@arstechnica Remember that the incentives incentivize this kind of behavior. If you want to get tenure you must publish at least seven papers in seven months, which incentivizes you to put quantity and massaging the ears of journal editors and peer reviewers instead of publishing quality papers that inform and inspire. Peer review needs to be replaced with peer replication, and if you want to publish you should have to release your source code and everything needed for a layperson to replicate your experiment with a step by step tutorial if necessary. Then your peers use your instructions to do your experiment exactly how you did it to see if it works, in exchange they get listed in the bibliography and get a thanks for helping prove the experiment worked.