We all know spring training stats are meaningless. So are horoscopes, but that doesn't mean I don't read them anyway.

Like horoscopes, we tend to read into them whatever we want to read into them. If the prediction is wrong, we ignore it. If it's right--wow, that's great!

In that spirit, I tend to seek out the stats which look like they're supposed to look. I know it's not logical, but they reassure me.

I'm currently finding a lot of comfort in these Cactus League numbers:

Player      ERA
---------- -----
R Harden    3.75
B Zito      4.00
D Haren     2.00
J Blanton   4.00
K Yabu      4.00
S Etherton  2.25

Hey, the starting pitching looks good! No worries.

Of course, I then go and rationalize away numbers like this:

Player      ERA
---------- -----
O Dotel     9.64
D Meyer    10.38

Those stats look bad, but that's because in this case, we're talking a small sample size.

Using a similar lack of logic, I have chosen to ignore all of the A's spring training batting stats, except Jason Kendall's. His are the only numbers that seem close to what I'd expect. So Kendall: wow, he's gonna be great, I'm sure.

Otherwise, the spring stats have been nonsense. Half the team's home runs have been hit by Eric Byrnes. The star hitter in camp has been Mark Ellis. Nearly everyone else is hitting around .200.

Using numbers like those to project the upcoming season is just crazy talk. It's as silly as astrologers predicting the future by staring up at the sky.

But hey, check out that big fluffy cloud over there. It kinda looks like Jason Kendall, doesn't it?