THE BOOK--Playing The Percentages In Baseball

Dan is missing a huge factor.  He never mentions in his article who those players would be replacing.  If in the rotation he’s be replacing Randy Johnson and in the bullpen he’s be replacing Dan Kolb, you’d obviously want him pitching out of the bullpen.  In this kind of article you have to take the situation into count too, because baseball is all about putting yourself into the optimal situation, into a situation which you could succeed in.

Tom,

Thanks!

You’re definitely right I should have mentioned leverage.  It’s baked into Rally’s numbers and I just didn’t call attention to it like I should have.

Aging is a definite issue too, but I was trying to keep the article somewhat high level.  You’re probably correct that Eck wouldn’t have kept up his starter performance in his 30s, but Smoltz did string together three seasons of > 4.5 WAR.

On the BABIP front, do you think either has demonstrated enough of a difference from average to make it meaningful?  I’ll be honest and say I didn’t really look into it very much.  As I said, I was trying to keep the fairly high level.

@Ari/2,

Yes and no.  The whole idea of replacement level is that the player is that much more valuable than what a team could pick up for free. 

That’s essentially it anyway, the actual calculations vary depending on the WAR methodology.

So the caliber of player he’s replacing in abstract is baked into the calculation.  There are theoretically different replacement levels based on the role for a pitcher, which I’m pretty sure Rally uses.

But you’re correct that it could make a difference in a particular situation. Although that’s more likely a consideration during the season than during the offseason.  Teams have a lot more freedom to make moves during the off-season because there are a lot more players available then.

In the particulars here, in 1987, the A’s gave 47 starts to pitchers with well below average ERA+.  I have to think Eck would have been better than them.  He was their best reliever, but they had some others who could have filled the role adequately I think.

Smoltz is a little bit of a different story because of his injury - so let’s consider 2002 instead of 2001. Atlanta had a great rotation that year.  The only relatively weak link (who was actually good for a 5th starter) was Jason Marquis.  So maybe there wasn’t really room for him in the rotation.