The Red Sox Have to Bench Somebody

After the first two games of the World Series, the Red Sox are probably feeling pretty good. With two victories already to their credit, they need to win just two of their next five contests to win this year’s title. The odds are in their favor.

As they begin the the first of three games in Los Angeles, however, they have a decision to make regarding their lineup, owing to the absence of the designated hitter in a National League park. The issue comes up every year, but rarely does it present much difficulty to decision-makers. Most American League champions feature either feature a DH who isn’t worth playing elsewhere or an obvious weakness somewhere in the lineup.

That isn’t so much the case for J.D. Martinez and the Boston Red Sox, however.

Consider some examples from recent World Series. In 2014, Billy Butler DH’d for the Royals in Kansas City but only pinch-hit in San Francisco. Kendrys Morales played the same role for the 2015 edition of the Royals. Last year, it was the Astros’ Evan Gattis. Only Cleveland, during their series against the Cubs, was forced to take some unusual measures to include their usual DH in the lineup, placing Carlos Santana in the outfield. That said, Cleveland also lacked a surplus of great outfield options at the time.

The last legitimately good AL player forced from a lineup when the World Series headed to an NL park was probably Kevin Youkilis in 2007, when the presence of David Ortiz sent him to the bench. But the current version of the Red Sox will be forced to take similar steps in Los Angeles. Martinez was a six-win player this season, with a 170 wRC+ as Boston’s designated hitter, marking the best season by a DH since Ortiz’s 2007 campaign for the world champions. Martinez needs to go to an outfield corner, and even if he we assume that he had just bested (worsted?) the single-season low for UZR (-36 runs, set by Brad Hawpe in 2008), he still would have recorded nearly four wins this season. He needs to be in the lineup, right?

It would seem so, yes, but once Martinez goes into the outfield, one of the following scenarios needs to occur:

  1. The Red Sox bench Jackie Bradley Jr.; or
  2. they bench Andrew Benintendi; or
  3. they play Mookie Betts at second base.

Mike Petriello made the case for the third option over at In that piece, he notes that the decision is easier in games Four and Five with a lefty on the mound, mentions the lack of potential opportunities for Betts at second base based on Red Sox pitching and Dodgers hitters, says the offensive numbers are a wash, and concludes thusly:

So if the lineup choices are a wash, this comes down to defense. Martinez is a sizable step down from Betts in right, but he’s got to be there if we want his bat in the lineup. Betts is likely a small step down from Bradley in center, probably a barely perceptible one, but that matters with Martinez in right. And at second, Holt is worse than Kinsler, but better than Betts. You could argue in either direction. Either way, Betts and Martinez must both play, and Bradley’s defense is more valuable than Holt’s.

Maybe putting Betts at second base is too radical for the World Series. Perhaps Cora will just keep it simple and bench Bradley, starting Holt or Kinsler at second base. It’s the easy way to go, and Red Sox fans know better than most how hard one poorly-timed defensive miscue can sting. But October is when the tough choices mean the most. Putting Betts at second is far from traditional. It just might be the right thing to do.

Petriello presents an interesting argument, but I’m not sure it is the correct one. Instead of separating offense and defense, let’s try to combine them. First, here are the player projections with a slight platoon adjustment for the lefties against a potential righty in Walker Buehler.

Red Sox Lineup Options
Name WAR/600 Projections
Mookie Betts 6.8
J.D. Martinez 4.2
Andrew Benintendi 3.7
Jackie Bradley Jr. 3.5
Ian Kinsler 2.8
Brock Holt 2.1
Bradley Jr., Benintendi, and Holt bumped up 0.5 WAR for platoon advantage

That’s what things look like when the players are all playing their normal positions. I’m leaving Martinez where he is, since the DH penalty is roughly equivalent to a -10 outfielder, a figure that seems about right. In the scenario where Betts goes to second base, we have to remove about a win off his value because of the change in defensive position. He’s a great outfielder, yes, but given his lack of recent play at second base, it’s probably not reasonable to expect even average defense from him, regardless of how many grounders he’s taken over the last few years. Putting that into the calculation, we see these numbers for the four outfield slots plus second base.

Red Sox Lineup with Betts at Second Base
Name WAR/600 Projections
Mookie Betts 5.8
J.D. Martinez 4.2
Andrew Benintendi 3.7
Jackie Bradley Jr. 3.5
TOTAL 17.2
Bradley. Jr. and Benintendi bumped up 0.5 WAR for platoon advantage. Betts loses a win for defensive change to second base.

Maybe Betts is better at second base than we think. Maybe Martinez is worse in the outfield. What happens if we replace Bradley with Betts in center field and put Kinsler at second base.

Red Sox Lineup with Bradley Jr. out
Name WAR/600 Projections
Mookie Betts 6.8
J.D. Martinez 4.2
Andrew Benintendi 3.7
Ian Kinsler 2.8
TOTAL 17.5
Bradley. Jr. and Benintendi bumped up 0.5 WAR for platoon advantage. Betts loses a win for defensive change to second base.

Betts isn’t likely to lose any value in center field given his range and arm are just as good as Bradley Jr.’s, even by Statcast metrics. Whatever he would theoretically lose in terms of fielding runs from the move, he would gain back by means of the center-field positional adjustment.

Whatever the case, it appears as though the difference between the two alignments — if one exists — is just a few runs. And while that’s minuscule over the course of one game, logic dictates that, when the radical solution doesn’t come out any better in the numbers, it probably isn’t worth the downside risk. The argument holds whether Benintendi or Bradley Jr. is benched. Given that Bradley Jr. seems more likely to hit the bench against a lefty on Saturday and Sunday (if necessary), I think it raises the argument that it is Benintendi who should sit against Walker Buehler.

Given Benintendi’s spot in the batting order, it is fairly clear that Alex Cora feels it is his left fielder who hits lefties (and righties) better than Bradley Jr. The projections (with defense) put the two players as near equals. Against Buehler, that means Cora is simply choosing offense or defense in Los Angeles with the other a potential pinch hitter. Neither option is a bad choice, but since Bradley Jr. still projects competently on offense against a righty, choosing defense over offense — in light of Martinez’s presence in the field — it makes sense to go with Bradley Jr. over Benintendi. The Red Sox center fielder has been unlucky with the bat this season by xwOBA, and it is possible projections are underrating his ability.

The graph below shows all players with 500 plate appearances with their xwOBA and wOBA, per Baseball Savant. Those players above the line are underperforming their xwOBA, while those players below the line are overperforming it.

Most of the players who undershoot their wOBA based on their xWOBA are slow as seen by the table below.

Biggest xwOBA Underachievers
Player wOBA xwOBA wOBA-xwOBA Sprint Speed
Kole Calhoun .283 .335 -.052 26.1 ft/sec
Victor Martinez .281 .324 -.043 23.2 ft/sec
Joe Mauer .319 .355 -.036 26.0 ft/sec
Chris Davis .239 .274 -.035 25.5 ft/sec
Nelson Cruz .361 .394 -.033 24.9 ft/sec
Salvador Perez .304 .337 -.033 25.1 ft/sec
Alex Gordon .305 .336 -.031 25.5 ft/sec
Jackie Bradley Jr. .311 .340 -.029 27.8 ft/sec
Joey Votto .370 .396 -.026 25.4 ft/sec
Ryon Healy .296 .322 -.026 26.0 ft/sec
Marcell Ozuna .327 .352 -.025 27.8 ft/sec
Yangervis Solarte .285 .309 -.024 24.7 ft/sec
Joey Gallo .343 .366 -.023 27.7 ft/sec
Jose Martinez .356 .378 -.022 26.5 ft/sec
Kyle Seager .288 .309 -.021 25.9 ft/sec
Jose Abreu .337 .358 -.021 26.7 ft/sec
SOURCE: Baseball Savant

It’s a bunch of really slow guys, a player who gets massively shifted in Joey Gallo, and then Marcell Ozuna and Bradley Jr. The Red Sox center fielder has hit a lot better than his batting line indicates. At least against a righty starter, he deserves to play. With the pitcher’s spot inithe lineup, Benintendi will still have an opportunity to impact the game off the bench. It certainly feels odd to advocate benching one of Boston’s best players in Andrew Benintendi, but I suppose that is one of those good problems to have.

Craig Edwards can be found on twitter @craigjedwards.

Newest Most Voted
Inline Feedbacks
View all comments
5 years ago

Bench Jackie, Benintendi’s bat is too good and JD gets subbed for a runner at some point anyways.

5 years ago
Reply to  nktokyo

Either for a runner or maybe more likely defensively.

5 years ago
Reply to  nktokyo

While JBJ is of course an all-world defender it seems funny to bench Benintendi for defensive reasons after the whole “Air Benny” experience in game 2.

Cory Settoon
5 years ago
Reply to  nktokyo

Second half numbers vs RHP. Tell me again who’s bat is benchable?

JBJ – .301/.369/.531
Benny – .273/.350/.385

You could argue how well Benny has played thus far (against LHP too), but most of those hits just sneaked through or barely dropped in.

5 years ago
Reply to  Cory Settoon

A partial season split of a platoon split? Am I still on fangraphs? That’s ~200 PA worth of data.

Across ~1100 PA, Benintendi is a career 125 wRC+ hitter vs. RHP.

Across ~1600 PA, JBJ is a career 95 wRC+ hitter vs. RHP

If you want to *weight* those for recency, use the whole of 2018. Jackie was a 103 wRC+ to Benintendi’s 135.

5 years ago
Reply to  nktokyo

OOF. Gotta hate it when your reasonable process leads to the wrong results like that.