Monday, July 21, 2008

The systemic bias in favor of check-swing strikes


While Jon Miller and Joe Morgan have undoubtedly inspired more suffering than intellectual stimulation during their time on Sunday Night Baseball, they made one pretty good point during yesterday's Red Sox-Angels game: the prevailing regime for evaluating check-swings is systematically biased in favor of strike calls.

The home plate umpire makes an initial call. If he says swing: no appeal available. But if he says no swing: appeal available. The result: false negative calls are occasionally caught and overturned (erroneously or not) while false positive calls always remain in place.

Is there any persuasive reason for this? Do efficiency concerns counsel in favor of granting in catchers a right of appeal that batters would use less judiciously or deservedly? Is accuracy instead the operative value--in which case the idea would be that home plate umpires simply render fewer erroneous false positive than false negative decisions (perhaps deliberately so, in light of their knowledge that only a non-swing call, or non-call, can result in a second opinion)? Is there some other explanation?

To be sure, the phenomenon of the one-way-appealable call appears elsewhere in sport. In the NFL, for example, a disputed fumble ruled a fumble on the field is subject to instant replay review, but the same play initially ruled down-by-contact is unreviewable. In that case, however, there's a good resason for the asymmetry: players stop playing when they hear the whistle that accompanies a down-by-contact call, and so any subsequent activity during the play might reasonable be viewed as tainted or less legitimate. At first glance, at least, it's hard to see a similarly compelling justification for baseball's check-swing appellate regime. Am I missing something?

8 comments:

Dork 2 said...

Here's what you're missing: you assume, perhaps reasonably, that the calls originating with the home plate umpire will contain, on average, an equal number of false positives and false negatives. What you neglect to point out, however, is that the home plate umpire need not honor the catcher's challenge: if he's sure it's a non-swing, he is free to (perhaps erroneously) rule it a non-swing without appeal to another umpire, just as he is free to (perhaps erroneously) rule it a swing without such appeal. The only cases in which the catcher's appeal will be honored, and the decision delegated to another umpire, are those where the home plate umpire is truly unsure of the correct call.

Thus, if anything, the bias is actually in the other direction, towards erroneous non-swing calls. Why? Because in order to get a no-call changed to a strike call, the catcher must issue a challenge. Without a challenge (and acceptance of the challenge by the home plate umpire), the other umpires are just as powerless to make the call as they are in any other case where judgment has not been referred to them from home plate.

No-calls will thus remain non-swings except in those cases where the catcher actively intervenes, and then only if the home plate umpire accepts the challenge and the other umpire deems it a swing. False positives and false negatives will cancel each other out due to non-referral, in both cases, to the other umpires.

You should know better than to listen to Joe Morgan.

Dork 1 said...

I agree that the discretion to "certify" a no-swing call for review does change the calculus--but as I see it, the effect is only to mitigate the bias, not to reverse its direction.

In principle, a home plate umpire requests outside help only when he's unsure of the call. In practice, however, "petitions for review" (in the form of angry pointing at the first/third base umpire by the pitcher/catcher) are very rarely denied. While the discretion to grant an appeal nominally resides in the preliminary factfinder, then, in all but the rarest cases, appeal follows as a matter of course from the preferences of the parties themselves.

But not from *either* party; just from one party. The pitcher/catcher gets a second bite at the apple, while the batter is stuck with the initial determination--unless, that is, his adversary gets a second shot.

To be sure, the fact that a home plate umpire may on occasion exercise his prerogative to deny review should eliminate some "false positive" calls that the first/third base umpire would render. But it can't eliminate all of them, and it's not a terribly pervasive practice in any case.

Dork 1 said...

Addendum: Yes, you're right, an umpire's exercising the prerogative to deny review not only prevents false negatives, but also preserves false positives. That's another sense in which the pro-strike bias is mitigated. On the whole, however, you can't seriously believe that this practice of review-denial occurs frequently enough to make a real dent?

Dork 2 said...

Let us grant, for purposes of the present discussion, that, as you put it, "'petitions for review'...are very rarely denied." If that is the problem, then I see a ready solution: make a more common practice of denying such petitions. Your argument rests on an unfortunate (and perhaps willful?) failure to acknowledge the difference between the rule itself and adherence to that rule. Your original claim, following Miller & Morgan 2008, was that problems in the check-swing arena called for an amendment to the rule itself. What you have subsequently helped me to demonstrate is that, instead, all that is needed is a modification of umpirical behavior under the existing rule, which is without flaw as set down by the framers.

To paraphrase the lobbyist's slogan: we don't need new laws, we simply need to enforce the laws that are already on the books. And maybe Coco Crisp should learn to lay off those down-and-away pitches.

Dork 4 said...

I was going to chime in with a complimentary point: that check swings are almost *always* strikes, if called by the book. (It seems that some combination of necessary bat speed, nerve-impulse propagation, and wrist tendon flexibility lead to an arrested swing invariably crossing the front of the plate whether or not it's visible to the non-electronic eye.)

Then I checked the rulebook and discovered that my long remembered definition (courtesy of Tim McCarver, perhaps?) was bunk and that it is officially a judgment call. Oops

Michael Eisen said...

I have two comments here:

1) I'm confused about the home plate umpires alleged right not to honor the appeal. I have probably watched thousands of games in my life and can not recall this ever happening. Perhaps this is because, contrary to previous postings, the home plate umpire is not at liberty to ignore the request for an appeal. The comments on Rule 9.02(c) in the official rulebook seem perfectly clear: "Appeals on a half swing may be made only on the call of ball and when asked to appeal, the home plate umpire MUST refer to a base umpire for his judgment on the half swing."

2) The issue of umpire psychology warrants further discussion. Dork 1 posited that home plate umpires might render fewer false positive than false negative decisions because they are less willing to make a positive call because there is no chance for review. But presumably most (if not all) umpires do not like to be overruled (and even if some don't mind being overruled, few actually like it). This would tend to result in more false positive calls relative to false negatives because umpires will tend to move as many ambiguous calls as possible into the strike category.

Dork 1 said...

Three more thoughts, in response to the new data that have recently come to light:

(1) Dork 2: Not so bright.

(2) To be sure, the original suggestion that home plate umpires might err on the side of no-swing calls in order to preserve the availability of review was charitable, at best. The upshot of the last commenter's more plausible assessment: this crisis is even worse than we'd imagined.

Here's a mystery, though: given that hitters outnumber pitchers in the players' union by something like a 14/11 ratio, how has such an offensive-unfriendly rule (and one with virtually no basis in reason or necessity, to boot) persisted so long? Why have the Coco Crisps of the world not organized to lobby their representatives? Is it that union reps are disproportionately pitchers, because they have less to do during the week? WE WILL GET TO THE BOTTOM OF THIS.

(3) Dork 4: what do you mean "it is officially a judgment call"? Judgment by what criteria? That is to say, when a 1st/3rd base umpire overrules the home plate umpire, the appellate adjudicator has to have some reason more grounded in objective factors than "my judgment differs from the home plate umpire's." Right? Or is it actually the case that it's an entirely standardless, discretionary decision? And if so, doesn't that violate the Due Process Clause or something?

Dork 2 said...

A few thoughts:

(1) What kind of a name is "michael"? That doesn't even come close to having "dork" in it. I'm not sure this person can be taken seriously.

(2) You will have noticed that the above rule-book quotation states that the judgment of "a base umpire" must be sought in the case of an appeal. It neither grants the catcher the right to designate a particular base umpire nor binds the home plate umpire in his choice of base umpire for purposes of making the call. If my fellow posters' speculations about the fragility of umpire psychology had any basis in fact, we would expect to see a far greater number of appeals to the second base umpire in cases where the catcher challenges an original call of no swing. What better way to keep from being overruled?

(3) Dork 1 wears a brooch.