A. J. Mass devotes a column to the nonsense of home and away splits:

As we always say here at “Playing With Numbers,” stats that identify trends in past performance don’t guarantee the same trends in future performance. But at some point, the proof clearly is there in your proverbial pudding, and it might be time to earn your just desserts.

No, no, no, no, no. There is no proof. There is just a probability distribution. The randomness that underlies these numbers means that the average player’s home/away split has no more likelihood of continuing forward from this moment than does its opposite.

Take a look at Vladimir Guerrero, one of the players on Mass’ poor home list. Does Guerrero have a bad history in Angels Stadium? No. In 2007 his home vs. away OPS was .954 vs. .946. From 2005 through 2007 (now our sample size is getting reasonably significant) the difference was .969 at home and .925 away.

Given Guerrero’s historical stats, the most likely thing to happen in the coming weeks and months is a return to the mean. Do you really want him on your bench when that happens? Look at your keyboard. That’s a board with which to hit yourself when Mass’ advice loses you a matchup.