I thought this was a good basic demonstration of Bayes, but I have to nitpick:
-The assumption that each x is picked at random is not really true, as it would assume the Browns second string RB who will get 4 carries is just as likely to be picked as Adrian Peterson.
-This appears to be a couple years old, threw me for a loop (Bryce Brown didn't have any regular season games against the Giants last year).
Yeah I wrote it when I initially set the site up (this is the third year).
In terms of the x's, the point was just to get across the idea you're drawing from a distribution. No matter what the distribution (whether it's just starting RB's or just Adrian Peterson or what), you could still sum it up in terms of stacks of little x's. So each x would be equally likely if you were drawing from the universe of ALL running backs. If you're drawing from the universe of starters, it'd look a little different.
