Joshua Cude <joshua.c...@gmail.com> wrote: So if the probability of a false positive is 1/3, and 1/3 of the tries are > hits, then that is consistent with all the hits being false positives. How > can you not get that? >
You are assuming that all hits are false positives. This makes no sense. There are 17,000 positives (false and real). If, as you say, only 1/3 of the tests work, that means there are 34,000 negative tests. The researchers are sure that it did not work 34,000 times, and they think it did work 17,000 times. Surely you do not claim that the researchers' are wrong about the 32,000 and they cannot even tell if a cell produced no heat?!? Their evaluations are not perfectly random. They are not flipping a coin. If the rate of false positives is 30%, then 5,100 of the positive tests were mistakes, and the other 11,900 are real. Or, if even ONE of those tests are real, then cold fusion is real. - Jed
