How Much Did The Mets Gain By Benching Castillo And Francouer?

As fans, we hem and haw and yammer about every move the team makes. That is our wont.

Sometimes, though, it's worth looking at exactly how much the moves we want the team to make are worth. For example, the Mets finally released Alex Cora and benched Luis Castillo in favor of Ruben Tejada and put Jeff Francoeur in a platoon role. We huzzahed, collectively, and Eric did a good job of illuminating the reasons why. Can we figure out exactly how much the Mets gained with those moves, though?

This is about as mathy as I'll get, so bear with me. First, we'll take rest-of-season wOBAs to see how many batting runs above replacement the team gained on offense. We'll have to regress Frenchy's left-handed split some, and we're relying on projections and some dartboard-style faithcasting when it comes to the rookies, but if we are conservative, we might still be able to quantify the runs the team gained on offense. Defense is much harder, but we'll do our best there, too. The numbers may be underwhelming, but at least they should be positive.

First, we have to figure out how many offensive runs Luis Castillo and Alex Cora would have been worth at second base. Just for simplicity's sake, we'll assume that Cora would have taken one game of every five from a 'healthy' Castillo. Looking backwards won't help in this regard because Castillo has not been healthy. So, 80% of Castillo's offensive run production plus 20% of Cora's.

We'll use wOBA to predict batting runs above replacement. If anyone is interested in the math, here's the link I'm using. Castillo is projected by ZiPs RoS to have a .316 wOBA for the rest of the season, and Cora would supposedly sport a .265 number. (For reference, .323 is average this year, so that's a putrid number.) Together, that would be a... .306 wOBA going forward. Converting that combination to runs, assuming 4 plate appearances and 49 games for the rest of the season, that's -2.9 batting runs. Yup, despite being terrible, that's all they would have cost us the rest of the year - about a third of a win on offense.

Of course, that's not the whole story here. Defense is tough, so let's take a three-year composite for the veterans. FanGraphs has Cora worth -1.6 runs fielding over the past three years on average. His 3-yr UZR at second alone (0.8) is a little kinder, but he's made so few plays, we'd be better off using his career UZR/150 at 2B (3.7) despite some obvious regression as he's gotten older. Let's use his 3-yr UZR since it lies in the middle of the pack. Castillo has stayed in one place and made a lot of plays, so he's easier. His 3-yr UZR is a whopping -14.7. If defensive statistics were better off, I might feel worse about this next part, but they are a little better than a flashlight in a dark room right now. Let's just take 1/3 of 1/3 of each of their 3-yr UZRs and use that as our "fielding runs above replacement." So that would be 20% of 0.1 and 80% of -1.63, or another loss of -1.5 runs. Combined, that's -4.4 runs for the veterans.

The replacement, Ruben Tejada, is much harder to project. He has a .227 wOBA right now and a -1.5 UZR. If he continued to be that poor, he would cost the team a whopping 16.9 runs over the rest of the season. But Tejada's BABIP is .234 right now and even his major league equivalent OPS for his 219 Triple-A at-bats (.567) is higher than his current .497 number. Per FanGraph's Jack Moore, his wOBA would be .271 if he had a .300 BABP right now. So let's do our own ZiPs RoS and give him that .271 wOBA... that gives us -8.9 runs on offense, or -9.4 runs total for the rookie. Of course that's a floor, but we'll sum up in the end.

When it comes to the outfield, it's once again easier to quantify what might have been had Jeff Francouer been allowed to stay the course than what will be in the future. With Bay out, Frenchy's .313 ZiPs RoS wOBA would have been playing every day, and would have cost the team 1.7 runs over a replacement player. Though his fielding has been better this year, it looks like it would have been worth -.8 runs through the rest of the season, putting him as a -2.5 run player - that's 2.5 runs below replacement, or craptacular.

We've talked here about how good Francoeur is against lefties, and his career .348 wOBA against lefties is sooo dreamy (like his hyena smile!). But Tom Tango at The Book taught us that we have to regress that wOBA against 2200 plate appearances for right-handed hitters because small-sample splits from right-handed hitters are poor predictors of future performance. Once again, here's the math if you are interested, but once we regress his past performances to the mean, Frenchy's projected .313 wOBA would actually be a .341 wOBA if he played against only lefties going forward. So far this year, he's played in 106 games and accrued 95 ABs against righties. If he did something similar going forward, he would see about 44 plate appearances and add about .65 batting runs above replacement to the lineup. This is a nice little swing, from -2.5 runs to -.15 runs overall.

The rub is, of course, that Fernando Martinez might just give all those runs back and more. He does have a ZiPs RoS projection of .316, which isn't terrible, and actually better than Frenchy's RoS - sort of surprising. He also had a better platoon split against righties in the minor leagues (.819 vs .689 OPS), but without many major league plate appearances to go by, we'll just have to give him a little tick upwards in the final analysis if he's going to avoid lefties. A .316 wOBA would cost the Mets about .92 runs. He'll get those runs back on defense - he's been a -5 or so center fielder in the minors, but that projects to be a positive defender in right. Let's call the combination a scratch-to-positive combo.

Okay, so let's sum up. At second base, we went from a Luis Castillo / Alex Cora combo that would "provide" -4.4 runs above replacement to the rest of the year and replaced him with a rookie in Ruben Tejada that might be as bad as 9.4 runs below replacement. At right field, we went from losing 2.5 runs above replacement to basically scratch by platooning Jeff Francoeur with Fernando Martinez.

Although exchanging these roles may have cost the team a quarter win (10 runs above replacement = a win), there is one last aspect of the moves that cannot go unmentioned. The two players that will receive the bulk of the playing time going forward are young. They threw wrinkles into our projection process by being so young - and that's a beautiful thing.

The Mets are, essentially, paying a quarter-win down on a bet. In the short-term that bet is that Tejada and Martinez can outperform a gaggle of veterans that would give us predictable - but poor - performances. If the young men don't step to the fore, the team loses a quarter win on their already out-of-the-money pace. If they do, the golden goose of upside may just lay a postseason egg for the Mets.

But the best part is, even if the bet doesn't pay off, the Mets gather more information about two young players going into 2011. As we know from doing this exercise, more information is always helpful, and it's probably worth a quarter win.