For all the discussion about whether the Cubs would dare to put Kyle Schwarber in left field, perhaps we focused on the wrong defensively challenged ex-catcher trying to make it work in the outfield for the sake of offense. After much speculation, the Indians decided to start Carlos Santana in left field for what should be a windy and wild Game 3.
They're doing that in order to keep Mike Napoli's bat in the lineup, and they're doing it because Coco Crisp (the likely left-field starter otherwise) is a bat you can live without. But defense matters, and Santana is a designated hitter, with professional left-field experience that consists entirely of a handful of games in the low Minors in 2005-06 and four mop-up innings late in a 14-1 Boston rout in '12.
There's little doubt that Napoli's bat is superior to Crisp's. There's also no doubt that Crisp's defense is superior to Santana's, and that's really the choice. Which carries more weight? Let's dig into the numbers and try to see if the decision holds water.
:: Complete World Series coverage ::
Before we begin, it's important to understand a particular concept. It's pretty easy to simply look at a player's career platoon splits and assume that's the player they are. It's also not really useful, because in order to get enough of a sample size to actually be meaningful, now you're years and years back in a player's career, and that info gets less useful the further you go back. (Crisp is nearly 37 and is a shell of the player he once was, for example.)
• Shop for Indians World Series and AL champs gear
So the way we'll do this is to look at a player's projected platoon splits based on Steamer projections. The point is, we've done the hard work for you and can look at this with the best possible data.
How much more offense should Napoli be expected to provide than Crisp?
Napoli is coming off a good season (.239/.335/.465, 113 wRC+, so 13 points better than average). Of course, he's long had a big platoon split, putting up better numbers against lefties, and the Cubs are starting right-handed hitter Kyle Hendricks tonight, who himself is tougher on righties, so that does take away some of his value. On the other hand, the switch-hitting Crisp has a mild platoon advantage that leaves him somewhat better against righties, where he's been a nearly league-average hitter over his career.

Though Napoli is obviously superior, the platoon split that works against him and for Crisp makes it slightly closer than you'd think. Napoli's performance against righties is projected to be just slightly above average. (That's a .329 weighted On-Base Average against a .323 Major League average against righties, for those inclined to know.) Crisp, meanwhile, would be projected to be easily below average against righties, at .311.
That's basically the difference between what Justin Upton and Andrew McCutchen did compared to the seasons from Maikel Franco and Travis Shaw. It matters, more so than the fact that neither Napoli (.206/.289/.382) nor Crisp (.167/.250/.500) have done much this postseason.
How much defense does Cleveland give up by starting Santana rather than Crisp?
This gets a bit more difficult, because we have no actual experience or data with Santana in left. We'll need to choose a stand-in, and a particularly good choice would seem to be another slugger who primarily plays the infield and attempted to convert to left field: Hanley Ramirez. Over the past five seasons, Ramirez had the fourth-worst Defensive Runs Saved (-19) of any left fielder who played 500 innings, and the seventh-worst percentage of balls in zone turned into outs (.826). It doesn't seem unreasonable to set that as a baseline.
On the other hand, Crisp was once a plus fielder, but as he's aged, he's no longer been much of an asset. Over the past two seasons, 30 left fielders have played at least 500 innings, and while DRS is neutral on him, he finished 30th in percentage of balls in-zone turned into outs (.870). Moreover, Crisp's arm is all but shot, as his Statcast™ arm strength of 78.6 mph on "competitive throws" is one of the weakest in the game.
Ramirez missed a playable ball approximately once every four games. Over a long season, that's awful. But over the course of six innings -- let's assume that if the Indians are winning, defensive changes will happen -- it's understandable to take the risk that no competitive ball heads out to Santana. (Remember, a huge percentage of balls to the outfield are either clear hits or no-doubt outs, and there's been no competitive ball to left by Cubs hitters in the first two games.)
Is the risk of going from a mediocre fielder to an awful fielder worth the upgrade of having Napoli bat several times rather than Crisp? Though it'll look a lot worse if Santana botches a ball in left than it would if Crisp struck out, it seems a defensible risk. Headed into three games at Wrigley, Cleveland might be underdogs. Underdogs have to take risks. All of manager Terry Francona's other ones have worked out so far.