Second Thoughts Game #134: Cleveland 3, Kansas City 2
On Saturday, Cleveland defeated Kansas City 3-2 in eleven innings. Cleveland struck out 13 Royals batters, allowed 2 runs, and used seven relievers, six of whom combined to make a grand total of ten outs.
Welcome to Terry Francona's Indians in 2014. Enjoy your stay.
The liberal use of the bullpen in 2014 has been a point of extreme intrigue, not that the entirety of the bullpen's usage has been solely at the manager's discretion; in 2014, Indians' relievers have thrown 441.2 innings, second-most in the majors next only to the Rockies, whom they trail by only one out. This factor is as much a result of pitchers like Masterson, Bauer, McAllister, and (occasionally) House failing to give the necessary length. Francona's hook is somewhat quick, yes, but if a starter reaches the end of his rope, it's hard to imagine that the population of managers would vary widely on when to pull that same starter. The bullpen's raw innings eaten is not reflective of Francona's judgment
Entirely reflective of Francona's judgment, however, is the fact that, despite being merely second in total innings, Indians relievers lead the majors in appearances by a vast margin. Cleveland's total reliever appearance total - 481 separate reliever-games after Saturday's game - is 33 appearances greater than second place; given that the team with the fewest relief appearances in the majors, Cincinnati, has 351 relief appearances, the gap between first and second is one fifth the difference between the most liberally-used bullpen arms and the most rarely-used.
As much as Francona has used his bullpen arms, however, he's gotten not only substantial length but also sound performance, as well. FanGraphs WAR, which focuses on fielding-independent outcomes, the Indians bullpen is merely above-average, at 14th in the majors. In terms of the processes that lead to effective run prevention, the Indians bullpen is good but not great.
Unlike the starting pitchers, however, the Indians bullpen has parlayed its good processes into exceptional results. Those processes, measured by the ERA estimator Fielding-Independent Pitching (FIP), regard the run prevention processes of the Indians rotation as virtually the same as the its bullpen; the starting rotation's FIP is 3.52, and the bullpen's FIP is 3.48. For the sake of comparison, the AL average FIP for starting pitchers is 3.90, whereas the AL average FIP for a reliever is 3.68. Because relievers are permitted to pitch in shorter stints than starters, a reliever is expected to be much more effective in the time he does pitch than a starters; hence, in FIP terms, while the bullpen is marginally better than the rotation in absolute FIP, the rotation is far better in terms of run-prevention processes relative to their roles.
Actual run-prevention, however, tells quite the different tale. Because the Indians' rotation has an ERA of 4.08 on the season, it has an ERA minus FIP (ERA-FIP) of 0.55, a half-point gap that is the second largest gap in the league. While previous paragraphs stressed run-prevention processes, the Indians' rotation has been indeed been below-average in terms of actually preventing runs.
The bullpen has quite the opposite problem - their process stats are somewhat middle of the road, but their run prevention is little short of stellar. Whereas their FIP was only slightly better than AL average, their ERA of 2.77 is third-best in the majors. While ERA is a problematic stat to use in isolation for bullpens, since inherited runners are not counted against their ERA, among other reasons, it's a stat whose systematic flaws apply to all teams about equally, and can probably be used to fairly compare bullpens against other bullpens. The bullpen's ERA, even when factoring in league and park factors, is third-best in the game.
In large part, this difference is fueled by an equal and opposite process as the one that has so inflated starters' ERA: Batting Average on Balls in Play. In other words, when batters *do* put bat on ball, the starters have been victimized more often than have the relievers. The league average BABIP is .295 - anything either substantially higher or lower than that for a pitcher is either unsustainable or as a result of an extremely good or poor defense. For the starters, they have a BABIP of .321, third-highest in the majors. Given the often disappointing quality Indians' defense, it should follow quite readily that more balls fall for hits relative to the league average.
The relievers, however, have done their best to undermine the narrative that the Indians defense is to blame for the starters' BABIP struggles: the .277 BABIP that the Cleveland bullpen has enjoyed has been the seventh-lowest among all MLB bullpens. This dissonance between starter and reliever BABIPs explains virtually the entirety of the gap between starter ERA and reliever ERA.
Ultimately, the bullpen run prevention has come. That the source whence the run prevention has sprung is inherently volatile is possibly concerning going forward, but what has happened has happened - regression is not some Lovecraftian horror that reaches into the past and takes away what has occurred. The Indians bullpen has prevented runs at an elite rate, and that cannot be denied. At least, the pen is very solid and slightly better-than-average. Perhaps it will not be an elite bullpen moving forward, but when Corey Kluber is one of your starting pitchers, even an average bullpen makes for an extremely valuable pitching staff.
John can be reached on Twitter at @JHGrimm. He can also be reached by e-mail at email@example.com.
That said, most games don't go more than 9, and most only ten. On balance, it's probably the case that Francona's reliever usage is a positive, but I haven't run the numbers.
That said, Cleveland took a series from the (then) AL Central leaders, have a chance to sweep, and have a winning record against the Tigers. The deficit is appreciable, but reasons for excitement absolutely exist.
At the same time, that criticism is both revisionist and unfair. Francona's strategy has its downsides, which almost bit us on Saturday. But to be fair, if Francona had not aggressively used relievers in the seventh, there's no guarantee the game would have even made it to extra innings. They might have surrendered several runs in the seventh rather than simply one. What we saw was a possible downside of the strategy on Saturday.
So ultimately, I suspect the fact that they rode Tomlin was less a factor of the relievers being deficient as it was that there simply weren't enough of them to sate Francona's need for constant reliever rotation. I expect that will change when rosters expand. A fifteen man pen might be sufficient for Francona's purposes - probably not, but it's possible!