Considering that
1. the plaintiff is starting with a rebuttal presumption of patent validity,
2. it is the plaintiff who makes the decision whether or not to sue, which should weed out a lot of weak cases, and
3. this was before software patents, back when examiners actually had a deep understanding of the art in the fields whose patents they examined, so there was not a flood of bad patents being issues,
I'd expect in a fair system for plaintiffs to win the majority of cases.
According to the article, a significant fraction of patent lawsuits were being deliberately filed by competitors to challenge the validity of patents that had just been issued. The reason to do this was so that you could choose a venue that was patent unfriendly and therefore gave you better odds of overturning the patent.
Consider the fact that these lawsuits were filed preemptively (they had to be filed before the patent was used to sue you, else you would not get to choose venue), and that they would be filed on the weakest possible patents. One would expect the patent holder to lose a very large portion of these.
I would suspect that a lot of the 80% of wins right after CAFC was due to these cases reaching the CAFC, and it asserting its power over lower courts.
The long-term slide since then is undoubtably due to people pushing the envelope as fast and hard as they can on how bad patents can be, and how broadly patent law can be filed in novel areas (ie ones that Congress never meant it to be applied in).
There were obvious issues in the pre-CAFC era, but I'd prefer those issues over the current status quo.
They were winning 20-40% of appeals, it doesn't say how often they won overall.
It's hard to say whether 20-40% of appeals is a good number in an absolute sense (percentage of successful appeals is a more useful number) but in this case, it's only to compare with the 50-80% figure that came after the change, and I'd think that is pretty significant.
Given this advantage, and the fact that defendants can always avoid court by licensing if a patent really is valid, I would expect defendants to win the majority of cases that actually make it to court (calling the plaintiff's bluff).
2. Sorry, all I can do is laugh. Do you actually follow the US legal system?
3. Again, whether a patent is bad or not has nothing to do with whether it's infringed in any particular case.
2. Yes. I acquired a passing familiarity with the US legal system in law school.
3. Again, read more carefully.
Do you follow it deeper than the propaganda about frivolous lawsuits?
> 3. Again, whether a patent is bad or not has nothing to do with whether it's infringed in any particular case.
He's not talking about particular cases, he's talking about aggregate statistics. A system with a lot of weak patents, where people might independently come up with the same design because it's obvious, is going to have a lot more infringement than a system with only strong patents, where infringement is likely to come from only purposefully copying a design.
The most depressing, even alarming, part of this is:
> Instead, patent appeals are exclusively heard in DC by judges who live and breathe patent law. Unsurprisingly, this leads to insular thinking. For example, when we interviewed Paul Michel, who served as the Federal Circuit's chief judge from 2004 to 2010, he didn't seem to understand the problems facing small software companies. "If software is less dependent on patents, fine then. Let software use patents less as they choose," he said, seemingly oblivious to the fact that software companies don't have the option to opt out of patent troll lawsuits.
Seriously? The court seems to have no idea of the Pandora's Box they've opened by loosening the "obviousness" constraints.
There's still plenty of drama. For example, a few years back, there was quite a fuss over a simple change to the filing date of something on the docket in East Texas:
http://thepriorart.typepad.com/the_prior_art/2009/09/troll-t...
That inventor behind all that mess was quite an interesting fellow, too:
http://thepriorart.typepad.com/the_prior_art/patent_troll_tr...
They don't necessarily have insight into the dynamics of how patents function in the software space, but that doesn't mean their actions are just a cynical money-grab.
Interestingly, similar charges have been made of the DC Circuit with regard to the Supreme Court's national security jurisprudence, especially Boumediene v. Bush.[1] (The US Court of Appeals for the DC Circuit has essentially become the US terrorism court — because among other things — they are the only court allowed to hear Guantanamo detainee appeals. But DC Circuit judges do hear normal cases as well.)
So it seems to me the problem with CAFC is not just in its "specialist" judges but also in allowing the law in one field to develop almost exclusively in one court. This also makes Supreme Court supervision much more difficult.
1.http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1838402
