To summarize poorly, he admits that there might be some selection bias in which patents are defended but generally finds that Patent Trolls are just about as effective (or ineffective) in court cases as the general patent-owner population. You suggest (and I agree!) that the courts and current patent systems are ill-equipped to handle software patents effectively, but his systematic look at the evidence, compared to the excellent anecdotal look TAL provided, suggests that for the largest, most systematic trolls, the patents asserted aren't much worse than non-troll patents.
I highly suggest the full-paper, or at least the final conclusions section.
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1792442
To summarize poorly, he admits that there might be some selection bias in which patents are defended but generally finds that Patent Trolls are just about as effective (or ineffective) in court cases as the general patent-owner population. You suggest (and I agree!) that the courts and current patent systems are ill-equipped to handle software patents effectively, but his systematic look at the evidence, compared to the excellent anecdotal look TAL provided, suggests that for the largest, most systematic trolls, the patents asserted aren't much worse than non-troll patents.
I highly suggest the full-paper, or at least the final conclusions section.