We're better in the rearview mirror than we are at predicting - 'cause you're never going to be right every time. You can handicap it. You can point to certain elements that make it work, and many of those elements come straight out of epidemiology, right?