Image credit: USA Today Sports

Has Mookie Betts surpassed Mike Trout as the Best Player in Baseball™? A week or so ago, this was the question du jour, at least among those folks who were sympathetic to Betts’ case (which is to say Red Sox fans). The response from the acronymati was a fairly quick dismissal of Betts’ case. Sure, Mookie is a talented lad, and perhaps there’s no shame in coming in behind a generational talent like Trout, but Betts would have to settle for (at best) a second-place trophy?

Why? Because the God of WAR said so. The general form of the #TeamTrout counter-argument was that someone got out the WAR chart from the past few years, circled a few numbers, and called it a day. Now, before we worship and bow down at the altar of WAR, I think it’s worth asking whether WAR might be underselling Betts a little bit. And the reason why comes down to one sentence: Mookie Betts is a reasonably competent center fielder. He’s a reasonably competent center fielder who has been playing right field for the Red Sox. And that has consequences WAR doesn’t quite know how to handle.

We know the Red Sox believe Betts is a competent center fielder, because in 2014 and 2015 they played him there on a regular basis, and when regular center fielder Jackie Bradley Jr. went on the shelf, the Red Sox have moved Mookie back to the middle. But in 2016 and 2017, the Red Sox, having JBJ on the roster, had the luxury of moving Betts to right field, and he played there exclusively. And so, for the purposes of WAR, Betts became a right fielder and was compared to other right fielders.

There is something of an app for that within WAR. The “positional adjustment” is supposed to account for that, and Betts himself is a decent illustration of how the principle works. Betts’ defense in right field–on all accounts–has been stellar. His defense in center field has rated out as slightly above average. Given that center field is a harder position to play (and those who play there tend to be better defenders than those in right field), it makes sense that the same player would get better results in the easier position. But before we call the case closed, we need to account for something.

Last fall, I started talking about how these positional adjustments are a little deceiving. They’re calculated by looking at the performance of players who appeared at both positions. The adjustment between center field and right field comes from looking into players who did enough of both to get a good read on their performance out there. But that’s a very biased sample. It’s not a random selection of right fielders who get to spend any time in center field. Some guys are in right field because they have a great bat and their team needed to hide them somewhere. Those guys never play center field, because if they did it would be downright hilarious to watch. That sample bias gives us a false sense of how different the two positions are.

If those bat-first right fielders tried their glove at center field, we’d have a better idea of how far apart the average right fielder is from the average center fielder, rather than how far apart the above-average right fielder is from the average center fielder. Maybe we’d even have a better sense of how that adjustment might not be so linear. The positional adjustment essentially says that a right fielder would lose X runs’ worth of value if he moved to center field, both because of the difficulty of the position and the better competition. But what if X varied to the point that good right fielders don’t lose as much value when they shift to center field, but really bad right fielders lose much more than X?

Given that MLB teams don’t assign playing time randomly, we don’t really have a way to know whether this is true or not, but it’s a reasonable theory. If that’s right, the fact that Betts has the defensive chops to play center field means he probably deserves a little more credit compared to his right field brothers than the defensive positional adjustment might give him. That would (in theory) bump his WAR up.

But that isn’t all Betts’ (hypothetical?) talents in center field meant to the Red Sox, even when he wasn’t using them for two years. (Part of that is because the Red Sox used Andrew Benintendi in center field as well.) Consider for a moment what Betts’ flexibility allowed the Red Sox to do—or at least could have allowed them to do. When filling out their bench, they didn’t have to settle for dedicating one of their backup outfielder spots to a guy who could “handle center.” Betts (and Benintendi) were their de facto backup center fielders. They could get away with finding someone who was merely an average corner outfielder and who had a decent stick to sit on their bench. And that makes a difference.

In 2017, the 30 players who logged the 10th-most plate appearances for each MLB team had an average run value produced (using linear weights) of negative 0.67 runs per 200 plate appearances. (This isn’t surprising. They’re on the bench because they’re below-average hitters.) Here’s the rest of that chart.

Batter Runs per 200 PA
10th-most used -0.67
11th-most used -2.25
12th-most used -1.20
13th-most used -4.66
14th-most used -6.38
Everyone else on the bench -8.06

If a player has the sort of bat to be at the top of that chart and has defensive skills good enough to be average in center field, he’s probably starting somewhere. So, teams are often playing a game of “one or the other.” You can have a guy who plays a decent center field but won’t give you much with the bat, or a guy who will probably be more suspect in the field but might give you more at the plate.

But what if they didn’t have to choose? They could still have the option of adequate defense in center field if their center fielder got hurt or needed a day off, and they could still pick a bench bat more likely to give a touch more production. It’s hard to know exactly how much that might be worth, but we can see that the order of magnitude is somewhere around five or six runs, and Betts’ flexibility creates the circumstances where that transaction can happen. In true WAR fashion, I propose that we should credit him with that value whether the Red Sox actually went out and did something about it or not, in the same way we credit a player for hitting a double and getting into scoring position, whether or not his teammates knock him in.

There’s a further effect down this trail if you want to follow it (although, fair warning, some of you aren’t going to like this scenery). Since the Red Sox are now freed from the obligation of having one of their backup outfielders be a guy who can “handle center,” what if they decide to use that roster spot not for a backup outfielder, but for some other use? What if, rather than carrying two backup outfielders–one a bench bat for pinch-hitting purposes and occasional starts in left or right field, one a bench glove–they simply carry the bench bat and move Betts (or Benintendi) over to center field if the game situation (or the need for an off day) calls for it?

Here, the fact that Betts is one of the best hitters in the league also plays a role, since it’s very unlikely that the Red Sox would want to tactically remove him from the game. If you have a good right fielder whom you never want to take out, it becomes a little sillier to have a guy on your bench whose primary role is “backup outfielder.” What’s he going to do? And more importantly, if he’s not doing much, could that roster spot be put to some other use? (Sure, there’s going to be risk that if you don’t have a couple of backup outfielders you get exposed if an injury pops up, but at some point, the marginal value goes down far enough that you have to start asking that guy “what exactly do you do around here?”)

Maybe the Red Sox could carry a platoon bat at another position. Maybe they could carry a guy who’s primarily a pinch-runner. Maybe … oh, let’s be honest, it would be another reliever. It’s always another reliever. (That’s the part some of you won’t like.)

This puts us further afield than we ever could have imagined. We began with the idea that WAR didn’t recognize Betts would have made an adequate center fielder, because he never actually got to play center field in 2015 and 2016. And now we’re saying one of the effects of the thing he never actually did (but could have done) was that it might have allowed the Red Sox to do better … on the mound?

Maybe that roster spot Betts saves is another LOOGY who grabs the platoon advantage a few more times during the season. If Betts isn’t a decent center fielder, maybe the Red Sox feel more pressure to dedicate that roster spot to a true backup center fielder, and they miss out on those extra plate appearances with the platoon advantage. Maybe the extra guy is a sponge who sucks up low-leverage innings and keeps the other “real” relievers a little fresher during the season. It’s hard to figure how much that might be worth, but the effect would probably be positive.

Maybe … well, remember earlier in the spring when the Rays were talking about “Bullpen Day”? I suggested that the true benefit of Bullpen Day wasn’t the actual fifth day when they didn’t really have a starter, but the fact that the Bullpen Day system meant they had a bunch of guys around who could handle pitching 2-3 innings at a time. And sometimes, that would mean being able to lift a starter earlier in the game than normal, saving his arm, but also saving him from going through the lineup a third time. A little math showed the effect of that sort of reliever might just outstrip the effect of the platoon advantage. What if the Red Sox, freed from the need to carry a backup center fielder, got themselves one of those and didn’t have to worry about shortening up the rest of the bullpen? Betts allows them that option, even if they don’t take it.

Would all of this be enough to vault Mookie over Mike? I don’t honestly know the answer to that. It’s probably “no” given the fact that Mike Trout is Just. That. Good. I can’t even really put solid numbers on any of the things I’ve noted above. (And that’s kind of the point. We haven’t explored them yet.) But I can make a decent case that the issues discussed are real effects that WAR just doesn’t know how to take into account, and even if we can’t put a number on them we know which in direction they probably point. The fact that Betts could have (even though he didn’t) play center field for the Red Sox may have had benefits to his team that WAR doesn’t reflect, under-selling Mookie. So when we pull out the WAR chart in the Betts vs. Trout debate, are we really getting the full picture? Is it perhaps a closer race than we thought?

I didn’t write this piece to wade into the actual Betts vs. Trout debate. I wrote it as a lesson in the limits of WAR. While it’s a wonderful stat and much better than what came before it as a measure of player value, it’s always worth thinking about the ways in which it might not be fully reflecting reality. The case of Mookie Betts just happens to show off a few of those weaknesses perfectly.

Thank you for reading

This is a free article. If you enjoyed it, consider subscribing to Baseball Prospectus. Subscriptions support ongoing public baseball research and analysis in an increasingly proprietary environment.

Subscribe now
You need to be logged in to comment. Login or Subscribe
Would Ohtani offer another example of WAR being inadequate since he, similar to Betts, allows his team another roster spot?
Russell A. Carleton
Ohtani is the answer to all questions! (And yes, I think he fits the bill.)
Mr. Carleton, I always enjoy you articles, the more Gory Mathematics the better. WAR is certainly a worthwhile judge of overall value but I will also aruge that WAR has some real problems. It simply has no way to adequately quantify the true value of a dedicated DH. As it stands the "position", which is, of course, no position, is penalized 17.5 defensive runs when the DH never even plays defense. It would only make sense to penalize the DH if the team was forced to play without a defensive player in his place. In reality, the fact that the AL uses a 10 player starting lineup makes it impossible to compare that with the NL where everybody plays the field. A DH should only be judged by his offensive contributions and anything else is silly. On the Trout vs. Betts debate, playing in Fenway skews the situation and does not give Betts enough credit. RF in Fenway is tougher to play than the symmetrical CF in Anaheim. While they are both great I give Betts a slight edge on defense with a better arm and he might be the best base-runner in the game since Jackie.
Byron Hauck
Even if you concede that there's real value in the roster flexibility, I don't think that flexibility can't be solely attributed to Betts. Bradley and Benintendi are both involved in creating it too, aren't they?
Russell A. Carleton
I think you're right that some of the credit should go to Benintendi since he is also good enough to handle CF. Maybe there's even something to the idea of "if only Betts, no added value; if only Benintendi, no added value; but because Betts and Benintendi are both there, the value is added." But _some_ of that value goes to Betts. And honestly, I'm only marginally interested in figuring out how many more WARs should be properly credited to Mookie.

Part of the problem is that we're very culturally focused on who exactly should get individual credit for everything, and we have no concept of or word to describe the idea that value can be created through the interaction of players or how a player fits within the system. WAR is a great stat, but it has no way of accounting for that sort of thing.
Russell, you missed the clearest example. When playing in NL parks, the Red Sox can substitute JD Martinez for Bradley in the OF rather than replacing Betts or Benintendi or keeping Martinez on the bench
Jonathan Beilfuss
So would Trout be more valuable if there was a better center fielder on his team, so he could play another OF spot and also be the backup CF?