> Chun explains: “There are 3,003 scenarios where 10 people can be chosen from the 15 applicants, and only 11 scenarios where all 10 are men. All things being equal, the chance that the top 10 applicants would all be male is 11 out of 3,003.” Or, as Chun calculates it, the likelihood that gender discrimination occurred in the tenure process last year is over 99.6 percent.<p>Everything else aside, this is a very poor argument. "All things being equal, the chance that John Smith won the lottery is 1 out of 10,000,000. Or, as I calculate it, the likelihood that the lottery was rigged in his favor is over 99.999%."<p>There are plenty of arguments to make to support the idea that gender discrimination occurred here, but this is not one.<p>This kind of reasoning has had disastrous consequences in criminal trials. People are inclined to think "given that this person is innocent, what is the likelihood of this outcome happening" when they should be thinking "given this outcome happening, what is the likelihood of the person being guilty", i.e. you need to consider actual evidence.