I knew that someone was going to pick up on the scent of Bayesian fallacy here ;-) ...though I think there's probably an economics-style-argument that can be made here that a market where companies reject a lot of candidates won't be in equilibrium if there are a lot of candidates who seldom get rejected. -- Whether the exact number is 9, 99, or 999, and what kind of measure of central tendency that number represents, whether it's a mean, a median, a mode, or whatever, has no real bearing on the argument I was making.