Tuesday, October 30, 2012

Ned Colletti and Brandon League ($22.5million): A Good Relief Pitcher Strategy

MLB 2K9 was a pretty good baseball simulator for its time. The game balanced the players well, had decent playing mechanics (aside from a propensity for outfielders to run into each other), and had a mediocre at worst General Managing experience. I spent perhaps a few too many hours constructing teams, simulating seasons and playing games, and because of the time I invested in the game I noticed one significant glitch: the designated closer on my team would rack up the requisite 30-40 saves, appear in 50-60 games, and usually post a decent ERA, but he would have only thrown about 25 innings!

It didn’t take me long to figure out that during the games simulated by the computer, the “designated closer” would almost always be brought into the game in a save situation to face only the last batter of the ninth inning. The closer would retire this batter, rack up a save, and look incredibly valuable to a surface analysis.

Though I had yet to truly become immersed in Sabermetric analysis, and would not have applied it to my video game teams even if I had, I noticed the potential to cheat the system. So I traded away the closer, installed a mediocre relief pitcher in the “designated closer” spot, and kept my best pitchers in the set-up and middle relief spots. The game, unable to perceive my genius any better than it could fix its flaws, doggedly kept trying to switch them back. My plan was brilliant, my best relievers threw lots of innings, my mediocre “closer” racked up saves and actually put up decent peripheral numbers because he was not actually facing the best hitters in the toughest spots, and my team won big.

I bring up my crowning moment in computerized baseball not because I think video games can apply directly to real baseball, I don’t, unlike fantasy baseball, in which success is directly predictive of one’s ability to be a major league GM (third place in my 8-team league!!!!). I mention it because the real world usage of major league closers is only slightly less absurd than that of the pixilated closers in my glitchy game.

I was certainly not the first person to notice how irrational closer usage is, entire pages of Moneyball and reams of articles and comments on Fangraphs are devoted to the topic. In terms of the most crucial article for interested fans, Joe Posnanski’s brilliant analysis and commentary on closers and set-up men in 2010 should still serve as the final word on the subject. In fact, Posnanski’s most easily implemented solution to the problem is one that my MLB 2K9 teams would readily recognize:

I think I would do this too -- put my best reliever as a setup man. I mean, yes, I would still love to see someone tear the whole thing down and try and create (a) bullpen without specific roles. But I don't think that will happen anytime soon, and I don't know -- human nature being what it is -- that it would work. I think there's a chance it would not work… So, assuming that we're not yet in a place where you can go with a no-roles bullpen, I think I would make my setup man my star. Sure, you would want a good pitcher as a closer. But I think that's enough. Put someone good in that role and you will win 95-to-100% of the games you lead going into the ninth inning.

This brings us to the most recent baseball news: the Dodgers’ signing of Brandon League, who bears the “Established Closer” tag, to a three year, $22.5 million contract to be their long term closer. This is a man who racked up 37 saves for a bad Mariners team in 2011, and stabilized the Dodgers bullpen with 27.1 solid innings at the end of last year. This is also a man who was at no point prior to 2011 considered a relief ace, and whose only repeatable skill over the last few years has been his ability to generate an above average number of groundballs.

It is all too easy to slip into the sabermetric cliché of bashing Ned Colletti for his propensity to throw big money to mediocre players, and his blatant ignorance of modern analytical techniques. The fact that this move was made by the Dodgers, who boast one of the best young relievers in baseball in Kenley Jansen, who had already proven himself capable in the ninth inning, only made the signing easier to lambast.

But perhaps this signing is not such a bad deal for the Dodgers after all. Though real world baseball is not quite as simple as MLB 2K9, the Dodgers, whether intentionally or not, seem to be doing almost exactly what I “pioneered” in my video game. Whether intentionally or not, the Dodgers have adopted a smart and forward thinking relief pitcher strategy: they are saving their best pitcher, Jansen, for the most important moments and leaving the ninth to a decent pitcher who can probably avoid giving up two or three runs each time out. This is exactly the strategy that Posnanski champions at the end of his article:

If you put in someone good -- your second or third best reliever -- into the closer role, then you will have your best pitcher to use in key situations. You will have him to secure the eighth inning, of course, but you could also use him at other crucial times.

Of course, after the fashion of the new Dodgers’ ownership, they spent way too much money to do it. They have also given no indication that they spent this absurd sum of money for any other reason than the standard dispiriting reason that League had racked up saves in the past. Colletti stated in the press conference announcing the signing: "We think, after what he did, closer is the role," and then declared that, “at the end of the season he was as good as anybody in baseball.”

So, the process wasn’t great, and the contract is certainly extravagant, but the Dodgers will start the year with their best relief pitcher available any inning or situation, and with a mediocre flamethrower pitching the ninth. For a non-sabermetrically inclined franchise, this is not a development to be discounted.


  1. Some good points, but then there is the excellent shut down work of Sergio Romo, this year's closer for the Giants.

  2. Romo didn't even become the de-facto closer until the playoffs, really. Sure he closed some games throughout the season, picking up more towards the end of the second half but he was primarily the high leverage setup man that the author is talking about in this very article.