I can’t imagine how aggravating this must be for both top-quality set-up men like Adams and smart bullpen strategists like Padres manager, Bud Black. Managers are expected to reward pitchers like Adams by making them closers, even when that decision isn’t necessarily in the best interests of the team. Carlos Marmol is arguably the best reliever in baseball (he led MLB relievers in WAR in 2010 and is having another excellent season in 2011). Lou Pinella leaned heavily on Marmol in ‘07 and ‘08, bringing him into a wide variety of high-leverage situations, because Marmol’s insane strikeout rate, combined with his durability, made him especially well-suited for that role. But in 2009, pressure mounted from the Chicago press and from Marmol himself (who naturally wanted to bank on being one of baseball’s best relief pitchers), eventually forcing Pinella to make Marmol the closer. The middle innings have been torturous for the Cubs ever since.
My question is this: How hard would it be to come up with a “stupid counting stat” which could do for the set-up man what the Save has done for the closer?
Sure, we have the Hold. Mike Adams was second in baseball in Holds in 2010. But the Hold is seriously flawed, even for a “stupid counting stat.” One can “earn” a hold without recording an out. In certain situations, one could walk six straight batters and still “earn” a Hold. And Holds, like Saves, only apply to situations in which the reliever’s team has a lead. In the case of the Save, that isn’t a huge deal, since managers rarely use closers in any other scenario. But set-up men are frequently employed in close games, whether their team is ahead, behind, or tied.
I offer instead a handful of alternatives which apply to particular game situations during relief pitchers are forced to prove their mettle.
The Rallykiller (RK) – Applies to any situation in which a reliever inherits runners in scoring position and succeeds in stranding them.
The Bridge (BG) – For recording three or more outs without surrendering a run when entering either a tied game or a one-run game.
The Dam (DM) – For recording four or more outs without surrendering a run in a game seperated by three runs or less.
The Dragonslayer (DS) – For retiring the opponent’s #3, #4, or #5 hitter in a situation in which one swing could result in a lead change (or tie).
The Pensaver (PS) – Any relief appearance of three or more innings.
Perhaps, the adoption of such stats would merely exacerbate the problems caused by Saves and Holds. However, I believe, as such stats would highlight other crucial game situations and provide relievers with alternative means of quantifying their contributions, it could very well diminish the idolization of the Save and eventually even make the salary margins for relievers more rationale (i.e. less overpaying of closers, more equal pay for quality set-up men).
It would take considerable time to compile a comprehensive leaderboard for these “inventions,” but I have put one together for the 2011 Yankees bullpen:
In retrospect, I wish I had also tracked opportunities for each category, as that would give us 1.) a clearer vision of how often these situations arise and 2.) a better sense of the efficiency of each reliever, as well as a deeper insight into Joe Girardi’s bullpen strategy.
That, I think, is perhaps the greatest gain to be had from further quantifying and tracking relief pitching. It could put in stark relief the habits of big-league managers and test their effectiveness (or lack thereof).
Even from the limited data I’ve gathered, it becomes clear that Girardi heavily favors David Robertson and Boone Logan in high-leverage situations (middle of an inning, men in scoring position), while Joba Chamberlain and Rafael Soriano usually come into close games at the start of frames. And, although small sample sizes should probably be taken into account, one can see how valuable Luis Ayala and Hector Noesi have been since joining the bullpen corps, pitching effectively in a variety of key situations.
What do you think? Would such stats be useful? What qualifications might improve them? What other situations might be worth tracking?