Is KSR Actually Helping Applicants at the BPAI?

Earlier in the month, the 271 Blog reported on post-KSR appeals at the BPAI related to obviousness (link), and reported that "only" 36% of obviousness cases were reversed on appeal. This inevitably lead to the question: "what was the reversal rate before KSR?"

This resulted in further digging, and a bit of controversy. The BPAI traditionally publishes statistics in the Official Gazette (OG), and similarly posts statistics on the BPAI link of the Office's web page.

Back in 2005, a widely-distributed analysis was made on reversal rates, based off of the OG, and reported that the reversal rate was around 60%:

Disposition-------% of Dispositions--------[% decided cases]
Affirmed------------------- 36. 6%------------------[40 % decided]
Affirmed-in-Part---------12.6 %-------------------[14 % decided]
Reversed-------------------40.9 %------------------[45 % decided]
Reversals-------------------53.5 %------------------[59 % decided]

Partial Remands, Dismissals--------9.9 %

TOTAL, 100.0%

Now, the 59% "reversal" number was a combination of decisions that outright reversed a rejection (40.9%), and partial reversals (i.e., "affirmed-in-part" - 12.6%). It is important to note here, that the reversals included ALL rejections, and not just obviousness rejections.

Turning to the published numbers on the BPAI site (link), the numbers tell a slightly different story:

FY 2005 (link) -- reversed = 39.6%; partial reversals = 12.5%; total reversals = 52.1%

FY 2006 (link) -- reversed = 34.8%; partial reversals = 12.1%; total reversals = 46.9%

FY 2007, through April (link) -- reversed = 29.6%; partial reversals = 13.9%; total reversals = 43.5%

Judging from the "reversed" numbers alone, the 36% rate of reversal on obviousness rejections for May 2007 suggests an increased rate of reversal at the BPAI. Again, the early numbers (2005-2007) combined all reversals, which may or may not have been based on obviousness. Thus, it can be extrapolated that the reversal rate of 36% on "obviousness-only" rejections suggests an increase in the overall reversal rate. It follows that the likelihood of reversal on 103 grounds increased as well.

The BPAI is expected to publish its May numbers shortly, and they will be posted when they come out.