October 16, 2014
The Molina Mind-bender
When Yadier Molina went to the bullpen to warm up Trevor Rosenthal in Game 3 of the NLCS, the speculation was totally warranted. Why not bring Molina in to catch with his oblique injury? That’s something that affects the swing, and if you can get some defensive work out of him, that’s where so much of his value is anyway.
In that light, we had received one of my favorite questions to the Effectively Wild podcast, which Ben and Sam passed along:
“If Yadi can do everything at close to 100 % but swinging would kill him, do you start him and just instruct him to take at the plate?” – Friend of the site Matt Trueblood
It’s a fun question and feels like a legitimate one. Is it possible that Molina’s defensive value, including the traditional measures found in fielding runs and in framing ability, completely outweighs his worthlessness at the plate?
We decided to take a stab at it, starting with the offensive side to see just how much he would have to make up in defense for the Cardinals to try this. Along the way, there will be several assumptions made as there would have to be in talking about a guy basically walking up to the plate without a bat (a common EW theme). This is not a request to accept them blindly – I’ll try to justify them and point out some sensitivities to changing assumptions – but don’t let them get in the way of your enjoyment of the hypothetical.
Assumptions about Molina at the plate:
· He cannot and will not swing the bat, but he can bunt.
· When he’s not bunting, he will take every pitch. Pitchers pretty consistently year to year throw 50 percent of pitches inside the zone and 50 percent outside, but that’s against hitters who can do actual damage. We’ll use 67 percent to reflect Sam’s research on the 3-0 pitch to pitchers batting – a good proxy for a guy just aiming for the strike zone against a take. This leads to taking a walk 10 percent of the time and striking out 90 percent, but that can vary based on our assumptions.
· He will bunt on basically the same schedule as a pitcher including the borderline case of man-on-second with one out. He will not squeeze.
· We’ll consider bunts a binary outcome with either successful sacrifice or strikeout/forceout, which are the same base/out result, and he’ll have three shots at it instead of swinging away on two strikes. This is an oversimplification, but since a double play is actually more likely with Molina running than a single, we’ll shade down the percentages in “The Book” and call it 70 percent success.
· Since he is batting after the pitcher, the likelihood of coming up in each base/out situation was determined by how many times the Cardinals’ leadoff hitter faced each of the 24 base/out situations in the last three years, ignoring the first PA of each game.
· He will not be pinch-hit for since that violates the spirit of the question, but much more on this later.
So here are the results, with the first column being a weighting based on how many times each scenario came up for the Cardinals’ leadoff spot hitter after the first inning.
Taking the weighted average of those changes in run expectancy, we get that a torque-free Molina costs the Cardinals 0.20 runs every time he steps up to the plate.
As it turns out, this is not all that sensitive to our assumptions, especially the more difficult assumption about the probability of getting the bunt down.
If he bats four times per game, which is reasonable for the No. 9 hitter in a game that goes nine innings (the nine innings staying consistent for defense purposes) at playoff scoring levels, that’s 0.81 runs per below average.
None of this is so easy with A.J. Pierzynski, whose plate appearances can’t be broken down so nicely given the number of outcomes associated with swinging the bat. So we’ll take his general batting runs above average from 2014, which is -10.0 in 362 plate appearances or -0.11 runs every four plate appearances.
Thus, we need Molina to make up 0.70 runs in the game on defense, which feels ambitious, but we’ll try.
- If we go back one year, Pierzynski is a -1.0 fielder in 721 innings, and Molina is a -1.2 fielder in 932 innings, which is almost an exact wash.
- If we go back two years, Pierzynski is a -2.2 fielder in 1,726 innings, and Molina is a -2.0 fielder in 2,047 innings, which is an edge to Molina of 0.002 runs per 9.
- If we go back three years, Pierzynski is a -4.5 fielder in 2,797 innings and Molina is a +0.4 fielder in 3,208 innings, which is an edge to Molina of 0.016 runs per 9.
So even if we use a three-year range, round up and assume we’re getting a completely healthy Molina, this is still only 0.02 of the 0.70 runs needed. So most of it will have to come in framing and blocking, and using the three-year range is also friendly to Molina in total receiving runs (framing + blocking).
- Pierzynski is a -20.1 receiver in 2,797 innings and Molina is a +34.8 receiver in 3,208 innings, which is an edge to Molina of 0.16 runs per 9 innings.
That’s not nearly enough. Even in these most optimistic of scenarios for him, with favorable windows on defense and only counting a bad offensive year for Pierzynski, there’s still half a run per game that he’s unable to make up with his defense.
The verdict on the question of playing Molina would be a pretty clear no unless you're willing to ascribe half a run to the non-framing elements of his handling of the pitching staff, which seems extreme.
However, there is another alternative to this method and to the idea of bringing him in late in the game, and it’s an idea that will work better in any Cardinals home games at the end of the series. The idea would be to start Molina, bat him ninth and then take him out when his spot comes up in the order.
(Personal aside: This was a strategy I invented when I was 11 and playing with the 1996 New York Yankees in Strat-o-Matic. Mariano Duncan hit .340/.352/.500 that year but was a bad defender, and since the 1996 Yankees obviously needed the help, I would start Luis Sojo at second base and just take him out when it was time to hit in a couple innings. You probably invented this strategy too.)
This way, you get a couple innings out of Molina, which you might not be able to do if you’re just saving him for a save situation. And you never have to sacrifice an at-bat, which you might have to do if a save situation goes awry.
It’s also flexible. If he’s up with two outs and nobody on, which is the least harmful ∆RE situation in the above chart, you can leave him and try to get a few more innings out of him. If you have a lead already and think it’s a spot where his defense will take on an exceptional amount of importance in relation to his offense, you can leave him. (Though this is not generally recommended considering how big the difference is.)
This is never an idea I’d endorse if there were only two catchers on the roster, but with Tony Cruz there too, you’re starting in the second or third inning just like you were starting a game with two (bad) catchers.
There’s just no way to make the math work on playing him throughout the game if he can’t swing, but there is a way to use his defense in some small ways.
Thanks to “The Book” and the Baseball-Reference.com Play Index for research assistance.