Even then the differences in statistics on rulings I believe can be explained by the relative sophistication of the plaintiffs who file there -- mostly patent trolls. Since trolls typically assert patents they acquire, they will go for "better" patents, and they can simply choose different patents to acquire when something like Alice comes along. Practicing entities have no such luxury and are stuck with the patents they were issued.
Lemley (the author I linked previously) and others actually have studies on these things. I'll have to dig them back up.
