MLB and Umpires reach agreement
Not getting a lot of press on Chirstmas Eve, either, but the
I’ve talked about this before, but it’s good news that this got resolved amicably.
The… less good news is that as part of the agreement, a couple of umps are coming back. It looks like Bob Davidson, Tom Hallion, and Ed Hickox will return. There’s also finally a settlement on severance pay and benefits for six umpires not re-hired after their disastrous mass resignation.
The two sides also seem to have reached an entirely reasonable Questec arrangement:
To resolve the grievance over the computer system, baseball agreed that umpires whose ball-and-strike calls are rated below standard by QuesTec will be evaluated by umpire supervisors based on videotape and in-game inspection.
I don’t like Questec, but the technology and the concept is sound, and I’m happy to see this resolved to everyone’s satisfaction.
Comments
21 Responses to “MLB and Umpires reach agreement”
Note to MBL:
Replace ’em with cameras already!!! Umpires get calls wrong. Cameras don’t. Or at least do a Big Ten style instant replay thingee.
Umpires with 27 years service get 350K a year? Wow! How much does the questec system cost?
Uhhh…
I don’t think the technology exists today that can completely replace umpires in all situations where their presence and judgement is required.
#1, yeah great, instant replay reviews… as if games aren’t long enough already. How often do we really see game-changing inarguable blown calls? I think it’s great that officials are conferring when necessary, as we saw several times in the past postseason, and we don’t have a lone ump holding out against his fellows like we saw in the past. When they do confer, I don’t think it would be bad thing if there was an official up in the booth looking at the replays and whispering in the field chief’s ear over the radio (and are we sure they aren’t doing it already?), but I’d hate to see it devolve into instant replay reviews on every close tag and challenges from the dugout and yet more delays so the networks can help their advertisers peddle beer, cars, and erectile dysfunction drugs.
There’s no way Questec is going to save them any money. Even if the guy behind the catcher was doing nothing more than looking at a couple of lights that told him to call “ball” or “strike” you still need someone there to call plays at the plate. And cameras aren’t the answer to everything. When Ichiro! is legging out an infield hit the play at first is often too close to call from every single camera angle, but I think the official almost always gets it right because he can hear the ball hit the first baseman’s glove while he’s watching Ichiro!’s foot hit the bag. Short of wiring the ball, the gloves, the bases, and the players, I don’t think you’re going to come up with a Questec system for baserunning.
Yes umpires sometimes make terrible calls. But removing them is hardly the answer. Umpires are just another factor of the game, like field dimensions or turf-type, etc. It seems like the game would become painfully mechanic if computers and cameras made all the calls. The best pitchers are able to adjust to particular umpires’ strike zones and even though its annoying when you see a bad call go against your team, the human factor and fallibility has always been a part of the game and should stay that way in my opinion.
I wasn’t advocating ditching umps, guys, and I’m sorry if I came across that way. My question was merely asked out of curiousity, and it still stands…
Well, according to this story on Questec’s on site it costs about $40,000 per stadium. Of course, there’s going to be an ongoing service/calibration contract that will quickly eclipse that, not to mention the replacement cost when Schilling breaks it.
I like the human factor that the umpires bring to the game, and nothing technologically sound will ever replace the feeling you have for umpires before, during and after games. The game may be called better here and there, but overall I think you’ll lose a lot of interest in baseball if the new system completely took over baseball. I like that the umpires are being monitored with this technology, however. It takes away a lot of the paranoia factor about umpires making bad calls in the Yankees’ favor…
Just musing on the finer details of Questec technology:
On Questec’s site they claim the system’s accuracy is 0.5 inches or better. I wonder about that… it would depend on the frame rate of the cameras, higher frame rates obviously being better.
I recall that the Mac-Cam which was introduced into tennis in response to John McEnroe’s tirades had a frame rate of 500 frames per second (!) which produced excellent slow-motion replays and removed all doubt about line calls. I’m guessing, though, that Questec uses “normal” video technology, i.e. something in the range of 30 frames per second (someone feel free to correct me on this if I’m wrong).
Given that a 95-mph fastball goes 60 feet 6 inches in only 0.43 seconds, that means the Questec operator gets only 13 frames of video for the whole pitching “event”, which means that successive frames are separated by 4.6 feet of horizontal distance!!… that’s a huge gap, meaning that the operator could quite reasonably fail to find any single frame of video showing where the ball actually crossed (or didn’t cross) the front of the plate.
Hmmmm. Can they really get accuracy of 0.5 inches by interpolating between data points which are 56 inches apart?
I sure hope they’re using a higher frame rate than 30 frames per second…
They’re not looking at frames, they’re looking at the whole path of the ball as extrapolated from positioning data taken while the ball is in flight.
Sure, but the ball breaks hardest near the end of its flight, so the data points near the end of the flight (when the ball is near the plate) are much more important than the data points at the beginning (when the ball is just leaving the pitcher’s hand).
That’s how it is with interpolation and/or extrapolation: there’s always a particular area of interest, which in this case is the plate. The data points closest to the area of interest are crucial, while the points further away are much less important. In fact, in the case of (for example) a slider with late break, the early data points may even be misleading and counterproductive. Changes in the turbulent airflow over the ball in flight can cause it to develop movement which it didn’t have as it left the pitcher’s hand.
So the more data points that are available *near the area of interest*, the better… which argues in favor of the highest possible frame rate. I’m just sayin’.
That’s…. not true. As it appears to the hitter, the break appears late, but the force on the flight of the ball is constant. See: “Physics of Baseball”, many other sources.
Um, yeah, see Newton, Isaac.
And Noel, with a name like that you coudn’t slip in a Merry Christmas at the end?
Re #14, good point, Dave. 🙂 In my defence, I’ll note that I was actually not a Christmas baby, so I don’t typically dwell on the Noel/Christmas connection. But a belated happy holidays to all, and good fortune to everyone in the New Year.
Regarding the topic of uneven or discontinuous flight of the ball: Newtonian physics is not the only influence – aerodynamics is also a huge factor, and that can be decidedly non-Newtonian.
Let’s take a more obvious example – the knuckleball. I think everyone agrees that a knuckleball can have absolutely unpredictable “late break” which is mainly due to aerodynamics, not Newtonian physics.
Would Questec be able to accurately track a knuckleball as it approaches the plate?… with an accuaracy of (allegedly) 0.5 inches or better?
It all depends on the frame rate, doesn’t it?… even at a mere 60 mph, the ball could be moving a long way between frames, and the “early” frames at the point of release would be utterly useless for interpolating the flight of the knuckleball near the plate.
If – and I emphasize “if”, because I don’t know, but if the frame rate is only 30 frames per second or thereabouts, then it would be darn difficult for the Questec operator, is all I’m saying. I’m not claiming that Questec is flawed, just that it could be less accurate than they’re claiming, unless they’re using a higher frame rate than normal video cameras.
Slight typo there – should be “accuracy”, not “accuaracy”. 😀
Aerodynamics are newtonian physics, unless you think there’s some kind of relativistic (fastballs!) or quantum {smallball!) factors to the airflow around the ball. I think what you’re trying to talk about is chaotic (turbulent) flow. This is still newtonian, just not linear. Nonlinear flow is notorious for being computationally intractable. So in theoretical terms, it may be extremely difficult to calculate the path of the ball in advance; however, in practical terms when measuring the actual path of ball the calculations become much simpler. I don’t know enough about Questec to comment on their system, but it certainly is possible and certainly doesn’t require non-newtonian physics.
Who MBL is?
Good point Jose, I shouldn’t have said non-Newtonian; air is pretty close to a Newtonian fluid. And you’re absolutely correct that it’s extremely difficult to calculate the path of the ball in advance – in fact it’s practically impossible. Fortunately, no-one’s trying to do that (I hope); not you, not me, and not Questec.
But you make the same point as I do when you say that “when measuring the path of the ball the calculations become much simpler”. I have no quarrel with the calculations: I’m sure Questec knows how to interpolate. My real question is, how accurate are the *measurements*?
Blimey, it’s Christmas Day guys. Lighten up. How come you aren’t too lit up to have such an academic conversation?
Hey Colm, good point, baseball isn’t a matter of life or death.
It’s far more serious than that. 😀
But actually I did have a good Christmas – excellent dinner with friends (unfortunately the family’s on the other side of the globe) – and hopefully you and all the other USSM contributors enjoyed your Christmas too.
By the way DMZ, you disagreed when I said that a pitch breaks hardest near the end of its flight. Doesn’t it travel on a parabolic path? Sure, the force (gravity) is constant, but the vertical drop increases quadratically.