Gavin Andresen writes: > Peer review works best when everybody involved is given time for conversation and debate without being contacted by reporters on deadli

Gavin Andresen writes: > Peer review works best when everybody involved is given time for conversation and debate without being contacted by reporters on deadline. In other words, the entire institution of science (computer science in this case) is better off if the individual researchers are careful about presenting things to the press and make sure to review things carefully before publishing. But the incentive for individual researchers may not be aligned with this. An individual researcher might be able to advance their career by going to the press (particularly about a hot-in-the-news-right-now topic like Bitcoin) and by making sensational claims ("Bitcoin is 'Broken'!") even if the overall scientific enterprise suffers. After all, had you ever heard of Ittay Eyal or Emin Sirer before this? Given these skewed incentives, at least a few people will "cheat" the system by going to the press, and those people will profit by it. Rather like Eyal and Sirer's claim for Bitcoin miners. - - - - - - - - - - To discuss the merits of the paper itself (rather than Gavin's response), if a mining pool were to utilize the strategy described in the paper it would result in that pool repeatedly releasing new chains that are longer than the existing blockchain but which fork one OR MORE of the previously released chains. While this is allowed under the Bitcoin protocol (this is the way that forks in the chain get "healed"), it is also rather obviously visible. So if certain miners were following this protocol then it would quickly become obvious that they were doing so. The community would then be able to take actions to redress the problem.