This would be a good point if patent cases were randomly assigned to districts. But they're not, of course - EDT courts are stuffed full of BS patent troll cases. So of course the patent suit win rate is lower there. It's still famously friendly to patent plaintiffs.
http://www.law.com/jsp/article.jsp?id=1103549728998&slreturn...
In ED Texas, the clerks all have hard-science undergrad (and sometimes graduate) degrees as well as degrees from top law schools because this is a major leagues of IP litigation. Want a guaranteed six figure income as a 26 year old? Clerk for two years in Marshall, then send your resume to Cooley or Wilson Sonsini or Loeb & Loeb and see how fast they return your call.
This is miles and miles beyond most lawyers, who struggle with basic math and still use dictaphones and WordPerfect and aol.com email addresses.
This also means that the litigation goes faster, which is always something that clients complain about. Win or lose, you want the dispute resolved.
A well-informed jury about the topic at hand is bad for one party of a lawsuit. It's better that a jury is a blank slate so that the lawyers can attempt to convince them their story is the one, true story in the debate. It's why during jury selection they often actively attempt to find a juror that knows nothing about the topic or carries no opinion about it.
Of course, that's not necessarily true across the board. Sometimes one side does want people that know the topic and have an opinion about it but I bet those are mostly criminal cases.
Plus, to me anyway, if the juries did actually have this supposed high-level of knowledge and experience then the results would be more balanced. Because there's no way you can tell me that nearly every single patent troll out there is in the right with these cases. More of these patents should be tossed out for the silliness that they represent.
Lawyers do their best to get into courts and before judges they feel will be sympathetic to their case. To do otherwise is a disservice to their client.