Saturday, 11 May 2013

Looking Back On 2012 (Part 1)

I guess most of the decisions taken by the Boards of appeal in 2012 have been published by now. Let us have a look at what the output was.

I have collected and analysed 1410 decisions issued between January 2 and December 20, 2012: my survey encompasses 755 decisions on opposition appeals, 629 on examination appeals, 16 on petitions for review, 9 decisions of the Legal Board and one “G” decision of the Enlarged Board.

In this post, I will only consider the 629 examination appeals. I intend to present the opposition cases later on.

Unsurprisingly, English was clearly the dominant language in examination appeals: it was the language of the proceedings in more than 78% of the cases. Far behind, German took an 18% share. French decisions are almost a quantité négligeable.

In principle, there are three possible outcomes to examination appeals. The Boards can order the grant of a patent, dismiss the appeal, or remit the case for further prosecution. There are a few other cases, e.g. when a Board only decides on reimbursement of the appeal fee or on re-establishment, but these cases made up less than 0.5 per cent of the examination appeals in 2012.

As you can see from the following graph, in a majority of cases (about 57%), the appeal was dismissed. An order to grant was issued in about 23% of the cases and about one case in five was remitted for further prosecution.

The outcome was not the same for the three languages – but before you get excited, please note that this may be an artefact due to the low numbers: there were only 20 French and 116 German cases, as compared to 493 English cases, so the statistical error is considerable.

(click to enlarge)

This time I undertook to directly compare the different Boards of appeal. The number of decisions taken into account for each Board is given in brackets. As there are 27 Technical Boards, the average number of examination appeals issued by a Board in 2012 was 23, but the differences between the Boards are considerable: I have found only 4 examination appeals for Board 3.3.06, whereas Board 3.5.06 appears to have issued 54 examination appeal decisions in 2012. Such low numbers make comparisons doubtful, but still one can see that the overall outcome of the appeals differed greatly between the Boards. Some frequently ordered grants whereas others mostly (or even always) dismissed examination appeals. Also, the part of remittals varied greatly.

(click to enlarge)
I also found it interesting to compare the outcome as a function of the technical domain:

(click to enlarge)

NB: I have tried to be very careful, but as such a survey involves big amounts of data, I cannot guarantee that the above figures are free of error. If you spot a mistake, please let me know.


Anonymous said...

I think one should make a difference between cases where the claims were amended such that grant or further persecution became possible. Very often the boards get completely different claims to decide on than the first instance division, especially in examination. Therefore, a different outcome on appeal does not mean that the first instance decision was wrong. Mostly it was correct, but because of a change of claims the outcome on appeal is positive.

george said...

I would think that “correct” and “wrong” are not the most helpful categories in this context, because they require some absolute standard. Even if the claims are the same and the Board sets aside the first instance decision, does this mean that the latter was necessarily “wrong”? The Board has the last word, but its majority opinion is not necessarily “correct”, is it? That being said, as the Board members are experienced people, blatant errors (at least of law) should be less frequent than in the first instance departments.

Anonymous said...

A board having a technically qualified chairman consists usually of four technically qualified members and three legally qualified members. If the chairman is always the same person there are 12 different compositions for a specific case. Divide the number of decisions you have found (ie 54 decisions for board 3506) and you get about 4.5 decisions for each particular board´s composition. Really a large number for making statistics :-))

Anonymous said...

Well, George, that is how most people interpret it, when a first instance decision is set aside on appeal. And one other reason for hopefully less blatant mistakes: although the work pressure is very high in the boards, the pressure to "produce" is much, much higher in the first instance. That said, if a first instance decision is confirmed it is not necessarily "correct". Sometimes one arrives at the right conclusion for the wrong reasons :-)

Anonymous said...

Statistics is not the correct word to use, I think. Numbers is more appropriate. Trying to apply statistics to small numbers is a mistake that is often made.

the dude said...

congratulation!you did a great job and everyone should say at least "thank you" for processing all this information with an interesting and somewhat hepful outcome.(instead of questionning the significance of this "statistic"/i am ashamed by the previous comments!)
the dude

oliver said...

thanks, the dude

yep, sometimes comments can be disheartening

as someone who has done quasi-elastic neutron scattering, I know more about bad statistics than most of my contemporaries ;-)


Anonymous said...

Yes, all the work done is certainly worth a "thank you". Nevertheless, a warning regarding the word statistics is also warranted, especially as the division of outcome by Board might suggest a trend or something. People are often very quick at finding such "trends", which aren't any. 630 decisions divided by 27 Boards (again divided by 3 possible outcomes).... not much room for statistics, I'd say....