In the New York Times, Dan Rosenheck wrote about the Diamondbacks’ improbable outperformance of their Pythagorean projection. Such musing was also the subject of THT’s recent “Outsmarting Pythagoras.”
One thing I didn’t make clear earlier this week was that not all Pythagorean variance can be brushed aside as non-repeatable “I-won’t-call-it-luck.” Rather, it makes up a greater portion of the variance than people are willing to admit to themselves. Dan did a very nice job of explaining that the D’backs were benefiting from both creative bullpen leveraging and a non-repeatable skill that I call “I-won’t-call-it-luck.” He correctly explained that both are factors in the success of the D’backs.
Dan also writes, “Arizona’s three best relievers (the closer José Valverde and the setup men Brandon Lyon and Tony Peña) have combined for an impressive 1.70 L.I., while its worst (Brandon Medders, Edgar González, and Dustin Nippert) have an 0.47 L.I.” (LI being Leverage Index) Along those lines, Jim McClennan, proprietor of the Arizona Snakepit blog suggested to me (by e-mail) that the Diamondbacks have a core/periphery bullpen that uses an A-bullpen to win close games and a B-bullpen mop up hopeless causes. Valverde, Lyons, and Pena have a combined 2.84 ERA, while the rest of the bullpen has a 4.77 ERA.
There’s a subtle distinction that needs to be made here, between that of performance and that of leveraging performance. Clearly, pure performance isn’t enough, since a good bullpen will allow fewer runs and therefore improve a team’s Pythagorean record. But properly leveraging performance can help if you let your bad pitchers eat up innings that don’t matter. This is the correct version of the oft-repeated but slightly incorrect meme that that a good bullpen will allow you to outperform your Pythagorean record.
Still, leveraging good bullpen performances can’t be everything. The Blue Jays also have core/periphery model in their bullpen (most teams do). Their top four relievers have a 2.30 ERA, the rest of the bullpen 4.14 ERA. Not as great of an A-team/B-team split than the D’backs, but still large. And those top Blue Jay relievers have a combined LI of 1.66. But the Jays are -3 games on their Pythagorean projection!