THE BOOK--Playing The Percentages In Baseball

I would have guessed that the penalty for short rest in the post-season has been worse than the normal penalty that we (Tango) found in The Book.

The reason I say that is that pitchers pitching on 3 days rest during the reg season are probably doing so because they are more able to, according to their managers, and because they had low pitch count or low stress prior outings.

In the post-season, pitchers tend to pitch on 3 days rest because their managers are desperate (they are down in the series) or because the manager has some lousy #4 and #5 starters and he’d rather pitch a tired #1 or #2 starter. 

Anyway, just knowing their ERA and w/l record in the post-season does not tell us much.  We’d have to compare that to their overall ERA and adjust for the batting opponents. 4.37 does seem high.  Maybe half a run high or so.  And you are clearly giving up some IP (90 pitches and 5.4 IP), which is probably not too bad as managers tend to use their best relievers in the post-season and you don’t usually run into over-use problems with all the days off.

Since 1995, the MLB ERA in the post-season is 3.89.  Among pitchers with at least 75% of the games as starts, it’s 3.97.  (Among those with at least 90% of their games in relieft, it’s 3.72.)

So, a 4.37 ERA (likely among the better pitchers… who else would pitch on 3 days rest), is likely way above what was expected.

The opposing batters are almost certainly representative of the population of post-season hitters. 

So, we’re talking about a 0.50 to 0.75 jump in ERA, most likely.  Plus a drop of close to one inning.

In short, they likely give up the exact same number of earned runs (2.6), while pitching one less inning (5.4 instead of 6.4).

Tango, how many runs in ERA did you find in The Book among all pitchers on 3 days rest?  I assume that was regular season only.  I forgot, did you control for pitches thrown in pervious start?  I think you had a minimum number of pitches thrown in previous start, did you not?

Did you track pitches or innings thrown after 3 days rest for the reg season pitchers?

I would also be curious if the amount of pitches thrown in prior start (even after some minimum) has an effect on the next start after 3 days rest.  IOW, if a guy throws 80 pitches is he better on 3 days rest than if a guy threw 115 (like Sabathia against the Yankees).  As I said, if a pitcher in the post season on 3 days rest is a 4.37 ERA pitcher, then limiting his IP/pitches is probably a good thing.  In fact, the proper strategy is probably to throw one of your better starters on 3 days rest but limit him to 4 IP or so (maybe 2 times through the BO) and then bring in a series of good relievers.  I also wonder how pitchers who threw on 3 days rest did on their NEXT start (after the one after 3 days rest).

If we’re talking about a .50 to .75 jump in ERA, shouldn’t the decision point be:

“Is my #1 starter at least .75 runs better than my #4?

If the answer is “Yes,” it seems that I ought to be bringing back my #1 on 3-days rest, no?

And isn’t that run differential very often the case, particularly when we’re talking about a true ace (Beckett/Sabathia/Webb) vs. an average-to-weak 4th (Wakefield/Byrd/Owings), or even with a strong #1 (Francis) vs. a Franklin Morales?

In other words, the results are suggestive of a “problem” with starting a #1 on short rest, but the conclusion—that it would be better to start your #4 on full rest—doesn’t necessarily follow.  In fact, I’d think it rarely follows.

If I have an awful bullpen I guess that loss of 1 inning may alter the equation, but given that this is the playoffs, with each series including 3 days off, I doubt that would have been a major consideration for any club this year other than the Phillies.

The team record is .419 on short-rest pitchers.  That would be a bit below what a #4 pitcher would give you against an average pitcher.  I didn’t look to see who the opposing pitcher was, but presumably it would have been a #3 or #4 pitcher, so we’d have expected to be closer to .500.

Good point, nonetheless.

Danno, you are right - that is the question. 

There is another consideration which I brought up in my last post.  Does that 3-day rest start affect the next start?  As I also said in my last post, I don’t think the short IP on a 3-day start is much of an issue, as post-season bullpens are usually quite good, and it would take an awful bullpen to be worse than a #1 starter pitching .5 to .75 worse than usual.

I think one of the issues here is that all the pitchers we can study are ones in which the manager thought that using them on 3 days rest would be O.K., or at least not disastrous. IOW, we have significant selective sampling problems.  It would probably be (or at least might be) wrong to generalize and say that Sabathia should be starting rather than Byrd, because even at .75 runs worse than his normal runs allowed, he is better than Byrd (which he is).  Maybe Wedge does not allow him to pitch on 3 days rest because he thinks (and he is right) that he will be 1.5 runs worse.  Certainly there are some pitchers better on 3 days rest than others. The question is can the manager and coaches identify that dichotomy.  I don’t know. Maybe yes and maybe no.  And as I also said before, the decision probably depends on how stressful the prior start was and how many pitches were thrown.  For exampe, Beckett only threw 80 pitches last game (supposedly because of back problems) but if his back is fine, he may be a prime candidate for pitching again on 3 days rest (even though I don’t think that he is nearly more than .75 runs better than Wakefield).  He also had that great game for FLO versus the Cubs on 3 days rest and you know that his manager is going to consider that.

Guys, I’m pretty sure that B-R list isn’t really what we’re looking for.  That includes anybody that had pitched at all in the previous four days, rather than anyone who had started in the past four days.  About 1/3 of those guys had short relief appearances, and that’s why they’re showing up there.  I ran the numbers for the guys who were on what we traditionally think of as short rest over at my site, and the jump in ERA is 1.09.