How lucky can one guy get?

I thought it was the case that all batters tend to regress to their own historic BABIPs as the season goeas along.

For example, we wouldnt expect Ichiro’s BABIP to regress from .366 to the league average (.300), because it is in line with his career BABIP of .357 (over 5600 career ABs).

Is that right?

Maybe Ichiro’s just been really, really lucky for a really, really long time. Or maybe the statement that “Equally well known is that players’ skills have little impact on their BABIP” is simply not true. I think I’ll go with the latter.

Mike: A batter’s BABIP tends to regress toward his “true” BABIP (which is probably higher or lower than .300). A batter’s recent historic BABIP sheds light on his true BABIP, but because of noise, we can’t pin down the rate perfectly. Certainly, for a batter with Ichiro’s career AB, we have a better idea of his “true” BABIP than we do for less prolific batters.

Like you, I would not expect Ichiro’s BABIP to regress to .300, and no sound projection system would, either.

Mike- Well, you can always obtain a “best guess”; what more AB gives is less dilution. With Marcels, for example, you add 1200 PA of league-average stats to the player’s weighted 3-year history. Even for someone with Ichiro’s massive playing time (2,200 raw PA from 2006-08; 8,800 weighted PA), league-average stats represent about 12% of his expected numbers. For someone with 200 PA/year, the proportion is 33%.

If you wanted no more than 20% dilution, then you’d need 400 PA/year (more or less). Etc.