that were based on parallel litigation and issued between May and October 2020. That six-month period opened the same month that Fintiv was designated precedential and ended approximately one year ago, allowing us to evaluate what actually happened over the intervening year when an IPR otherwise would have taken place and reached a final written decision within the 12-month timeframe required by statute.
The Board was almost always wrong when predicting trial dates in parallel litigation
Our results confirm the prior criticism. Out of 55 discretionary denials, only seven cited a trial date that proved accurate. Notably, in four of those, the cited date was correct because trial had already occurred when the Board denied institution. Apple Inc. v. Unwired Planet Int’l Ltd., IPR2020-00642, Paper 15 (PTAB Sept. 9, 2020); Apple Inc. v. Optis Wireless Tech., LLC, IPR2020-00466, Paper 13 (PTAB Sept. 15, 2020); Apple Inc. v. Optis Cellular Tech., LLC, IPR2020-00465, Paper 13 (PTAB Sept. 17, 2020); Amazon.com, Inc. v. Vocalife LLC, IPR2020-00864, Paper 22 (PTAB Oct. 28, 2020). When evaluating future trial dates, the Board was wrong 94% of the time (48/51).
The discrepancies were often substantial. Out of the 51 cases where the Board relied on a predicted future trial date, only three occurred on time. For the others, one was delayed by less than one month, five were delayed by 1-3 months, 17 were delayed by 3-6 months, three were delayed by 6-12 months, and seven remain pending pre-trial, well beyond the earlier trial date the Board accepted at face value. Another 15 litigations were terminated without any ruling on validity (for reasons including settlement, bankruptcy, and summary judgment on other issues).
The Board’s reliance on scheduled trial dates has proven remarkably inaccurate, and our results contradict the Board’s stated practice under Fintiv of simply accepting nominal trial dates at face value under Fintiv factor two. Trial dates in patent litigation are not stable and make a very poor barometer for evaluating the potential efficiency of denying institution based on a parallel proceeding.
 Our methodology counted AIA trials individually, including when multiple petitions were related to the same parallel litigation. The percentages remained approximately the same if related AIA trials were grouped by litigation, with errors in predicting future trial dates occurring in 95% of related proceedings.