Thanks, but I'm making a lot of assumptions. It may be that the panels are not formed at random. They may, for example, have to have at least one DC judge. It's easy enough to check my conjecture if you can match the rejections with the panels that rejected them.