BP Comment Quick Links
February 25, 2015 Pebble HuntingWhat the Heck, Tigers?
Has there ever been a team that entered an offseason more obviously in need of bullpen improvements than the Tigers? They finished 13th in the AL in bullpen ERA, ahead of only the Astros and White Sox; 14th in WHIP, ahead of only the White Sox; and 28th in all of baseball in FRA, which is park- and league- and era- and luck-adjusted. Notably, they managed that while throwing the fewest innings in the American League, a reflection of their strong starting staff but also a boost to their overall relief numbers—if they’d had to scrounge up an additional hundred innings to match the Angels’ total, most of those extra 100 would have come from even worse options.
And then came October, when the ‘pen turned a one-run deficit into nine, and turned a two-run lead into a loss. In all: 11 runs in 2 2/3 innings, if we except Anibal Sanchez’s anomalous relief appearance. In Tim Dierkes’ pre-offseason predictions, he put Andrew Miller on the Tigers. In R.J. Anderson’s, he shipped Luke Gregerson to Detroit. As it turned out, Randy the Random Number Generator knew better: The Tigers, Randy said, ain’t gonna do squat.
And re-signed Joba Chamberlain for a million bucks, but only because Chamberlain "turned down more lucrative offers." Here’s one: The Tigers had one reliever last year (Chamberlain) produce a FIP lower than 3.25 in at least 15 innings. The Angels had eight.
So, what the heck?
Possible Explanation 1: Last year was just a fluke.
What you can say is that the postseason was pretty much a fluke. The difference between the Tigers bullpen and a good bullpen is maybe a run per series. That the Tigers very loudly, very visibly gave up 11 runs in three games is easy to draw conclusions from, but they’re likely to be the wrong conclusions. That was a freak occurrence and not proof of Joakim Soria’s inevitable failures.
Possible Explanation 2: No, I mean every year is just a fluke.Two years ago, I wrote an article called The Most Fleeting Way To Win, in which I supposed that teams with good bullpens were likely to have average bullpens the next year, and that teams with bad bullpens were likely to have average bullpens the next year. That it’s all an illusion.
A good bullpen is ephemeral, and the correlation from one year to the next is basically insignificant.
Is this the Tigers’ strategy? Simply waiting for the inevitable bullpen regression to relump every baseball team’s bullpen into a sort of new pangea each year, from which teams will drift off toward positive or negative outcomes based on no factor we can anticipate? Should we expect nothing specific about Detroit’s bullpen going forward?
If it is the Tigers’ strategy, it’s probably not a great one. And we probably should expect something specific about Detroit’s bullpen going forward.
That original article looked at team ERAs from year to year, and limited itself to one league (because year-to-year ERA comparisons would be skewed by DH and league difficulties). It also had the convenience of running after 2012, a year that was an outlier for bullpen-quality shuffling. That year’s bullpens showed virtually no correlation, and that’s probably why the article got written. But most years showed some, and the years since have shown some. Let’s take another look, using not ERA but team bullpen FRA.
Last year’s bullpen FRAs showed a .26 correlation with 2013’s, which in turn showed a .26 correlation with 2012’s, which in turned showed a .46 correlation with 2011’s, and so on. If we go back a decade, we see a .31 correlation from year to year, which isn’t overwhelming but is much more compelling than the .18 we found for year-to-year ERAs. (Team success generally—as measured by wins—show a .47 correlation in the same timeframe.)
This is an even bigger difference than it appears, because ERA is heavily influenced by park factors that would promote stronger year-to-year correlations. We can deduce that the year-to-year variance of bullpen performance is strongly influenced by those factors that differentiate FRA from ERA: Strand rates and batted-ball luck. The skill aspects of pitching, however, persist a bit more durably, even in bullpens.
The Tigers’ issues were, largely, skill aspects of pitching. As noted, they were as lowly by FRA as by ERA. They were 14th in the AL in FanGraphs’ model of WAR, 13th in the AL in FIP.
So, yeah. Bullpens are less stable than any other part of a team’s roster. But up in that block quote where I said “that’s not hyperbole”? That was, as it turned out, hyperbole.
Possible Explanation 3: They did add awesome top-flight relievers.
Doing your Christmas shopping early isn’t the same as not doing your Christmas shopping, obviously. The Tigers added Soria, and will add last year’s preseason closer candidate Bruce Rondon, who should return healthy from Tommy John surgery. Neither is so certain as Holland and Davis—or even Robertson and Duke—but they provide reasonable expectations of two innings filled.
Possible Explanation 4: They’re building a more sustainable model for bullpen success.
Despite the relative lack of big names, this is how bullpens are built. Alburquerque and Hardy were both minor league free agent signings and they contributed 2.9 rWAR out of the bullpen last season. Joba Chamberlain was another buy-low signing and he was the team's best reliever for nearly all of 2014. …
I’m not sure 11 is really all that many—you have to draw a line somewhere on who gets counted, and minor-league free agents might be on the wrong side of that line—but, sure, the point stands. After 20 years of this site’s authors criticizing big-dollar reliever contracts, it’s hard to get mad at the guy who doesn’t. Still, it brings to mind Jerry Dipoto, who also has famously eschewed offering closer-type contracts and preferred to gamble on numbers. It didn’t work at first, then last year it did. So where did he find his guys?
Two things to note: It’s not all plankton Dipoto scooped up. Even he was willing to give up C+ prospects in trade or spend $5 million a year on a reliever, whereas the Tigers this offseason did not. And the best performances did come, usually, from the guys who cost something. Not always—Mike Morin was awfully good, and Yoslan Herrera produced a little something—but otherwise every name that produced for the Angels last year was a name you’d heard of already. The Tigers’ strategy, if it is one, involves a lot of pitchers you haven’t.
Possible Explanation 5: They’re broke.
But, as it turns out, Dombrowski, the Tigers tight-lipped president and general manager, might actually have been dead serious when he told local reporters earlier this month that the Tigers bullpen could really be all set.
It always—always—comes down to Possible Explanation 6: A Little Bit Of All Of These, and so it probably is here. The Tigers avoided overreacting to two bad October games and had faith that their bullpen wasn’t as bad as the commemorative DVD of the 2014 postseason will imply. They knew they had already started improving their bullpen as soon as Bruce Rondon began his rehab, and that they had added a relief ace already at the trade deadline. They learned from the Joe Nathan signing and, instead of spending half their bullpen budget on one guy, they spread it among a bunch of different guys with various strengths and risks. And they were probably less aggressive than they’d like to have been because they are already pushing a $170 million payroll, with huge commitments to older players extending for the next half-decade and beyond. They made do with what they had, and the result was… well, it was underwhelming. The Tigers bullpen will probably be bad again, but the Tigers probably won’t be.
Sam Miller is an author of Baseball Prospectus. Follow @SamMillerBB
10 comments have been left for this article.
|
Well, it's a lot easier on the ego to send Shane Greene and/or Alfredo Simon to the bullpen during the playoffs, so maybe that's their plan*
*Assuming their plan doesn't interfere with actually making the playoffs