The Closer Role
Over the last ten years, baseball managers' rigid usage of the closer role has been frequently criticized in sabermetric blogs. The criticism boils down to this: strictly using closers only in the 9th inning sometimes prevents the team's best relief pitcher from being used in the highest leverage situation.
After the local beat writer expressed disdain for the "moneyball crowd's" criticism of closer usage, Charlie Wilmoth at Buc's Dugout launched a nice defense of the saber position, and it's worth a read. The article traces the modern closer usage to the invention of the save statistic in 1960. I think he correctly identifies some obstacles to a change in managers' bullpen usage: closers' fear that changes will reduce their salary potential and managers' reluctance to be second guessed if they decide to make a bold departure from normal closer usage. He speculates it may be a generation before managers' adherence to the closer role changes.
Teams that are more likely to change the traditional closer role are likely to be either desperate or bad teams with little to lose. These are not the best proving gorund for changing the closer position, since both circumstances are predisposed to failure. However, recent failures by closers have led both the Brewers and Giants to contemplate "closer by committee" role changes. Chron.com's Zachary Levine labels Wilton Lopez as the " pseudo-closer" for the Astros because Brad Mills says that Lopez will be used mostly in the 9th inning, but may be used on occasion in earlier innings. However, this makes me wonder if the Astros could become an agent for changing the closer role in the future.
Although we frequently juxtapose sabemetrician and baseball traditionalist as opposites, this topic is an occasion when the saber position is closer to the traditional usage of "ace relievers" instead of modern closer usage. In an earlier era, closers could be used less rigidly, allowing the manager greater flexibility in using his bullpen. Closers, who were capable of doing so, were used to pitch multiple innings in a game. Mike Marshall won a Cy Young Award as a relief pitcher, throwing as much as 200 innings out of the bullpen. Arrangement where the two best relief pitchers shared the 9th inning were not uncommon. I particularly liked the lefty/righty combo closers. The decision on who pitches the 8th and who has the 9th can be decided based on the batting orders expected in those innings. Joe Sambito and Dave Smith represent the L/R relief pitcher duo used by the 1980 Astros team. Wilmoth mentions the Orosco/McDowell duo of the 1986 Mets and the Myers/Dibble comination for the early 90's Reds---again, left and right platoons at closer.
I realize that it's not always easy to determine a priori which inning will be the highest leverage inning. But sometimes it becomes obvious that you won't have a save opportunity if you can't get out of an excruciatingly tough situation in the 7th or 8th inning. In 2004, Phil Garner's had one excellent reliever, Brad Lidge, and frequently poor results from the rest of his bullpen, leading him to depart from the normal closer rules. Lidge pitched 2 innings in a game on 7 occasions during the final three months of the season. If a strike out was needed in the 8th inning, LIdge was as likely to produce a strike out as anyone in baseball during that season. Lidge ended up with the 7th highest average leverage index among relief pitchers in 2004.This leads me to wonder if the Astros may be in a position to challenge the normal closer usage patterns next year. Some factors to consider: (1) the Astros traded their closer this year, and have no established closers to complain about losing the role; (2) the front office is saber knowledgeable; (3) as Wilmoth suggested, bad teams are in a better position to experiment with the closer role. The Astros are likely starting from scratch in building a bullpen next season. And, if the Astros sign one or more good relief pitchers, the most effective way of using them is to target their usage to high leverage situations, even if it's outside the 9th inning. And keep in mind that teams with losing records are likely to have fewer pure save situations.
A counter-factor is that accumulating saves may build up the value of a pitcher who will be traded. In my view, this was part of the front office's thinking in moving Brett Myers to the closer role. It's also possible that it's easier to attract a free agent relief pitcher if the closer role is promised. But aside from these issues which are unrelated to on-field winning, the Astros may be a promising proving ground for changing the normal closer usage patterns.
Managers' Use / Misuse of "Hot Hand" or Slumping Hitters
This sabermetric topic can be introduced by a quote from Jeff Luhnow, himself. In a recent New York Times interview, Luhnow indicates that he would like managers to be weaned from decisions based on hitters' recent hot or cold performance:
For one thing, Luhnow said, the notion that a player is hot or cold is largely a myth, and managers should recognize it.
"If you end up changing your strategy based on hot or cold tendencies, more often than not, you’re chasing your tail and you’re actually destroying value rather than sticking to what you know is right based off the data over a longer period of time," Luhnow said. "It’s easy to say from up here, and it’s easy to prove from up here. But it’s a lot more difficult to implement down there."
This says a lot about how Luhnow believes sample size should inform the managers' judgement. Fans frequently criticize lineup or roster decisions based on recent small sample size performance, but this isn't a view that is likely to be shared by the Astros front office. And, more specifically, Luhnow wants to avoid managerial decisions grounded in whether a player is perceived as hot or cold at the moment.
Let's describe the position that is supported by sabermetric research. Obviously, hot streaks and slumps are a fact of life for hitters. Many factors could be causes, including random chance, ballparks, opposing pitchers, mechanical changes, fatigue, injury, etc. But from an on field decision perspective, the question is whether you can predict what the batter will do in the future based on a current hot streak or slump. And the answer is largely "no," with most studies indicating that recent current batting performance is unlikely to predict whether the batter will be better or worse than his normal performance in subsequent at bats. There is some evidence that certain hitters may be more streaky than others, but even this conclusion is difficult to prove over a multi year period.
Here is an academic paper on the subject by a mathematics professor, if you like reading about statistics at that level. This Baseball Prospectus article and study by Russell Carleton is more accessible and useful. David Cameron writes here about The Book's findings on hot hand hitters and how it should affect specific managerial decisions of the Mariners.
The B-Pro article describes a specific managerial decision:
Let’s go to a real game situation. Bottom of the ninth, tie game, runner at third with no outs. A manager will sometimes call for an intentional walk here because the walk means very little in the grand scheme of things. Indeed, the only runner that matters is on third. (Some managers will issue two free passes to set up the force at any base. Let’s discount that for the moment.)
The decision of whether or not to issue the walk comes down to a choice between which of two hitters he’d rather face: the guy at the plate, or the guy on deck. Suppose that the choice is between a hitter who is clearly inferior overall, but has been hot lately, or a better hitter who has been cold. I’d argue that every time, the manager should pitch to the inferior hitter. There’s no evidence that Mr. Hot Hand is any more dangerous than he’s ever been, and if he’s the inferior hitter, that’s who you want at bat. But managers will sometimes betray how much they believe in the hot hand theory by walking the streaky hitter. This is a behavior that has to stop. Not that I’m holding my breath.
I recall similar issues arising when Cecil Cooper managed the Astros (like a 2009 Cubs' game when he ordered a walk of Theriot in order to face Soriano in the bottom of the 9th, who then produced a walk off hit). In game threads this year, I have seen complaints that the Astros shouldn't pitch to a hitter because he has already gone 2 for 2 in the game---but that isn't necessarily the sabermetric answer.
The larger question is how Luhnow's view on issues such as use of hot and cold hitters might affect decisions on the next manager. Luhnow has stated that he will evaluate Brad Mills at the end of the season to determine whether he will be retained or replaced. Luhnow points out, in the Times article, that Mills has been open to data based methods. But if he decides to look for a new manager, this might be the type of question that comes up in iinterviews.