THE BOOK--Playing The Percentages In Baseball

I’m surprised that the playing catcher disadvantage isn’t larger than the playing DH disadvantage for guys who spend significant portions of their careers in both roles.  Victor Martinez and Mike Napoli are guys who seem to bat better when not catching, not that two anecdotal examples without looking up whether they even hold water proves anything.

Any plans to investigate that more directly?

Now, how about 1B?  Surprisingly, it’s harder to hit at 1B!  It was .347 at 1B and .352 at the other six positions I noted above. 

I’m inclined to think that 1B may be used as an injury position as well.  So, a guy who is not 100% to play the field, but the manager wants his bat in the lineup, is placed at 1B to a disproportionate extent.  And this injury affects his hitting a bit.

Just guessing.

Mike Napoli as catcher in 2011: 354/437/667

only 54 games, but sweet lord!

Career: 265/360/516, compared to 263/358/509 overall (including his time at catcher).

Tango, even though you only used players aged 25-29 (peak years on both sides), I would still like to see the average age for each position at each pair.  There could be some age bias going on.

Also, hopefully, there isn’t too much of a selection bias going on, but there could be.  For example, maybe a player starts out at 1B, doesn’t hit well, and his team says, “He doesn’t hit good enough for a 1B, let’s move him to another position.” So it seems like it is hard to hit at 1B.

Same thing at third, which may be why third baseman also seems like a “hard” position to hit at.

Great stuff, BTW!

Could there be platoon effects here?  Eg., At position A you have a LHH who doesn’t hit LHP.  So you platoon him with a RHH.  Against RHP the LHH is at position A while his platoon partner, who hits RHP reasonably, but not as well as he hits LHP, plays position B.  This would provide a systemic bias between that player’s performance at the 2 positions.  Aren’t the differences we are seeing on the order of magnitude of average platoon differential?

Excellent point about platooning affecting what position he plays.  Like in the other study, I should include handedness, force it only against SP, and might as well control for times through the order.

***

MGL: it’s simple enough for me to just include it as well, so let’s see what it gives us.  It’s important to note that if we find say the C/1B pair as these average ages:
26 C
28 1B

Then that’s perfectly fine, IF we accept the peak age is 27.

The reality is that 25-29 is such a plateau, that any difference from the theoretical peak (whatever that is) is going to be slight in impact.

In any case, I’ll add it, and let’s see what the data says.

One thing about the catchers, is that playing catcher could have a negative effect on hitting, even when they are playing first base for the day. 

I don’t know how you could find a better answer for how much playing catcher affects batting, unless you look at players who played catcher a lot one season, then not as much the next season.  But again, you’d be running into questions of why they were playing more or less catcher.

Anyway, lots more work to be done.  But, without question, we need to apply a huge bonus to guys who DH,

Now, I did not try to distinguish between “lifers” at DH, and just those who are transient.  It’s possible the reason some players were DHing was because they had some injury that was hobbling them on the field, but was hoped to be masked while hitting.

Thanks for the study, Tango. However, it’s not sufficient to find correlation.  One must also attempt to discover causation.  In this case, before trying to apply a hitting bonus across the board to all DH’s, I think the nagging injury issue should be addressed.  I could imagine someone (perhaps sloppily) doing the same study involving just the DH/other position data sets and concluding that minor nagging injuries costs a player 16 points of wOBA rather than concluding that being a DH somehow makes it harder to hit a baseball.

"However, it’s not sufficient to find correlation. “

Obviously.  That’s why I presented a list of causative agents.

As for doing the work to determine the impact of the causative agents, well, the world is full of aspiring saberists.  I did my half, and someone out there can be inspired to do the other half.

Until that happens, I am not going to just presume that the best adjustment to use is zero, until we get things straightened.  I have to use *something*.

This is like the MLE issue.  Yes, there are a bunch of nagging issues to consider.  But, in the absence of answers to those questions, you take your best shot AND accept a certain level of uncertainty in the results.

That’s how life works.

I am grateful for the work, Tango!  I have been thinking that the DH penalty is a bit too severe, given just how unique a skillset it appears to be and how rare it is to find a guy who succeeds there...but I don’t have the statistical chops to do these regressions.  Thanks for devoting your valuable time and energy to this issue.

"But, without question, we need to apply a huge bonus to guys who DH, and to guys who play catcher.”

Without question? Forgive me, I’m still skeptical of the idea of a DH penalty. Have you looked at how DHes do in their first plate appearance of the game, when DHing and when playing the field? There was a study in the BBBA years ago and found that players hit less when DHing then as well, when the there was less of a difference, which led the one studying (I think Tom Ruane) to conclude injury was the reason. It might have been overreaching as well, but it does lend itself to that conclusion.

There might be something to this for catchers as well.

Could you take the guys who played X number of games at DH in a season (say > 80) and compare their batting to other positions for that season only?

Ok, I included a match on:

batter hand
pitcher hand
times through the order
starter/relief

The results for DH:

Things overall did not change (16 point difference remains), though you had a bit of movement for each position.

By the way, the pattern repeated itself for players aged 18-24 and 30-49.  (Above chart is for ages 25-29.)

***

I also looked at pinch hitters compared to other non-starters (i.e., defensive subs).  They were .291 as PH and .308 as defensive subs.  That’s a 17 point difference.

Note that Andy in The Book noted that the difference between PH and being a starter was around a 30 point difference.

So, it’s interesting that the gap between PH and defensive sub (remember, it’s the exact same players, with the same conditions of handedness and times through the order, etc), is identical to the gap between DH and starting player.

This would seem to me to show that “playing the field” is the variable that links to the two.

Are you comparing only same-season plate appearances or are you aggregating seasons?

Aggregating over ages 25-29.

MGL: to do the age thing will be a bit harder than I thought it would be…

OK. And mgl brought up the age idea already, which might explain something too (though MLB teams use the DH extensively in the minors for reasons that escape me; when did they care about developing DHes?).

Another one, unrelated to DH, is playing out of position. If a player is playing a position he doesn’t play often, is there a difference?

I really wish there were ample data to do this for pitchers.

Charlie: your idea is similar to what I did with comparing starting pitchers and relief pitchers.  If you have a starting pitcher pitching in an emergency situation in relief, his performance is not better (indeed, it’s a bit worse).  Basically, a full-time starter is not going to start to change his approach.

So, your question would be along those lines.  I can break down the players into groups, based on which is their primary position over their age 25-29 seasons.

Exactly what I was thinking.

At some point, this all needs to go into a summary. “Computer baseball sim designers? Some tricks to help you improve the realism of your game!”

If there is something about playing defense that helps with hitting (e.g. staying warm, loose, or “mentally in the game") then I wonder if batters hit better when they’ve had more fielding opportunities in the previous half inning?

If so, that would also suggest that DH’s might improve their hitting if they could establish a routine between AB’s that emulates whatever advantages come from fielding.  (Though possibly that tradeoff isn’t worth it if such a routine brings about more fatigue that eventually drags down performance over the course of the season.)

Michael, good point.

What’s interesting to counter that however is that a player hit better when he was in the outfield than in the infield.

So, it could be that you want to be involved, but not TOO involved.

I seem to remember seeing a study of how a hitter performed if he made the last out as a fielder, and was the leadoff batter of the next half inning.  From what I remember, there was no change in performance from his seasonal performance.  I can repeat that study, but if someone else wants to do it, please feel free.  (Where are all the aspiring saberists anyway?  As Bill James once said: I can’t do this alone.)

***

Or, maybe the guys hit better in the OF than in the IF because they were seen as struggling on defense (which affected their hitting) and then moved to the OF, where they could relax.

We can sit here and come up with dozens of plausible explanations!

Relating to Tango’s point in #22:  I think that there’s an ideal level of stimulus when you’re hitting, just like most tasks (think about your most productive hours spent writing, reading, etc.)

The ideal time to read a book for comprehension isn’t when you are jogging at 9.0 on the treadmill.

So...extrapolating to baseball:

*Catching is not the ideal level of activity for a hitter - you come into the dugout more tired than anyone else, quickly remove gear, scramble up to the on-deck circle, etc.

*DHing/pinch-hitting is too lazy, or too little stimulus for your brain.

*Pitching, like catching, is a grueling task that leaves you tired when you re-enter the dugout and if we had an adequate sample of pitchers playing another position, my hunch is they would perform better as hitters when they were position players (and not just because they took BP more regularly).

A way a manager could test this:  Have your DH be one of the ballboys in a folding chair down the foul line during the half-innings where the team was in the field. 

He would be almost as mentally involved in the game as the seven hitters in the field.

The catcher thing is interesting too.  We can look at how catchers do when leading off an inning.

And we can look at if their “times through the order” effect is stronger with catchers.

I mean, at least one of the two has to happen in-game.  If there’s a powerful game-to-game difference when going catcher to a non-catcher position, then in-game, he’s either getting beaten up a bit at a time during the game, or he’s getting beaten right after catching the third out of an inning and coming to bat to leadoff.

Those are worthy things to look at.

This isn’t really related, but has nagged me for awhile: do batters (really, just the lefties) who are platooned have bigger platoon differentials than those who are not?

I’d check first to see if Andy did the study in The Book.  Can you see?  If not, then I can maybe do a study on it.

I’m 99.9% sure there’s nothing in The Book about this—that there was a platoon skill among right-handed batters at all, albeit a small one, was handled as a major revelation—but I’ll look when I get home.

"So, it could be that you want to be involved, but not TOO involved.”

I think this is very close; there is probably more stress when you are involved in most plays (including holding runners on, anticipating bunts for 1B).  And, of course, the hot corner is known as such for a reason- and 3B tend to be more tensed when preparing to field their position than other players, which may be a factor as well.

It’s interesting to note that this argues for a fairly large mental-preparedness component to hitting.  I guess all those sports psychologists have a reason for being in business after all.

I wonder if this might be some sort of chemical/hormonal effect, also.  I know I read a hypothesis that home field might actually be due to elevated testosterone levels when ‘defending your home.’ Could players involved in the whole game have higher adrenaline levels, testosterone levels or some other chemical effect that simply can’t be replicated on the bench or by being a ballboy or trying to stay mentally focused. 

Perhaps your DH should play ping-pong against the clubhouse attendant so he can keep his (literal)competitive juices flowing.

TT: Nothing about it in The Book. (I have it by my computer to know when to issue a free pass in DMB.

Looking at the data in #14 again, it occurs to me that 2B & CF (which have the greatest diffs) are probably the least likely (after SS) to be used as DH’s unless they are either hobbled with a minor injury or the manager is trying to give them partial rest (in some cases because it’s a day game after a night game or a doubleheader).

Ooh, excellent point.  Ok, I can buy the CF.  It’s not like you have competing CF and you are going to put the other guy at DH.  At worst, you put him in LF.  So, to actually put a CF at DH, and you are likely finding someone who can’t play the field much.

The others, I’m not so sure.  But CF I think sounds great.

This time I broke it up based on how often they batted at each of the paired positions.

For guys who spent at least 90% of their time at one position (and less than 10% at the paired position), I Call these guys “regulars playing out of position”.

Excluding C and DH, these guys were .342 wOBA at their primary position, and .337 at their out of position.  (Based on 22,830 PA.) The difference is 1.5 SD from the mean.  MAYBE there’s a small impact to playing out of position.

For the semi-regulars (played 75% to 90% at the primary position): .338 at the primary position and .334 at the secondary position.  (90,894 PA.) That’s 2.4 SD. 

Put the above two together, and we can say that there’s some small effect to hitting when you play out of position.  Not much.  Maybe 1 or 2 points.

***

As for catchers: full-time catchers are .341 at C, and .318 elsewhere.  Presumably they are hurt, but they want to keep their bat in the lineup.  Only 990 PA, so 1.4 SD.

Among all the other catchers (less than 90% of their playing time at catcher): .321 at C and .350 (!) elsewhere.  6916 PA, so 4.8 SD!!

In this case, these are probably guys who tried their hand at catching, and they got moved from behind the plate.  With shackles removed, their hitting exploded.

This is just a big wow as far as I’m concerned.

***

With DH, not much difference between full-time DH and emergency DH. They all hit much worse at DH.

Thanks. That was worth knowing.

Interesting stuff.

I don’t suppose that it could be that big of a difference but could their be a league affect here?  AL pitchers have been tougher than NL pitchers over this span and a chunk of this sample could be from interleague play (AL DH’s playing the field against some NL pitchers, and NL position players getting a break and playing DH against AL pitchers).

The other thing I’d be interested in (if there’s enough data to see it) is whether there’s a night/day or hot/cold split here if it’s really about getting warmed up or accustomed to the lighting or whatever.