Statistics don't lie, but they don't always tell the whole truth.
If you look at a ballgame through the prism of low-, medium- and high-leverage situations, stats make a compelling case for using a Mariano Rivera before the ninth inning to put out a fire with the game on the line, or using a bigger chunk of the bullpen to close by committee. But that doesn't necessarily mean it's a good idea.
It's the human factor.
"Stats are a derivative that work, and they show a pattern," Padres closer Huston Street said last week. "But you can't quantify what it feels like down there when the phone rings every night and there are six guys who have no idea what's about to happen. Look at the military -- they don't go general by committee. There is a rank and a file. More than anything, [the role of the closer] provides a routine."
Yankees reliever Shawn Kelley was a member of a Mariners bullpen that in early 2012 used situational stoppers in the ninth inning.
"I think a lot of times when it's bullpen by committee, it's that you don't have one guy that you fully trust," he said Wednesday. "It's not so much that you have seven great arms, it's more that you don't have a closer. I think any time you have somebody step forward and solidify the ninth inning, and then other roles can be determined off of that, it's better. A chaotic bullpen, not having set roles, is hectic."
The Tigers were the latest team to learn that lesson, opening the season with an any-arm-can-do-it approach that lasted less than two weeks before manager Jim Leyland named Joaquin Benoit the primary closer.
Yet stats show that the average team can add about 1.6 wins just by using its best reliever in the highest-leverage situations. By examining real play-by-play data to assess expected wins, Baseball Prospectus ran a computer simulation in 2006 that concluded more than 75 percent of teams would win more games by changing how they use their closer.
Closer by committee often is misunderstood -- and incorrectly applied -- as using a different closer every night. The theory actually seeks to identify a bullpen ace and use him during the most crucial point in a game.
The closer role was born in 1988, when manager Tony La Russa turned to future Hall of Famer Dennis Eckersley in the ninth inning of any game in which the Athletics had a lead that needed protection. Before that, relief specialists such as Hoyt Wilhelm and Rollie Fingers finished many games but also were used in other roles.
There were 42,440 save opportunities resulting in a save or blown save from 1916 (when baseball-reference .com's game logs begin) through 1988, the year before the A's went to Eckersley. Even without teams employing a single ninth-inning stopper, a save occurred 73.4 percent of the time. From 1988-2012, there were 42,669 save opportunities ending in a save or blown save. The save success rate -- in the era of the closer -- was only 67.8 percent.
So wouldn't teams be better off using their best reliever in the biggest spots?
"I think there's something to that," Kelley said. "It's not always the ninth inning where the game's won or lost . . . It's more important to let a guy prepare for his role and for his situation than to say, 'We're going to use this guy because he's our hottest hand against their hottest hitter.' "
The Yankees and Rivera have refined the closer role to a point of near-perfection.
Since Rivera became the Yankees' closer in 1997, when "Enter Sandman" is played over the loudspeakers, the Yankees and their fans are conditioned to believe the victory is all but assured. That's because Rivera always knows what he is required to do and when it will be required.
"It's tough when you bring a closer into a non-save situation," Tigers catcher Victor Martinez said during the 2011 playoffs. "As a closer, you're used to coming into the game with the game on the line -- the adrenaline when you come in the game takes over."
Rivera and Street know their role and know how and when to prepare for it.
"There is something to be said about the eighth and the ninth inning," Street said. "We don't have five more at-bats to pick up the slack. There is a weight there, and that weight carries some level of burden."
As helpful as stats can be, there's no scale that can measure that.