Tuesday, March 15, 2005

Analysis of CAFC decisions: invalidity down, noninfringement up

from an article by Glynn S. Lunney, Jr., Patent Law, the Federal Circuit, and the Supreme Court: A Quiet Revolution, 11 S. Ct. Econ. Rev. 1 (2004).

page 15:

As Figure 2 reveals, patent invalidity is significantly
less likely to be the reason why a claim of patent infringement fails under the Federal Circuit.
Before the Federal Circuit, invalidity accounted for fully three-quarters of the cases in which claims of patent infringement failed. In contrast, after the Federal Circuit's creation, invalidity accounted for just more than one-third of the failure results.

As Figure 3 reflects, an inability to establish infringement is
significantly more likely to be the reason why a claim of patent infringement fails under the Federal Circuit. In the pre-Federal Circuit era, a failure to establish infringement accounted for roughly one-quarter of the cases in which claims of patent infringement failed. However, under the Federal Circuit, an inability to establish infringement accounted for more than
two-thirds of the failure results.

This observation (also related to an earlier post of an observation by Judge Rader) has relevance to the patent quality debate involving Quillen/Webster on the one hand and Clarke on the other hand. If patent quality were going down significantly, one might expect the fraction of invalidity decisions to increase. The analysis by Quillen and Webster is discussed in 4 CHI.-KENT J. INTELL. PROP. 186 (2005).

Also, of "expectations" for patent grant, note the following text from the patent pending blog.
An inventor's statistical chance of success.

Sometimes an inventor asks about their chance of success. If they are using an invention promotion company, I would say their chance of success would be better if they bought lottery tickets. However, we have noticed some statistics that are useful in this evaluation.
Our usual practice is to do a patent search, then file a patent application.

Over a period of time, we have observed some general statistics, which of course do not take into account if the product is a good idea, if the invention is at the right time, if it can be built for a price to make its sale price attractive, if the inventor is competent to get it to market, design it, hire the right people
at the right time, etc. Just looking at numbers, here are the numbers:

100. Of every 100 inventions we do patent searches on, about 60 percent of them turn out to have patentable subject matter.

50. Of those 60 searches, about 50 of them go ahead with a patent

45. Of those 50 applications, about 45 of them result in issued patents.

25. Of those 45 issued patents, about 25 inventors do nothing with the invention, or are not successful. They may run out of energy, inspiration, time, dedication, or they may make dumb mistakes about marketing, packaging, designing, unable to find a licensee, or run out of money.

20. Out of the 45 issued patents, about 20 inventors do some kind of test marketing.

5. Of the 20 that do test marketing, about 5 will fail. The item may not sell for the price they have to get in order to make a profit, or people may not be interested in the product.
15. Of the 20 that do test marketing, about 15 are at least marginally successful.
10. Of the 15 that experience marginal success, about 10 will quit for lack of interest, resources, dedication, inspiration, changes in circumstance. Some of these could have been saved if the inventors could learn from purchasers what they really want in this kind of product, and if the inventor can change his product. Most inventors don't realize how long it takes to make a product a
success, and don't have the stamina to follow it through.

3 and 2. Of the 5 that are left of the 20 who did test marketing, about 3 will grow into successful single product businesses, and 2 will license their product to a larger company.
[end text from patent pending blog]

These numbers are of relevance to the Quillen/Webster discussion. Of the 50 applications for which a search finding patentable subject matter is performed, about 45 lead to issued patents, a success factor of 90%, less than the 97% of QWI, but greater than the 85% of QWII. However, that is the success rate with a prior search. If no search were performed (none is required), then the success rate would be 45%.

The fact that the raw PTO number for issuance is about 2 patents issued from every three applied does suggest that (some) applicants (and/or attorneys) do their homework, and do not expend the money for filing applications based on disclosures not studied for patentability.


Post a Comment

<< Home