Five Thoughts on a Non-Simmering Closer Controversy

Whatever else was made of the Royals’ off-season makeover, most agreed that the bullpen was a strength. Tim Collins and Aaron Crow had shown to be good relievers who might even be worthy of closing for some teams, but were only the the third and fourth options in the Royals bullpen. Kelvin Herrera came up and dominated in 2012 with a fastball that touched triple digits, and projected closer Greg Holland had been one of the best relievers in the American League over the two previous seasons. There were serious questions about the reconstructed rotation aside from James Shields and an offense that would no longer have Wil Myers projected to come up and save them from The Right Field Horror. However, the bullpen seemed safe.

The first week of the season, particularly the weekend series in Philadelphia, has gone a very different direction. The rotation has not dominated, but other than Wade Davis‘ start on Friday night, the starters has gone at least six innings in every game so far. The offense was dormant in the opening series, but exploded against the Phillies for 25 runs in three games. However, the bullpen — more specifically, Holland — was shaky. Holland blew a save on Saturday in one of the most frustrating ways possible, by walking three batters in a row, striking out two more to almost get himself out of a jam, then finally losing the game on a hit by Kevin Frandsen. Holland looked bad again on Sunday. He came into the game in the ninth after J.C. Gutierrez managed to turn a 9-4 semi-laugher into a 9-7 semi-nail-biter. Holland gave up two hits before Ned Yost had seen enough and brought in Herrera, who himself just eked out of the game, 9-8, after giving up a run.

Naturally, fan insta-reaction to Holland’s problems has been, shall we say, less than measured. There has not been obvious buzz coming from the team about a closer switch, but many fans would welcome it. One way or the other, the over-reactions will likely be forgotten by mid-season. We are not to mid-season yet, though, and the situation provides a nice opportunity for a few reflections on this sort of situation more generally.

1. Time for a classic: small sample size. Yeah, it is trite, and everyone (supposedly) understands it now. But it has to go in here, both because it is true and as a qualification for what follows. Anyway, it is worth remembering that Holland got off to a bad start last year, too, having not just reallly bad BABIP luck, but control issues, too. He managed to finish this season with a 2.96 ERA despite a .346 BABIP. He did go on the disabled list after his bad stretch at the beginning of 2012. Still, the Royals did not overreact. After Jonathan Broxton was traded, they made Holland their closer, a role in which he pitched well.

2. It could be the case that there is something wrong with Holland’s mechanics or something like that. Maybe he is hurt or on his way to injury, and that is what is giving him problems. I have no idea. If it is an (impending?) injury, then this is obviously an issue for medical personnel, and sort of makes the “who should close?” issue irrelevant. If he is having mechanical issues, then it might make sense to put him in a lower-leverage situations until it gets worked out. This is not something I am qualified to analyze, but it is worth at least mentioning.

3. Even granting the sample size issue, and even if Holland if healthy and his mechanics or whatever are fine… maybe Herrera should be closer (read: relief ace), anyway. I still believe that Holland is an excellent reliever. He is something of a classic reliever-only type. He has a fastball that hits 95 and and a good slider. His strikeout rate (30.7 percent career) speaks for itself, and though his issues with walks were already evident last season, he his whiffs were enough to allow him to work around them. One would think that a fastball-slider pitcher like Holland would have platoon issues, and maybe as his sample gets larger, they will become be evident, so far he has not. In fact, Holland’s career wOBA against versus righties is .296, versus lefties it is .261. He does walk left-handed hitters a bit more often than he does right-handed hitters, but his xFIP against both is almost exactly the same (3.03 versus lefties, 3.06 versus righties).

As good as Holland is, Herrera might be better. Herrera does not strike out out as many batters as Holland, but his control is far superior. Herrera has also had shown a ground ball tendency. Although Kauffman Stadium might exaggerate velocity, Herrera’s fastball is incredible even if it “only” averages 97. His change-up is inspiring to the point of being creepy. While Herrera may not have the basically neutral split Holland has shown so far, his repertoire is such that it seems unlikely to be much of a problem. Moreover, while the fastball and change might be more than enough to make him an outstanding reliever, I have talked to at least one person who thinks his curve might be good enough for him to merit consideration as a starter. (That issue is a whole different can of worms that I will leave unopened. Or maybe just slightly opened.)

I am not trying to settle the issue of whether or not Herrera or Holland is the better pitcher. The point here is that the Royals might very well have good reasons at hand for switching to Herrera as their closer/relief ace, and it could have nothing to do with how Holland and Herrera pitched last week week. Based on their prior performances, one could make at least a reasonable argument that Herrera is, and thus a better choice for high-leverage situations (however those are identified).

To be fair, it is hard to imagine any team making that decision in the Royals’ situation. Making the switch now rather than during the off-season would just lead people to think that they would be doing it because of Holland’s bad showings. Perhaps there is not that much of a difference between the impact of pitching the eighth or ninth inning in those games in which both pitchers would usually pitch. I am simply using sample size to go the other direction: one might be able to make a case (although I have not made it here, at least not in a way that would convince me) that Herrera should be the top dog in the bullpen based on prior performance and repertoire.

4. Maybe the real lesson to be drawn from the comparison of the two pitchers’ skills is that there does not need to be a designated closer/relief ace at all. Platoon issues were mentioned in the previous point as a potential reason to pick either Holland or Herrera as top dog. Assume for the sake of argument that that Holland really is better against lefties than Herrera, while Herrera is better against righties. The obvious usage in close games, then, would be to try and use the lefty-killer in the inning (we are asssuming one-inning usage here, which is not ideal but simplifies things for the time being) in which he is more slated to face more lefties, and the other pitcher in the other.

One need not think of think of this in terms of traditional platoon issues. Between scouting reports, PITCHf/x, and all that stuff, teams have a pretty good idea of what hitters do and do not do well against. This could be worked out in various ways, but the most obvious would be that if the likely hitters as a group have trouble with sliders, send in Holland. If they have have more trouble with change ups, send in Herrera. One could imagine a number of more sophisticated ways of doing this, but I would think most of the useful ones could be simply coded on a chart that would fit on a sheet of paper for the manager to use when the situation called for a decision.

Of course, teams almost never use stud relievers in this way. They do it in the middle innings with middle relievers. Sometimes, teams with particularly mediocre relievers or no “proven closer” do this a bit in the ninth. But Holland and Herrera are not “specialists,” so they have to have set roles according to inning. There may indeed be psychological reasons why set roles work better in bullpens, I do not know, but I would also hazard a guess that it is a self-fulfilling prophecy: if you tell a reliever that his set role is pitching the eighth, maybe he comes to believe it. Maybe it would help a player to think of his role as being “making guys who can’t hit a slider look foolish” rather than “pitching the ninth when my team is leading by no more than three runs.”

Psychological speculation aside, I am not here to issue yet another call for rethinking reliever usage. Well, okay, I sort of just did. Still, doesn’t it make more sense to think about using a pair of excellent relievers in ways that maximize their match ups rather than getting locked into managing to the save statistic or static roles? To me, it makes sense not just tactically, but also has the added benefit of sidestepping public “closer controversies” from the start.

5. Whatever happens, this post is going to seem really silly in a few weeks.