Wednesday, 18 July 2012

2012, a difficult year

According to the results as published by the EPO, the pass rate for the C exam was a little lower than in the previous two years. Of 1410 candidates scoring more than 0 marks, only 33.12% passed the exam, and 11.13% scored a compensable fail.

No examiner's report has been published yet. At this moment, we can only guess what problems lead to this disappointing result.


  1. I guess you have an automatic system for the statistics analysis? If so, it would be interesting if you could post the statistics for papers A, B and D too. I have a feeling this years EQE has historically low results on all papers.

  2. Paper D: 35.87% pass, 11.78% compensable fail. Better than last year, worse than in 2010.

    Statistics for A and B are useless, because I cannot distinguish between chemical and electromechanical candidates.

    Full sit (four papers): 539 candidates, 167 passed (30.98%), 53 of them needed compensation. This result is not as good as in the previous two years (around 34.5%).

  3. OK, thank you very much!

  4. Disappointing indeed! My solution was rather close to yours, Joeri, yet I failed hard (37!).
    This was what I had:
    1. N A3
    2. IS A2+A6
    3. IS A4+A6
    4. IS A4+A6 (because 4 is broader than 3)
    5. N A3
    6. N A4
    7. IS A6+A5

    I'm really curious for the Examiner's report. Of course, my argumentation may have failed, but looking back at my paper, I'm pretty confident that was not the case. Anyway, we'll see... Congrats to the ones who passed!

  5. I passed with 62 marks.
    I did the 3-day DeltaPatents C course.

  6. Joeri, how do you extract the raw data from the .pdf?

    Got to wonder why the EPO made the document like that...

    1. There are some tools for that on the Internet. Some of them are free and can handle protected documents.

      I don't know why they do it this way. Maybe, they don't like people to calculate their own unofficial statistics. Maybe, it's just company policy always to protect outgoing pdf-documents.

  7. hello, i also passed and i attended the Delta patent C course in Eindhoven this year - thanx for the great job, Joeri! steph

    1. I attended Delta C in Munich with Joeri and passed, too!

  8. I passed with 64 marks. My expectation was actually a little higher - I had a very good overall feeling about the exam, and my attacks were almost identical to Joeri's.

    I guess this year they marked heavily on argumentation, and less on "use of information". I always found it difficult to get a feeling on what level and style of argumentation is desired by the C exam comittee. Maybe the Examiner's Report will shed some light...

    Still, I passed the EQE and that's what counts! Thanks to the excellent Delta training (goes for all modules).

  9. I failed C but thought I nailed it. I am interested in seeing what the Examiner's comments are like. I passed A, B and D.

  10. Carmo, expectations concerning C results are dangerous... Last year, I had all attacks but one correct, and scored a little over 30 points. This year, I missed at least four attacks and scored 41. By the way, this is the third time I score over 40 points in paper C, with a personal best of 47, but that was before the amended compensation rules... If I do not pass the exam next year, I will change carreer.

  11. I think I failed with 41 points because they subtracted points for not recognising that Art 53 (c) turned two product claims into process claims. I noticed the correct attacks and wrote them down completely, but I left the wrong attacks in the papers I submitted, plus the page where I wrote down that the 'for' in a claim 'product for' is not a limiting feature, also for these two claims

  12. 51% for paper C, far lower than for the other three.

    I thought this was a very odd paper - simple in some ways, but perhaps some very subtle issues that (I guess) many people including me missed.

  13. I have critics to exam. Report C2012. They wrote, Claim 3 was to handle for 2 embodiments (A and B), due to information [0009] in A1. But, then they have to start with that on claim 1 already! and didnt. Further, english and german best solutions published do not have this on any of claims, also not on claim 3. Crazy examiners? Please explain, if You see another reasons for 2 embodiments just in claim 3 and why not for all other claims.

    Next: for claim 1 and claim 5 they want to use also A5 to destroy novelty? Claiming that fibre layer is same as textil. This is a big work of fantasy; even if they want to switch of our own knowledge about textil, there is no such easy definition of textil in any of documents.

    Claim 6: A4 ist clearly destroying novelty by information in [0005], what they mean to be not combined as solution. So, if that embodiment is using aluminum, and there is in any place of that dicument written that further metalls like steel, cupper may be also used instead of aluminium, it is not combined? Maybe the examiners are beginners in first year? Look above what they did with textil and fibre layer for novelty destroing over A5 on claim 1. This crazy examiners seem to spot over all of examinees who are working successfully since years in jobs and training hard for passing the exam.

    Also on exam. report A 2012 I have critics to offer: the main claim of best solutions and in report is just describing what they often argue in real proceedings to be just description of inventions target, and not a description of technical means able to lead to that target. So they have just functions there, without any means. English solution does have a little hint in this direction but german is a pure target description. In such a way it is always possible very aesy and short to formulate a claim, would be very nice easynnes. They score it both english and german with 50 points, full points. OK, if so, we all have to use this case for future passing exam A and for the real life proceedings too. Since this year 2012 we are all absolutely alowed to formulate claims like that, just describing the target. All, who have tryed to formulate the claim by MEANS and functions, landed by much longer and complex formulations, which in my case the examiners has given just 10 points, even if it is a correct claim. Further: does anyone see a published best solution, where dependened claims and description does score more then 30 points? How is it possible? No one can score full points for that? It seems they score it by "political" decision, not for the performance writen there.

    1. We can have an endless discussion on whether claim 3 should be attacked once or twice; we can argue in favour of both approaches. I think it is more useful to focus on future exams. It seems highly unlikely that we will get this construction back in near future exams; so it is better to focus on things you may get. It is our guess that having only one out of the two attacks on claim 3 will have cost no more than 5 marks. It will also have saved time for another attack, so not many candidates will effectively have lost a lot.

      With respect the textile:
      A1 [05] discloses (shortened): "The textile
      layer can be formed from any nonwoven or woven material which both can be made
      from various fibres".
      A5 discloses "a nonwoven fibre layer (54)" and even for same function (reinforcing, acting as a support).
      Difficult to see why this would not be novelty destroying. Candidates had no problem with this at all in the exam.

      I fully disagree that A4 is clearly novelty destroying; it clearly is NOT. Under novelty a feature from one embodiment may only combined with another embodiment if the combination is clearly indicated. That is not the case here. Admittedly, the skilled person may consider combining both, but that needs to be assessed under inventive step. For novelty this concept is already long time established, e.g. in T305/87. For added subject-matter a similar concept applies, e.g. in T1206/07 "It was not sufficient for the combined characteristics to have actually been disclosed individually in the initial application in order to make their combination comply with the requirements of Article 123(2) EPC. This conclusion is also in line with the EPO's established jurisprudence on this question, which says that the content of the application as filed cannot be regarded as a repository from which individual characteristics can be combined for the artificial creation of a particular combination."