Against Instant Replay

I always have been opposed to instant replay. This, of course, puts me in a distinct minority (maybe even minority of one) among sports fans and commentators. And it certainly puts me against the flow of history, as replay is becoming a part, and an increasingly greater part, of most sports. The NFL has raised replay use to an art form and college football is following suit. The NBA recently announced increased use of replay to review fights and flagrant fouls. And it probably will be only a few years before it finds its way into Major League Baseball.

But I stand by my opposition to replay for a number of reasons. I do not like taking the human element out of the game--if we accept inevitable mistakes by players trying their best, why not also accept inevitable mistakes by officials trying their best? I do not like the way it breaks the flow of the game for the ref to spend 10 minutes staring at a TV. How many times do we now have to wait to celebrate a touchdown while officials try to re-figure whether it was, indeed, a touchdown. I do not like the way challenges, at least in football, can be used simply for strategic purposes, to get the equivalent of a long time out and a chance to rest players late in the game. I do not like the effect it has on game officials--at least in the NFL, many refs seem hesitant to make firm calls one way or another, knowing that they basically can guess, then rely on replay to have the final word. I fear the same thing will happen in basketball on, for example, buzzer-beating shots. I do not like the inherent inconsistency in determining which plays or calls should be subject to review and which should not.

Replay supporters (which is to say just about everyone) insist that any of these drawbacks are outweighed by the increased accuracy that replay provides. All the drawbacks are worth it because, by going to the videotape, we get the calls "right" because the video shows the "right" answer.

But do we? And does it? My current writing project is an essay on the Supreme Court's 2007 decision in Scott v. Harris, where the Court found that a video of a high-speed chase was such an omniscient, unquestionably accurate, and entirely objective (i.e., not dependent on any inferences, perception, or interpretation) depiction of events that the Court could ignore eyewitness testimony that differed from the video, because such testimony obviously was wrong, without the need to submit the case to a jury for factfinding.

In thinking through the question, I drifted to instant replay. Replay rests on similar assumptions about the objectivity and accuracy of what Professor Jessica Silbey of Suffolk Law School calls "filmic evidence." The video replay will reveal the "truth" to the referee, showing, plainly and objectively, what happened on the play--he controlled the pass or not, his knee was down or not. No interpretation or inference is required, no perspective or perception comes into play. The ref can play a clip multiple times, slow it down, even look at it frame by frame; in doing so, the ref will be able to see the unvarnished "truth" about the play.

The problem is that assumption probably is wrong. A video depiction of an event neither is the event itself nor speaks for itself about the event. It still requires an act of mental and sensory interpretation from the viewer (here, the replay official). Two people could view the same video and reach different conclusions in many cases, just as two people could view the same play live and reach different conclusions. Being able to slow down and speed up the video provides a different perspective, although it does not necessarily mean that this different perspective is "right." It means only that it provides a different narrative, and thus a different interpretation, of the events depicted. Silbey argues this is what happened during the state prosecution of the LAPD officers who beat up Rodney King in 1991. Defense attorneys slowed the video to frame-by-frame inspection in presenting it to the jury, which allowed for a very different narrative of the events depicted as to what was taking place on the video, who was the aggressor in the altercation, and what anyone's intent was. This was one key to the officers' acquittal. But that tells us nothing about which narrative is objectively correct.

It is worth noting that NBA refs will be doing something similar in using video to review flagrant calls--slowing the tape down and rewatching to figure out exactly where Player A hit Player B and when and what A's intent may have been. But the repeated viewing simply allows the reviewing ref to construct a narrative based on his perception of the tape.

This is not to say that a video never is accurate or that repeated viewing at different speeds will not provide a more accurate conclusion than a one-time look at the play in real time. It is to say that any increased accuracy does not derive from the fact that we are using video and video inevitably reveals the most accurate answer. It derives instead from the same reason that having an appeals process increases (or is thought to increase) accuracy in the judicial system. Having additional people (new "eyes") consider an issue, especially by spending more time with it (think about the time appellate courts have to analyze and decide issues compared with how long trial courts have), increases the likelihood that we will reach the best or most correct result.

But if replay functionally is an appeal, then the possibility of increased accuracy may not alone be sufficient to justify it. Within the judicial process, for example, accuracy is not the sole value; other policy concerns often outweigh accuracy in dictating the rules for appellate practice. This explains the "Final Judgment Rule" in federal court, under which only final decisions by a trial court are subject to immediate appeal, with certain limited exceptions. Go here for a good discussion of the ins-and-outs of appellate jurisdiction. Whatever increase in accuracy may come with appellate review, systemic preferences against piecemeal review of every discrete decision that the trial court makes and in favor of cases moving forward at the trial level expeditiously and without delay lead us to create a system in which most decisions are not subject to appellate review, at least not immediately. Accuracy becomes secondary to other values, at least when the accuracy benefits are relatively slight.

Similarly, whatever increase in accuracy comes from having another set of eyes looking at a play may be outweighed by the harms such review imposes in terms of delay, loss of flow in the game, risk of challenges to increasing numbers and types of plays, quality of on-field officiating, etc. That ultimately is a judgment call that turns on how one weighs competing values. I weigh them in favor of keeping the game fresh, moving, and human; others may disagree.

My point is that we should have the debate about replay without the assumption that the availability of video review necessarily will increase accuracy simply because it is a video and not the human interpretation of live events that lead to the original call.

Posted in: