Looking at Abstract Idea Appeals by Tech Center

Anticipat Beta Research Database recently released the tech center search filter and the tech center column in the table display. Appeals data can now be examined for trends using tech center groupings. In this post, we will showcase one example of how this might be useful using the “abstract idea” ground of rejection.

In the past 7 months, 123 decisions have been decided on the “abstract idea” ground of rejection. Here is the breakdown for the “abstract idea” statutory subject matter rejection across all tech centers.


1600 1700 2100 2400 2600 2800 3600 3700 3900
6 1 7 9 7 2 75 13 1

Tech center 3600 is software and business method-heavy. In the immediate aftermath of the Supreme Court deciding Alice, many suspected that these types of patents/patent applications would be on the chopping block. And many examiners have sought to enforce such an approach. Small wonder, then, that this tech center leads in abstract idea appeals.

To put context to these numbers, we looked at the intake number of appeals per tech center. This intake of FY17 is only a rough comparison of the decided appeals in FY17 since it takes many months for an appeal taken in to be decided.


Comparing intake of appeals and abstract idea decisions, appeals of tech center 3600 are very much overrepresented compared to the decided abstract idea appeals. With the exception of the two lowest tech centers, the appeal intake of tech center 3600 is anywhere from 2-4 times the intake of other tech centers. But tech center 3600 decided appeals are a much different story–anywhere from 5-36 times the decided appeals.

An obvious reason for this overrepresentation is that the abstract idea doctrine does not affect all tech centers equally. Tech centers 1600 and 1700, for example, are focused primarily in the biotech and chemical arts, and other judicial exceptions to statutory subject matter are more applicable (e.g., law of nature, naturally-occurring phenomenon). Other tech centers are focused on physical inventions where the abstract idea ground of rejection is theoretically less applicable. For example, tech center 2800 includes inventions in semiconductors, memory, circuits, optics and printing. And as can be seen, the ratio of abstract idea appeals in tech center 3600 to tech center 2800 is far greater than the ratio of the corresponding intake.

Another reason for the overrepresentation stems from the large number of abstract idea rejections introduced as “new.” For the relevant time period, there were 16 new rejections in this tech center alone. This number is more than any of the other tech centers’ total number of abstract idea decisions of any disposition. This shows that the abstract idea doctrine is top of mind for judges deciding these types of cases, even if the Examiner did not apply an abstract idea rejection for the appeal.

A final reason for the overrepresentation may stem from the high volatility of the abstract idea doctrine. We generally assume based on anecdotal evidence that filing for an appeal is a (comparatively) arduous alternative to working with the Examiner through argument, interviews, and making amendments to overcome rejections. Thus, appealing a case requires a level of dedication that is born of coming to an impasse with the Examiner. Often times such an impasse can take place when the law is in flux—the Examiner can hold to one position and the applicant can hold to a contrasting position. And with software-focused inventions, such as those in tech center 3600, often times the impasse is the Examiner asserting that there is nothing patent-eligible in the application. Such a position leaves little room for any other procedure but appeal.

It would be discouraging if the judges in tech center 3600 supported this abstract idea chopping block attitude reflected by some. But the outcomes data show that they do not. We previously reported that the rate that abstract idea rejections are reversed are in line with other Section 101 nonstatutory rejections: 25% wholly reversed and 27% partially reversed. As we drilled into the abstract idea rates on a tech center level, we found that the same reverse rates hold true for tech center 3600. That is, 19/75 decisions (25%) in tech center 3600 were wholly reversed and 20/75 decisions (27%) were at least partially reversed. This shows that in the face of significant abstract idea rejections in tech center 3600, judges reverse at the same rate as other tech centers.

The above discussion shows that, by using tech center-focused appeals data, a quantitative assessment of how strong an Examiner’s position for a particular ground of rejection can be deduced by comparing that ground of rejection by technology center.  While we have done this for “abstract idea” grounds in this post, this same analysis can be done for the other grounds of rejection reviewed by the PTAB.  With this information, the odds of success on appeal can be better calculated based on which technology center is examining a particular case.


Section 101 – nonstatutory subject matter decisions: different categories = different reversal rates

One of the sexiest topics in all of patent law has become §101, specifically, patent-eligible subject matter. Part of the recent appeal stems from high volatility and uncertainty in the law. But not all categories of patent-eligibility grounds are in such flux. Some §101 nonstatutory grounds of rejection (e.g., reciting a propagated signal) are relatively predictable and stable while the so-called judicial exceptions are more unpredictable. So we drilled deeper into the types of §101 rejection to get a more complete picture of reversal rates. We found a big difference in the observed reversal rates of particular categories.

The following chart shows a breakdown of the past seven months of decisions on grounds of §101 – nonstatutory subject matter. Data for this chart was pulled from the past seven months using Anticipat’s research database. Anticipat keeps track of issue-specific tags to allow for better identification of sub-issues within issues. So while the Examiner and PTAB may decide a particular issue on §101 – nonstatutory subject matter grounds, Anticipat goes a step further to delineate the specific type of §101 – nonstatutory subject matter ground.


  1. Statutory Classes

Section 101 nonstatutory rejections include the statutory class variety (e.g., does the claimed invention fall within the statutory classes? recite more than one class? claim a human?). This is otherwise known as step 1 of the Mayo/Alice framework. Of the 186 substantively decided §101 decisions since July 25, 2016, these step 1 types accounted for 38 or 25%. Twelve were wholly reversed, a reverse rate of 32%.

This higher reversal rate for classes makes intuitive sense. Administrative patent judges must understand that statutorily, patent-eligibility is broad. A process, machine, manufacture, and composition of matter originally allowed almost any human innovation at that time of the enactment of the 1952 Patent Act to be patent-eligible. As technology has since changed, not all inventions fit into this framework, such as propagated signals and software per se. But for the most part, the courts have fit inventions into these categories—from non-transitory computer readable media to engineered bacteria. The observed reversal rate indicates that judges may reverse the Examiners in an attempt to be more faithful to the statute and to case law than the Examiners are.

  1. Judicial Exceptions

The judicial exceptions to patent-eligibility, such as abstract ideas, law of natures, and natural phenomena, have surged in popularity in recent years. And the appealed decisions show it. Of the 186 decisions within the past seven months, 146 have been judicial exceptions. The most popular of the exceptions is the abstract idea.

Of the 119 abstract idea cases, 30 were wholly reversed and two were reversed in part, or a complete reversal rate of 25% and an at least partial reversal rate of 27%. This falls squarely within the overall §101 rates that we previously reported. Natural phenomena/product of nature types are slightly higher at 31% while the law of nature ground reversal rate is markedly lower at 7%.

  1. Analysis

The number of decisions for some of these categories should be more reliable as the number of decisions increase, but some take-home lessons are clear. A judicial exception rejection has a lower chance of getting reversed than the statutory rejections. The PTAB judges are likely averse to overruling an Examiner’s finding of a judicial exception, especially when there is a great deal of uncertainty in the courts.

Furthermore, law of nature rejections are very infrequently reversed. Part of this may be a lack of positive case law to specifically support law of nature rejections being erroneous. By contrast, several Federal Circuit decisions have been decided within the past year that are positive for the patentee/applicant in showing that the claims are not an abstract idea.  Because of this, the judges have more material to work with in finding a particular claimed invention passes the Alice framework.