Everyone in the patent world now knows of inter partes reviews (IPRs), and knows IPR proceedings are killing patents. Click here for current statistics. Currently, about 65 percent of IPR proceedings result in all challenged claims being canceled. Id.

So with a 65 percent “kill rate,” you’d expect that nearly all petitions are being granted, right? Otherwise, how could enough petitions survive settlements, disclaimers, withdrawals, and losses to arrive at 65 percent? So writing a petition that gets a proceeding instituted and leads to cancellation can’t be all that hard or require all that much attention to detail, right?

Unfortunately, for many, that sense of easy petitioning can be false to an extreme. In fact, certain Patent Trial and Appeal Board (PTAB) judges write petition decisions that deny a full half of the petitions under their review. Moreover, by example, they deny some petitions and grounds for some interesting reasons. Here’s a short case study.

A particular PTAB judge, as author of institution decisions, has granted institutions of 22 petitions and denied 21. (The judge will remain nameless; we may appear before this judge.) The grants, however, include partial denials. More than half of the asserted grounds of challenge of all the 43 petitions were denied by this judge. The denial of half the asserted ground by itself is interesting, in that it raises questions. Is this cherry-picking going on? Are the judges right-sizing their workloads per case? Answers may show up in the reasons for denials.

Here are a few of the reasons for denial. We note that many of the decisions of this judge include criticisms of petitions and patent owner preliminary responses over petty issues. These include typographical issues, such as having more than one subsection A in a section, citing to a “VI (A)” instead of “VII (A),” and the like. The judge also plainly recognized these were harmless typographical errors. Yet, it is possible that these criticisms reflect a curiously high standard of judging. 

On substance, in a first case, issues of claim construction and application to the prior art concerned the word “attempts.” The patent claims required action based on a number of attempts. The prior art taught action based on successful attempts—a subset of all possible attempts. The judge interpreted the term “attempts” under broadest reasonable interpretation and based on the patent specification to include all attempts, both successful and unsuccessful attempts. As a result, the judge wrote, the claims did not read on the prior art, because the prior art had action based on attempts, but based on successful attempts, not all attempts. For this reason (and others), the petition was denied.

In a second case, the petitioner and patent owner agreed on claim construction. The judge disagreed, and broadened an interpretation of a term. The parties had good reasons for their agreedto construction of the term, since the term at issue was stated in the claims to have three functions and these were part of the parties’ construction. The functions were: receive, combine, and output. The judge broadened the construction of the term to eliminate the functions from the term’s definition (the functions remained expressly in the claims). The patent owner argued, however, that to meet the claim term with a prior art item, the petitioner had to show that the prior art item met the three functions, and another function not a part of any interpretation. The judge agreed.

A third case concerned communications, limited in type in the dependent claims. The prior art disclosed communications, and only of the particular type. The patent owner argued, however, that the petition did not disclose how the prior art limited the communications to the type. The judge agreed, in spite of the prior art limiting the communications to the type by the essence of what it was and did. He denied the petition as to the dependent claims.

All three of these case decisions have oddities. In the first case decision, “attempts” was said to be broadened in being limited to all attempts. It would have been broader, it would seem, if it meant any attempts. But it was not interpreted in that way.

In the second case decision, the oddity of claim construction appears to have been based on analysis of the phrasing of the functions by the patent owner and judge. The claims stated the functions were to receive, combine, and output, i.e., to receive two modified inputs, combine the received inputs, and have an output. Not expressing a claim interpretation of the functions, the judge nevertheless interpreted the functions to mean that to meet the claim, any structure of prior art referenced by the petitioner, as matching the item of the claims, had to be proved to receive modified inputs but also generate the modified inputs, and then combine them.

A device seemingly cannot both receive inputs and generate them (almost, of course), but rather than not take the tack taken, or find it incorrect, the judge required the prior art to do both, receive and generate the same thing. Another interpretation would have been that the word “modified” was not expressly stated in the function of combining the “received” inputs and that the function of combining the received inputs was necessarily a function of combining received “modified” inputs. Then the functions would have been simply receiving modified inputs, combining the inputs (the “received modified” inputs), and having an output—sensible. But because the challenger did not argue this, or do the impossible, prove both receipt and generation of inputs in the prior art item, a ground of challenge was denied. 

In the third case, the oddity was a requirement that the prior art show how the prior art limited communications to type, as well as have the limiting of them, where the claims at issue only required limited communications.

There are obviously many institution decisions by this judge, as this writing is kept short and reflects only three decisions. In the many decisions, there are denials of institutions for quite reasonable reasons, such as failure to explain reasons for an asserted combination of prior art, lateness of petitions, lack of proof of publication of asserted prior art publications, and the like. But, among the many are these three and some similar decisions, and interesting reasons for their denials.

An overarching lesson here is what was stated at the beginning, that with a 65 percent “kill rate,” you’d expect that nearly all petitions are being granted. Many petitions lead to settlements, disclaimers, withdrawals, and ultimate losses. But just because the claim cancellation rate is 65 percent, don’t think writing a petition that gets a proceeding instituted and leads to cancellation can’t be all that hard or require all that much attention to detail. A sense of easy petitioning is false. Arguments available to patent owners may seemingly make no sense, but some of them are being adopted. Some claim interpretations that appear to make no sense are being volunteered by IPR judges. Some cases have a curiously high standard of judging.

Because there are judges at the PTAB who write petition decisions that deny a full half of the petitions that these judges consider, and because you never know who will be your judge, petitions need to be written for these few. Petitions must be written with a sense of exploring and preparing for the odd. The alternative is risking denial of IPR petitions.