Getting Replay Right The First Time

“Replay is coming! Replay is coming!” You can almost hear Ben Revere running through the streets of Philadelphia sounding the call, if I remember my American history correctly. I imagine that 98% of FanGraphs readers read the news last week and reacted with only glee, because decades after some of their major sports competitors started using instant replay, MLB is finally getting with the times. Anything that can stop preventable mistakes from happening is a good thing, as I think you’d agree.

I probably don’t need to regale you with dozens of examples of terribly blown calls, from Phil Cuzzi blowing up the 2009 ALCS to Jim Joyce ruining Armando Galarraga’s perfect game to Don Denkinger’s infamous error in the 1985 World Series, so I won’t. You know the game needs replay, and so do I. Still, for the one straggler among you who might just be on the fence about it, I absolutely can’t pass up the opportunity to reuse this still of arguably the worst call in the history of sports, Tim Welke calling Jerry Hairston out on this play in Colorado in 2012:

That’s a real thing that happened, and somehow replay wasn’t instituted by the end of the night.

So if we’re agreed that Bud Selig finally letting go of years of protests to allow replay into the game is a good thing, what might not be so sure is the way they’re going about it, at least by the admittedly yet-to-be-finalized details that have been reported so far.  While I’ve assumed that most of the FG readers are on board with replay, many “casual fans” and some media aren’t, with defensible worries about the length of games, and so this has to be done right. If you’ve been paying attention to the national news at all over the last few weeks, you’ll have noticed that a certain highly sensitive issue has been plagued by awful implementation, and the opportunity to make a good first impression in that arena has been lost, no matter whether you backed the idea or not. Baseball needs to be aware of not doing the same.

Most of what we’ve heard makes sense, at least. Ball and strikes are not going to be included as reviewable plays, nor are check swings, and that’s a good thing. The technology just isn’t there for it yet, really, not when you account for all the variations in park & camera angles, differences in physical batter sizes and stances, and the simple matter of the actual strike zone not really reflecting what’s on the rule books. Check swings are a related issue, because the rule is just so subjective. More replay won’t fix that. (Though I haven’t seen it mentioned yet, please, please, let’s keep balk calls out of reviews, because there’s just never been an acceptable standard in calling those.)

It sounds like just about everything else will be reviewable, however. The obvious fair/foul and safe/out calls will be a boon, and so will eliminating the trap/catch mistakes that often look so terrible on television.  (The counter-argument here is that for some plays, replay won’t provide a clear answer, with a good example being whether or not Mark Ellis was actually tagged out by Yadier Molina in Game 1 of the NLCS. I imagine the answer there will be an NFL-style “irrefutable evidence” to overturn the call, and that’s sufficient.)

Replays will be handled at MLBAM headquarters in New York rather than by the game ump, similar to the NHL’s system of sending all replays to Toronto, and it makes a ton of sense. As much as this move is about taking the “human element” out of play, it’s impossible to ignore the fact that — consciously or not — the umpire on the field may not want to make himself or his crew look bad by overturning a call. This takes the onus off the men who have to go face 50,000 angry fans and puts it into the hands of a theoretically impartial observer.

There remains some logistical questions about exactly to set the baserunners on certain calls, like the trap/catch issue, but that can be thought through. Initial reviews of the system’s test run in the Arizona Fall League earlier this month came out sounding very positive, even in an environment that almost certainly doesn’t have as many cameras as a typical MLB game. So far, so good. Right?

Except I’ve purposefully buried the main component here, because the more I think about it, the more it seems like a mistake. The new replay system will including a managerial “challenge system,” much like the NFL, and I think this is going to be a point of contention for a lot of people. If you’ve ever thrown things at your television while watching Don Mattingly try to think his way through a sacrifice bunt, or wondered what in the world Dusty Baker is thinking when he lets Logan Ondrusek lose a game while Aroldis Chapman relaxes in the bullpen, you know that putting the burden on the managers seems like a situation ripe for disaster.

Including replay really ought to be about right and wrong. Either an outcome happened, or it did not. It shouldn’t be “an outcome happened, but only if the manager asks about it, and then only if he still has challenges remaining,” since the current reports indicate a system of two incorrect challenges. (There remains the possibility that an umpire could independently call for a challenge in that case, though that is not yet certain.)

Adding to that mix is the likelihood that the manager will not have the benefit of having actually seen a replay or receive communications from someone who has before he challenges, and you’re now asking a man who may have had the worst view in the park of a certain play to make a decision. Even plays that are close to the dugout will often be obstructed to the manager; the fan sitting 10 feet behind him will have more information if he’s fired up MLB.tv on his mobile device than the manager making the call.

This is just going to lead to more heat for the skipper, especially when the inevitable situation comes where he incorrectly challenges a call he may not have seen because his player insists he was safe, even though everyone on the planet with the game on television may know that it’s the wrong call before the flag (or whatever the mechanism will be) comes out. It’s not fair to the manager, and it’s not good for the simple goal of getting calls right, especially when some managers may choose to abuse the system for other goals. (It’s not hard to see a Joe Maddon type forfeit one of his challenges on a clearly losing call just in order to gain a few more warm-up pitches for his desired reliever, is it?)

The impetus for this idea on MLB’s side is that with limited challenges, they can attempt to avoid the average game reaching the four-hour mark. That’s admirable, but there’s better ways to do it. Here’s two.

1) Eliminate the manager entirely. Getting the call right shouldn’t be about strategy like it’s merely bringing in a pinch-hitter, it should be about getting it right, and that should be on the league and the umpires to ensure. But you can’t reasonably expect the field umpires to constantly second-guess themselves, so a simple solution here is to implement an idea that’s been kicked around for a while — hire a video tech at every park, and pair him with an umpire who has decision-making powers. Maybe that’s a retired umpire who is local to the area; maybe it’s adding a rotating fifth full-time ump to each crew who can spend 20% of his time in the booth rather than on the field.

This two-man team would watch every play in real-time from multiple angles, and though it sounds like this might add an egregious amount of time because every play would technically be reviewed, if done right, it wouldn’t. 98% of the time you’d never know they were doing so, because most plays are routine. When a replay is required, they would simply buzz the crew chief to pause the game. To avoid undue stalling by a pitcher or batter to get things moving in hopes that a replay might come down, simply enforce the currently existing yet rarely enforced rules about how quickly a pitcher and better must be ready to play. (And, it should be noted, count a pickoff throw as “the next play,” rather than just a pitch.) For managers, eliminate time spent arguing — which can often be three or four minutes, sometimes more — by following the lead of basically every single other sport and demand they stay off the playing field under threat of ejection, especially relevant now that Lloyd McClendon is back in charge in Seattle.

Other than the NFL, this is how most other sports that have a replay system use it. That’s because it’s the right way.

2) Allow the manager to receive assistance. If you absolutely must have the manager involved, then it seems unfair to ask him to do it blindly. In the NFL, coaches are constantly on headsets with staff members who can tell them when to challenge plays, and while that level of communication wouldn’t fly in baseball — imagine the sign-stealing implications — a simple “yes/no” buzzer would at least allow them some level of competency in making the call. If we’re trying to speed up the game, the last thing we need is for managers to be wasting time on obvious losers — a feeling the managers would probably agree with, especially when a very obvious outcome is the advent of sites like ours beginning to track stats of the success rates of these calls.

The mere fact that more replay is being added is a good thing, almost no matter how it’s implemented. But the idea of having managers be the mechanism for getting replays started is almost hopelessly flawed, and seems to be setting the sport up for a lot of problems. It might be too late to change it now, but it might not: there’s a better way to do this.