"This seems true enough - from a (rational) game-theoretic perspective, why should any victim really fight hard to overcome a troll, if in doing so they risk bi

"This seems true enough - from a (rational) game-theoretic perspective, why should any victim really fight hard to overcome a troll, if in doing so they risk big losses, and don't gain any advantage over their competition even if they win?" This might be true over a single iteration, but I don't think it holds well over indefinite iterations. Developing a (credible) reputation as an entity (government, corporation, person, etc.) who doesn't "negotiate with terrorists" can be quite a valuable asset. Just like productive companies have an incentive to give in to the demands of patent trolls, patent trolls have an incentive to avoid companies who "don't negotiate with terrorists", in part because there is a very high (> 50%) rate of claim invalidation across the industry (what does that say about the quality of the original examination?), and in part because it costs the patent trolls just as much money to go to court as it does the productive companies.