How often do you get 5 minutes into an office action and you start thinking "here we go again, another idiot patent examiner that has no clue and is just trying to pad his/her stats!" I bet most of us patent prosecutors have that thought way more than we would like to admit (except, of course, on a rare first action allowance -- then the examiner "get's it" :-)). The fact of the matter is that most examiners are not idiots, and most are not trying to game the system. So how can we separate the wheat from the chaff? What metrics could we use to objectively assess a patent examiner? Not only to justify or dispel our initial knee-jerk reactions, but also so that we can tailor our responses to their tendencies and get our clients better patents in a more-efficient manner. In this post and my next I am going to look at some examiner metrics I find useful. I would love to hear others' suggestions in the comments (good ideas may be incorporated into the next revision of the bigpatentdata examiner metrics tool).

Allowance Rate

Allowance rate cuts right to the chase: does this examiner allow applications or not. If an examiner's allowance rate is high relative to her peers, it may be worth an extra response or an interview to get that allowance. If an examiner has a very low allowance rate, perhaps proceeding to appeal sooner will be the better move (BUT, be sure to check out her appeal statistics first). Although there are a number of ways allowance rate can be measured, on bigpatentdata.com it is the number of allowed applications divided by the total number of allowed and abandoned applications. Abandonments for an RCE are not counted as abandonments. Considering all examiners who have examined at least 100 applications in the past 3 years, the average allowance rate is 75%, and the median allowance rate is 79%. The highest allowance rate seen in the wild is 99% shared by 10 examiners. Of those, the one with the most allowances is Sophia Chen with 313 allowances and 3 abandonments in art unit 2852:

The lowest allowance rate among the group is Examiner Richard Scheuneman with 2 allowances and 122 abandonments (1% allowance rate) in art unit 3624:

Actions per (Final) Disposal

This is the number of actions (here actions include: restrictions, non-final rejections, final rejections, Examiner's Answers, and Ex-Parte Quayle actions) divided by the number of applications in which the actions were taken. Actions per disposal can tell you a lot about whether an examiner is efficient at getting to an allowance or an abandonment, or if she generates lots of actions that don't meaningfully advance prosecution. On bigpatentdata.com this is further broken down into actions per allowed application and actions per abandoned application to help assess whether allowances are quick and abandonments drawn out, or visa versa. The examiner with the fewest actions per disposal (0.1) is Hieu Vo:

The examiner with the most actions per disposal (6.5) is John Winter:

Relation between Allowance Rate and Actions per Allowance/Abandonment

As suggested by the previous two figures, there is a strong negative correlation between allowance rate and actions per disposal. This is shown explicitly in the figure below. Allowance rate is strongly negatively correlated with actions per allowance (correlation coefficient = -0.75) and moderately negatively correlated with actions per abandonment (correlation coefficient = -0.58). The correlation is stronger for examiners with higher allowance rates and weaker for examiners with lower allowance rates. Whether or not your examiner is one of the outliers can inform how to respond to that examiner. For example, if your examiner has a high allowance rate and a relatively high number of actions per allowance, they may just want to pad their numbers with an RCE before allowing the application. Conversely, if your examiner has a low allowance rate and relatively low number of actions per allowance, your examiner may make up her mind about your application early in prosecution and an appeal may be the better option than wasting time and money on an RCE. Of course, all of this should be considered in conjunction with the Examiner's appeal statistics, which is discussed in Prosecution Poker with Examiner and Art Unit Appeal Statistics. Also, when relying on allowance rate, make sure it is a relevant allowance rate, as discussed in Lies, Damned Lies, and Examiner Statistics

IF you find this article enjoyable, I would be super excited if you subscribe and follow BigPatentData on LinkedIn and Twitter. Thanks!