SPRING TRAINING STATS: ANOMALIES OR HARBINGERS?
This story is on ESPN.com. The first line of the story is:
Spring training stats: anomalies or harbingers? Sorry, we'd probably need Bill James to answer that.

Actually I'll answer it. Spring stats are meaningless because it's too small of a sample. How can you gather a solid opinion based on 60 or 70 at-bats? Do you know what the difference between a terrible spring and an awesome spring is when a player only has 65 at-bats?
15 for 65 = .230
18 for 65 = .276
21 for 65 = .321
24 for 65 = .369
27 for 65 = .417

So basically if a player is swinging the bat well and gets a few bloop hits while only having 65 ABs he could go from a decent spring of .276 to an amazing spring of something like .417.

Pitching is the best example because one bad outing (especially for a reliever) can ruin a pitchers' ERA. When you throw 10 innings in the spring, imagine what a 4-run outing in 2/3 IP does to your ERA.

If a player has a horrible spring, for instance Scott Podsednik (who hit about .153), then proceeds to have a bad season people will remember the poor spring he had. It simply doesn't matter though, because the sample is too small to formulate an accurate opinion. Plus a lot of players in spring training are experimenting with new mechanics and new approaches or just simply trying to get work in. They aren't approaching every at-bat or every inning pitched like they would during the regular season.

I know that in this article the question of "how important are spring training statistics" wasn't really the focus of the article but it reminded me of a rant I wanted to go on.