THE BOOK--Playing The Percentages In Baseball

In the other thread that MGL is referring to, I had said this:
http://www.insidethebook.com/ee/index.php/site/comments/catchers_strikeout_rate/#15

Anyway, if you presume the starter is going to pitch 6-7 innings, that gives your main catcher 3 PA, or a gap of 0.26 runs.  That’s on offense.  Defensively, say that Molina cuts that down .11 runs, so that the gap is now down to 0.15 runs.

So, I said Posada is +.26 runs on offense and -.11 runs on defense.  For an overall +.15 for Posada.

MGL’s expectation is 6.17-5.85 = +.32 runs on offense for Posada, and 4.18-4.05= +.13 runs on defense for Molina, for a total of +.19 advantage for Posada.

In either case, whether you do it the quick way like me, or the simulated way like MGL (albeit with different inputs), you get pretty much the same answer.

***

And his sim says .7215 minus .7133 = .0082 wins (about half what we expected, actually).

The Tango Distribution had .6784 using MGL’s runs scored/allowed for Posada and .6672 for Molina.  The difference is .0112 wins.  But, I think MGL must have done something different for him to get such high win% given that scored/allowed.

I tried posting this in the other thread yesterday, but it didn’t go through for some reason:

Looking at the numbers on Fangraphs, Posada’s RAR was 39.6 in 111 games this year, or .35 runs per game. Molina is essentuially a replacement level offensive catcher, thus the difference between the two is about .35 runs. This is very very close to what you get with your sim (.32 runs)

As for defense, you could look at their stolen base percentages and figure out each catcher’s value per stolen base attempt. I looked at their career stolen base percentages and the run values for a SB and a CS. I found that per stolen base attempt, Molina was about .07 runs better. Both catchers averaged about 1 stolen base attempt against them per game. It should be noted that even though I used career numbers, Molina was well below his career CS% this year.

That leaves .28 runs that Molina would have to make up on other defensive factors (fielding bunts, pop-ups, and nubbers in front of the plate) and calling a better game.

I agree with you MGL, it’s pretty much a wash. And I’m a Yankee fan

Not sure if this is indicative of how he’s played all season, but Posada looked absolutely terrible behind the plate on Wednesday, missing multiple catches on balls in the air when it didn’t even seem like signs got crossed.

I’m not sold on this analysis.

First, Jorge has been batting sixth, not seventh.  Also, he isn’t hurt at this point in any meaningful sense of the word.

More importantly, if the difference between one of the best hitting catchers in baseball and one of the worst amounts to less than a 1% improvement in the chance to win, why isn’t baseball full of Jose Molinas starting games?  You pay them a lot less than you do the Posadas.  I’d give up the less than 1% improvement in order to pay a major league minimum salary.  Don’t you think most teams would?

Or is it because the Yanks are so much better than the Twins anyway, you could change ARod for Cody Ransom and get the same answer too?

wave, you are right that the WP should be more than it is.  If it were only 1%, then yes, it would not make any sense to pay someone Posada money when you can pay someone Molina money.  No argument there.

A few things:

One, as I mentioned, you have Posada likely coming into the game in the 6th, 7th, or 8th innings, and being available to pinch hit even if he doesn’t come into the game.

Two, I don’t know why the sim only has the Yankees winning 1% or so more with Posada than with Molina.

There, whether he IS playing hurt or not, I programmed the sim the have him playing hurt, so that is one reason why it is only a 1% difference.

(Whether he hits 6th or 7th is not going to make any appreciable difference.)

Because the Yanks are so good, I get .0112, using MGL’s inputs.  If it was a .500 team, it would be closer to .019 wins using those same inputs.

I got similar results when looking at the win probability for tonight.  I held everything constant except for two things:

1) Scenario 1: Posada bats sixth
2) Scenario 2: Molina bats ninth

I get a Yankee win probability of 64.3% for scenario 1 and a Yankee win probability of 61.7% for scenario 2.  Like mgl said, if you do think there’s some benefit to having Molina catch Burnett it is effectively a wash.

More importantly, if the difference between one of the best hitting catchers in baseball and one of the worst amounts to less than a 1% improvement in the chance to win, why isn’t baseball full of Jose Molinas starting games?

One percent a game becomes a lot more significant over the course of a season.  In my scenarios above, I had the Yankee lineup scoring around 6.4 runs per game and allowing 4.29 runs per game with Posada.  With Molina it was more like 6.0 runs scored per game and 4.23 runs allowed.  That’s effectively the difference between being a .688 wpct team and a .664 wpct team when plugged into Pythagenpat. 

Over a full season that’s about 4 wins, 3 if you use about 120 games instead to account for the fact that catchers play less frequently.  I’d say the difference between Posada and Molina would be about 3 WAR over a full season.