Three Way News

Your Source. For everything. Really.

Thursday, August 31, 2006

Stat geeks?

Baseball, more than any other sport, is reducible to numbers. It almost makes me want to be a statistician, but not quite.

When I read something like this, though, I wonder if the stats have gone too far:

Park factors can be evident in the performance of a single player, such as Carlos Beltran (.347-23-71 on the road and .225-15-37 at Shea Stadium), either because the park has been particularly tough in a given season or because the player is a poor match for his home park (witness Jason Bay's long-term struggles at PNC Park). This week, though, I will be looking at aggregate park factors at the team level.

So Carlos Beltran has struggled mightily at home and performed very well on the road. Interesting (to some) right? Maybe. Maybe not.

There are 30 teams in the major leagues. Each team has a 40-man roster, but a 25-man active roster. Each 25-man active roster has about 12 position players. That gives us roughly 360 active heaters in the majors for this season. Beltran has about 440 at bats. Roughly 220 at home, 220 on the road.

Now, if you flip a coin 220 times, you expect that you'll end up with about 110 heads and 110 tails. It's possible, but very unlikely (2^219:1 against, I think), that you'll get 220 heads or 220 tails.

But, if you do that 220 flip sequence 360 times in a row, you're very likely to get a couple of unlikely results. You would expect to get roughly 3 "1 in 100" results.

Maybe there's a perfectly sound baseball reason why Beltran is hitting better on the road than at home. But it might just be luck. With 360 position players, chances are someone is going to hit far better on the road than at home. Rather than puzzling over "Why him?" maybe we can just shrug and say, "Why not him?"