THE BOOK--Playing The Percentages In Baseball

I’m a Phillies guy but it’s enough to make me root for them darn Pirates.  Go Bucs!

Again, I am unsure of the pitcher batting 8th thing, and of course it is going to entirely depend on the composition of the other batters in the lineup.

I have also always wondered if the extra PA that a pitcher gets batting 8th can be mitigated by simply taking him out earlier for a pinch hitter?  Or maybe that even happens naturally.  Does it? When we compute the number of PA each spot in the batting order gets per game and per season, we assume that no one comes out of the game for a pinch hitter. If you are always going to pinch hit for the pitcher, say the third time he comes up to bat, then where he bats in the lineup won’t make a difference, right?  If you are only going to leave him in if he is pitching really well and his pitch count is not high, then he might get more PA from the 8th position than the 9th, but not as many as if you never took him out for a pinch hitter after X number of PA.

If the pitcher does NOT get extra (or even THAT many extra) PA batting 8th, then it is a clear advantage to bat him 8th in most cases, right?

How many extra batters does a pitcher face batting ninth rather than batting eighth? There would be a few times each year when the eighth place hitter would make out and end the inning, whereas were the pitcher batting there, his manager would pinch hit for him. If he bats ninth, he’s still in the game for at least the beginning of the next half-inning.

Charlie, since we believe that teams would be better off replacing their starting pitchers earlier (at least if they are not an ace), it would actually be a good thing to get them out of there earlier.

Here’s the OPS each time through the order for a SP in 2009:
http://www.baseball-reference.com/leagues/split.cgi?t=p&lg=MLB&year=2009#times

1: .727
2: .766
3: .807

And a relief pitcher’s 1st time through the order is .725.  Even a sucky reliever would be better than a starter’s 3rd time through the lineup.

Well, count me in a different camp, then.

Tango #4:

No gripe with your general point, but wouldn’t you want to tailor your choice to the actual situation, i.e. your actual starter and actual bullpen?  Those figures are for the average pitcher, but that animal never actually starts a game, it is always one particular player, for which you have his specific stats.  And for that matter, he’d be backed up not by an average reliever, but by a particular set of individuals, each with his own stats.

So, perhaps the Pirates ought to factor into their decision 8th vs. 9th the identity of the starter and the condition of their bullpen (i.e. who is available, and who is likely to be first out the gate in the 5th, 6th or 7th inning.) It wouldn’t matter most of the time, of course, but on average 1 time out of 9 they could sneak another inning out of a strong starter without sacrificing hitting production of their own in a possibly high-leverage 2-out situation.

Greg, sure, that just sets the baseline.

A sucky reliever has an OPS of .775.  So, a sucky reliever his first time through the order is the same as an average starter the 2nd time through the order.

That’s why, when you are now into the third time through the order for an average starter, he should be pulled, and pulled fast.  Whatever you have in the bullpen is better than your 8MM$ starter.

Now, if you are talking about Lincecum, then yeah, bat him 9th maybe.  You don’t want to pull him early.  But, we’re talking about just 10 or 20% that are in this boat.  The other 80% of the starters should be pulled earlier rather than later.

Why does the pitcher’s 4th+ time through the lineup drop the opposing OPS down to .776? Is it because managers tend to only leave good pitchers in for that long?

Melvin: that’s part of the reason.  But Tango also found an increase in pitcher effectiveness, on the 4th PA even controlling for pitcher talent.  I think the reason is that there is some tendency for a pitcher who is pitching well on a given day to keep pitching well.  It’s not a strong tendency, and over the first few innings it’s undectable.  But if you’re being left in for the 4th time through, you’ve been pitching well for at least 6 IP, and at that point there’s some truth to the notion the pitcher is “on” that day.