June 17, 2004
Improving Umpires
Derek Zumsteg pens a piece about the state of umpiring. In general, it's positive.
The state of umpiring today is amazing. While umpires are devoid of the kind of personality that, say, Ron Luciano had, as a group they have improved so markedly since baseball broke their union that it's amazing to watch old games on ESPN Classic, even playoff games. Umpires today are faster to get into position and more observant. They're willing to consult other umps who might have a better view of a disputed play. They're far more professional than their predecessors.
I'm not surprised by this. In my initial rant about
the union stopping the A-Rod to Boston trade I wrote:
It's my opinion that the MLBPA is the greatest labor union the world has ever seen. Most labor unions set up systems that insure the continued employment of their least talented workers, while limiting the advancement of their most talented. The MLBPA has done just the opposite. They set up parameters for salary growth and negotiation, but how much you make depends in great part on how well you perform. They protect players rights extremely well, make sure they have generous benefits and proper legal counsel. It should be the model for how all unions operate.
The umpires union was protecting the bad umpires, bringing down the overall quality of their crews. If, like the MLBPA, they had worried about salary and benefits, and let umpires ability to umpire determine participation in baseball, they still might be in business.
Derek, however, believes things still need to get better.
I am more convinced than ever that the umpires have demonstrated the need for better strike zone measurement tools. We haven't heard much about Questec this year, due in part to Tom Glavine enjoying a bounceback year. But I watch so much baseball it frightens small children, and I see blown balls and strike calls all the time. And I don't even mean close calls, either, I'm taking about clearly up, down, or off the plate and my favorite, Ye Olde Hit the Target Strike. Like the other parts of the umpire's game, it's gotten better, but it's still not as good as it needs to be.
Derek feels we need some automatic tools to help umpires. The problem is,
automatic tools make mistakes also. In my previous job at the University of Massachusetts, I worked in a lab that conducted research in information retrieval. What I basically learned is that these types of problems are very difficult to solve exactly. So, if you go to an automatic system, you're still go to have bad calls; the difference, I believe, would be that the bad calls would be on the edges, rather than the flagrant calls Derek discusses above.
One thing that would help in a video system are more cameras at more angles. But that's a big cost in both equipment, and time to analyze all that data. A cool solution would be to use GPS. A GPS reciever would be built into the center of each ball. The problem with this, of, course, is that GPS is only accurate at best to about 1 meter, which is way to big for calling strikes. However, if instead of using the satellites, a large number (>5) transmitters were built into the structure of each ballpark (on each light pole, for example), I bet you could get very good accuracy for the position of the ball during flight. You could also put four sensors on the batter, to frame his strikezone.
But would all this technology really improve things? I think the correct use of any ball/strike location system is the way Questec is being used; to teach umpires what they called wrong so they are less likely to make the same mistakes in the future. Calling balls and strikes acurately is as difficult a problem for machines as it is for humans. Let's use our machines to work with our umpires not replace them.
Posted by David Pinto at
02:33 PM
|
Umpires
|
TrackBack (0)
Would the GPS inside the baseball get smashed to smithereens every time a player hit the ball?
I don't think so. There's a lot of yarn insulating the center from the outside of the ball.
Put the receiver in the middle of the rubber center? I smell "they juiced the ball" if hitting goes up or "they deadened the ball" if hitting goes down.
I definitely prefer to have human umpires who are going to make mistakes. Sure, they should always try to get better, but if a team's season comes down to one blown call, then they really weren't a better team than others.
And hitters make mistakes all the time, either swinging and missing, or taking a called third strike. Likewise pitchers make mistakes, usually those're called hits.
David, I couln't agree with you more. I like the human element, and like the players, who make mistakes, I think the umpires should be expected to be as well trained and educated as possible. But having a computer and a set of cameras call balls and strikes, would, for me at least, take away some of the soul of the game.
Granted, automatic tools make mistakes also. However, the whole point of the automatic tools is, I think, that you can know the boundary cases going in. I'm not sure if there's been any comprehensive research into this, but I'd guess that human (i.e. umpire) error is almost totally random; they're just as likely to miss a pitch up in the zone as one down in the zone, or one inside as one away, etc. With an automatic tool, you can pretty much eliminate the randomness factor; errors will almost always occur in known problem cases, and an umpire or operator can adjust accordingly.
So basically, I agree with you. Bad calls would be on the edges. But if they're on the edges, they're not really "bad calls", are they? They're just questionable calls, in which cases the decision could either be left up to the umpire, or it could be decided that "all ties are strikes," or something similar. Which I would think would be a marked improvement over the essentially random "flagrant" bad calls that currently appear.
There must be some unobtrusive way to implement some technology elements into umpiring the game.
I know this idea has been given the pinata treatment in the past, but I'll float it anyway, as foul-ball HRs only happen a few times each year, and I can take advantage of the hubbub. Why not give each manager the opportunity to instant-replay one on-the-field call per game, which would be limited to:
1. foul/fair balls called
2. force/tag outs
...no ball/strike challenges
One per game, limit the review time to 30 seconds. HRs would quickly be restored/overturned, runners either on base or the bench.
Yes, it smacks of the NFL, but if done properly (which it is in that lesser sport), much of this discussion gets mooted.
I may be delusional, but I don't think IR is needed in MLB, because I think the VAST majority of times, the umps get the calls right.
The foul home run last night of course is a glaring example of it being wrong. But you see something like that once a season.
If we follow the logic that no external mistakes should color the outcome of the game, what about the Bartman situation.
Should fans not be allowed to sit close enough to the action to potentially interfere with a playable ball?
I think the Bartman play was certainly unfortunate for the Cubs, but certainly within the scope of the game, and didn't involve any umpiring errors. Blown calls on forceouts, tags, etc. are mistakes, but I can handle them being part of the game. It's things like the Jeffrey Maier "home run" and yesterday's screw-up in Montreal that they need to be able to do something about. How about Instant Replay, but not at either of the teams' requests, at the umpiring crew chief's request? They say the cardinal rule they want to follow is to get the call right. I wouldn't want to see somebody like Lou Piniella or Lloyd McClendon arguing with an ump that he ought to look at the IR, though. (On second though, maybe I _would_ want to see that.)
Adam
David, I had the same idea for the "FoxPuck Trax" system and sensors on the players... but not for calls, but for analysis.
I want to know the speed, spin, trajectory and location of the ball at all times, as well as where all the fielders are, at every moment of the ball's flight. Then, we can get near-perfect fielding metrics.
I don't see umpires as a problem. They're the least of the issues that need to be tackled, and consequently, the least of the effort/money to spend on.