That’s it, Silenius. I think Bill Veeck had that installed, perhaps to pacify umpires irritated with things like his exploding scoreboard.
Hey, that could become one of the ballpark parameters that home teams get to adjust, like moving the fences or cutting the grass. I think more variation in baseball field conditions would be fun. As long as it’s applied fairly and accurately: the same for both teams, for all players, in all game situations.
Ok. So that puts the camera on 1st base (and hopefully 3rd for guys batting left). That doesn’t encompass the two corners of the strike zone so I assume they have to calculate if the ball is in those regions which will sound like magic to some people. But if they had the camera perpendicular to the line between home plate and second base the ball could be seen on video by the umpire moving through the entire zone. The umpire can decide if it’s also conforming to the batters knees and mid-torso.
That’s accurate enough to allow replay confirmation.
Let me also say that as long as we’re stuck with human umpires calling balls and strikes, I still don’t want them reviewable. Not even once per game. Like I said before, as long as umps are consistent, I can live with weird strike zones. I wouldn’t want a guy who’s been getting an inch off the plate all night to get screwed by PITCHF/x in the 9th inning.
I thought it was slang for the telephone under the plate for the league officiating office to tell the corrupt and incompetent umpire how he was supposed to call.
:rolleyes:
Agreed. Ball and strike calls are unreviewable for that very reason. I also agree that umpires should be consistent, and their training should focus on that. If MLB finds that an ump is applying their strike zone in a partial manner, that ump can be disciplined or fired.
This isn’t a logical fallacy, and I’m not suggesting that it’s a binary problem.
My argument: Error and bias should be minimized in all forms.
The argument no one is making, that I’m not arguing against: Error and bias are great. We should have more everywhere.
The argument several people are making, which I am arguing against: It’s complicated. Sometimes error and bias are good, and sometimes bad.
My challenge to anyone making the third argument is to name any other place in the game where adding in some error or bias would improve things. If you can’t think of a single one, then either we’re extremely lucky to have happened upon the perfect combination of empirical measurement and human adjudication that makes baseball the best possible game, or you’re suffering from status quo bias.
Maybe we should just set up some cameras and, based on the evidence, fire all of them, because a human being can’t possibly do the things we ask umpires to do reliably. Training helps, but it can’t overcome physical limitations.
I agree. I was initially lukewarm on the idea of computerized pitching calls, but I’ve long since come to accept that this really is the best and fairest solution. I see no good reason not to implement it, and I’m a guy who doesn’t like the DH and is wistful for all day games at Wrigley, too. I see no way errors in officiating create a “better” game by any measure.
The problem is there are calls that need to be made instantly.
Such as a ground ball ( with runners on) directly over 1st/3rd. Was it foul or fair?
Infield fly rule.
Foul tip.
Do you think people are faster or better at making those calls than computers? I’d bet that they are not.
I will admit that I’m not sure how to programmatically determine if the Infield Fly Rule should be in effect, and we should probably keep a human official for those calls for now.
But “did the ball pass through a well-defined volume” is a question that is demonstrably better decided by computers than humans.
Really? Like this one?
The strike zone is not that well defined. We know the pentagonal column of space that extends infinitely upwards from home plate, but the area between the knees and mid-torso is not so well defined, and it’s definition is not intended to provide parameters to a machine, it is intended as instructions for human umpires using their judgment.
I disagree. The rules are clearly intended to identify a defined space for the strike zone, with specific and defined parameters. How well these parameters can be reliably implemented by human observers is debatable, but the definition is specific. In fact, it’s a good example of how technology can better employ the codified rule than a human observer.
There’s also nothing in the definition of the strike zone that explicitly states that these parameters may be implemented or modified at the discretion of the umpire.
In the kind of exact terms necessary for a machine to work with, how do you define the location of the knees and the mid-torso?
[quote=“koufax, post:91, topic:728149”]
Really? Like this one?
[/QUOTE]To be fair, the location that is being highlighted in that video is the location of the catcher’s mitt when he catches the ball…which is not necessarily the same as the location that the ball was in as it passed over the plate. If a pitcher throws a ball that has some pretty good lateral movement, it’s quite possible for it to pass over the front corner of the plate, and then be several inches off the plate by the time it gets to the catcher (whose mitt is several inches behind the back of the plate).
That said, most of the pitches in that particular were nowhere close to being strikes in any case…
But this video is a good example of why ‘pitch tracking’ used to be a demostrably unsatisfactory alternative to umpires; they could pinpoint where the ball was caught, but not where it was when it passed over the plate in midflight. I assume that current technology is vastly superior, but I’m not certain if it’s advanced to the point where it can reliably determine is the ball has touched any point of a variable (in height) three-dimensional rectangular prism while it passes through it on the way to the catcher.
You adhere to the baseball rulebook. At the extreme, you place a tiny marker that the machine can detect in each player’s uniform at “the hollow beneath the kneecap” and another at “the midpoint between the top of the shoulders and the top of the uniform pants” when they are in their natural batters stance.
At the less extreme, you have those points estimated without requiring the physical markers. They must already be doing that for current strike zone technology, right?
I’m not clear yet what current technology does. Yes, you could add physical markers but that adds new rules. I’m just saying it’s not as simple as taking our best technology and implementing it today. My personal preference would be to use video solutions for replay capability, but I’ll focus now on the technical aspects, and while I think the equipment is accurate enough to register where the ball is relative to the plate it will still take more to have it replace a human umpire altogether.
Perhaps another approach is to define a specific strike zone height starting from the knee caps which must somehow be clearly marked. Or for all around fairness (for some values of fairness) the strike zone is defined as a height range over the plate and no longer based on a players movable and indistinct body parts.
How would what I wrote add new rules? I was quoting specifically from the MLB rule book where the strike zone is defined.
No machine can use that definition of the strike zone without marking the uniforms or using some kind of sensor to determine the points on a players body unambiguously. How does a machine determine what a player’s natural stance is? How does it see a hollow below the knee through a uniform? How does it determine midpoint from a 2D image of a players irregular torso in an arbitrary position?
One could place markers in the uniform at these points. That no more creates a new rule than the grounds crew is creating a new rule when they chalk in the foul lines before the game.