The flaws in the closer model are starting to show (MLB Bullpens) [View All]

[nickgutierrez]

Background: Since Tony LaRussa's Oakland A's in the 1980's (correct me if I'm wrong on this), at least, teams have used an incredibly predictable and utterly useless scheme of bullpen management where they save their best pitchers for the last inning of games in which they are leading by three runs or less. The purpose of this usage, as far as I can see, is to maximize the total of a particularly useless stat known as the "save", or because the magical and unquestionable Book says so. Either way, it has some problems.

We're starting to see the crack in the closer model, and it comes in the form of one of the most unhittable machines the league has ever seen. For years, Mariano Rivera had the reputation as a machine, an absolute lock with the game on the line for the New York Yankees and the Evil Empire of the late '90s and early 2000s. But now, after seeing him for something like 1,000,000 innings over the past three seasons, the Boston Red Sox have, apparently, started to figure him out.

Rivera has blown four of his last five saves vs. the Yankees' hated rivals. Part of this is that Rivera is a one-pitch pitcher. All he does is throw that cut fastball, in towards a rightie and away from lefties. It works. As I said, the man was a machine from about 1996 to this very day. But when a team of good hitters gets enough looks at a one-pitch pitcher, no matter how good that one pitch might be, they start to hit that pitch.

The other problem, and the one that nobody seems willing to touch, is his usage. If a team has a reliever's number, as the Sox seem to have Rivera's number, you STOP USING THAT PITCHER in key situations against that team. The Yankees have capable relievers. Tom Gordon, Steve Karsay, Paul Quantrill, and Felix Rodriguez are all very good pitchers, even if they have seen better days, and Joe Torre's refusal to use either of them beyond the 8th inning shows a level of myopia that isn't helping anybody.

In the first of the two games, the Yankees had a one run lead going into the 8th inning, and they brought in their designated Eighth Inning Guy, Tom Gordon. Gordon pitches a perfect 8th inning. A series of questions leaps to mind:

1. If this were the 7th inning, would he not be sent out for the 8th without a second thought?

2. Putting aside the closer model for a second, what is really so different about the 9th inning that renders a guy who just pitched a perfect inning incapable of finishing?

3. More generally, is there any real reason to change from a guy whose first inning was perfect to another reliever?

4. In light of these questions, why the hell would you pull Tom Gordon for the 9th inning?

This isn't all about the Sox "owning" Rivera. This is about an inefficient, myopic, and just plain ineffective way of running a bullpen. When a manager is so ready to use lefty-righty strategies in the 6th, 7th, 8th innings, why is it that the presence of a guy that has the shiny label of "closer" suddenly overrules these concerns?

If your team has a robot for a set-up man, as Rivera was in the late '90s before the departure of then-closer John Wetteland, as Guillermo Mota was for the Dodgers the past couple of seasons before being traded to the Florida Marlins, what's the risk in letting that guy go an extra inning now and then?

If your closer is one of the most effective pitchers in the game, as Eric Gagne is now for the Dodgers, why do you use him with a three-run lead in the 9th inning, but not in the 9th inning of a tie game? For that matter, why not the 8th inning with a one-run lead? Or, to take it to the extreme, the 7th or 8th of a tie game? If the man's job is to help your team win close games, why not put him in when the game is at its closest?

When The Book stops working, it becomes time to write a new chapter. Joe Torre needs to be the man to step up and write this new chapter, but he probably won't do it. Who will it be?