This is not some kind of intractable problem requiring a 12th level intellect. There are, of course, finer points of implementation that would have to be worked out, but it’s not a moon shot by any means.
Whether we should do so or want to do so is clearly a separate question. And if the argument is that there’s not enough desire for it or not enough comfort with the concept, hey, that’s well and good. I’ve always thought errors made by human officials were part of the game.
As long as they don’t favor one team over the other and aren’t egregious, it’s something we enjoy grousing about on Monday mornings. But if the goal is to eliminate such errors (which I’m not personally all that keen on), relying on humans is never going to get there.
Yeah, me too, as I’ve said. It’s like, regardless of everything else that’s changed, we still do this one really old thing. And watching them fuss with it raises anxiety, like everyone holds their breath waiting as they put the pole down with the tight chain, then lets it out with relief or a curse (depending on which team you’re supporting). And when it’s an inch away you just see it on the screen, it’s a clear thing that just shows how close it is.
But, objectively this is faster and it’s not like the technology is a novelty, they’re just finally catching up to the rest of the world. It’s for the better.
I hope they have their shit together on this. One big problem is that stadiums have some big variations in their camera coverage and placement. I recall a couple times this past season where plays couldn’t be reviewed because the home team didn’t have a sideline view of the pylon. Whatever this technology is, I hope they already have all the hardware installed and validated at every stadium. Also, I know they are keeping the chain gangs as a backup, but I can foresee this falling apart in bad weather.
As far as I can tell this isn’t actually improving the situation with the spot of the ball. It’s only removing the need to walk the chain gang onto the field for short yardage calls. We will still have idiotic spots screwing things up for years to come.
Ooops…should have finished reading the thread before responding.
I imagine from a skills standpoint a kicker is always going to have an easier time adapting than a punter. If it’s still coming off a tee, you have to expect that the punter is going to have a higher risk of a shank. This really only makes sense if you get lucky and your punters full power place kicks naturally land in the landing zone.
No, the real problem is the subjectivity. The issue with spotting the ball isn’t generally knowing where the position of the ball is, the issue is with knowing where the ball is at a specific moment in time when some other action happens.
Where was the ball when the ref decided the runners forward progress stopped?
Where was the ball when the runners knee/elbow hits the ground?
Where was the ball when the runner crossed the sideline?
At some point AI vision could probably solve a lot of these problems, but that’s going to be a long ways off. Chips are almost certainly not part of the solution; it’ll be camera based when it happens (see baseball, self driving cars, etc.).
All of that is solvable pretty easily, I would imagine, if you really wanted to go to electronic spotting.
Two chips per ball, maybe a quarter of the way in from the skinny sides. These two chips would allow calculating orientation of the ball so you could tell if it broke the plane regardless what angle the ball was being held at.
These chips would also have some sort of time component, meaning you could tell the exact location of the ball for every millisecond. Then all the humans have to do is determine when the runner is down. That would let the chips tell you exactly where the ball should be spotted or if it’s a touchdown. Sync up the high def slow-mo replays with the timing on the chips and you could quickly and easily make precise determinations on when the runner is down.
AI Vision is a million times simpler and more reliable. You might as well be proposing the horse and buggy. It would be almost trivial to train a system to constantly review spots if you have sufficient camera coverage.
It may be dramatic, but measuring with chains has been fundamentally broken from the start. I get that the chain is ten yards long, but how accurately can they position it? A team gets a first down, the refs spot the ball, and the guys on the sidelines try to place a stick in line with the nose of the ball. But those guys are 70 feet away; wouldn’t surprise me if they were off by a couple inches one way or the other. And yet, when they trot those sticks out onto the field three plays later, everyone treats them like they’re exactly right. It’s nuts.
Agreed. I watch a play with my own eyes, where you don’t have a mass of bodies, and it’s pretty obvious even in real time where the ball was at when the runner was down. And an official places it at some crazy location. The “magic robot chain” or whatever they’re going to call it isn’t going to fix that, just make it faster to figure out if where the official put the ball is a first down.
I mean, we have reviews and challenges, and lots of camera angles, so obvious blunders like this get overturned much of the time, but still. The issue is where a human is putting the ball, not how it gets measured afterward.
It’s called Hawkeye. I suppose they could end up rebranding it, but I’m guessing there’s a deal in place with Sony to co-brand it.
The more I think about, the more convinced I get that this is just the first phase of this. The problem they are solving is borderline negligible. Reporting is that this will save around 40 seconds on each measure. If there’s maybe 2-3 measurements per game that’s a savings of a whopping 2 minutes out of a typical 3.5 hour game. And in general, this isn’t solving a accuracy problem. The measurement is still based on the official’s manual placement of the ball, and I don’t think anyone was upset about chain gangs screwing up the stick placement.
I think this is just the trail balloon. Over time they will be using this as the starting point to roll out additional AI vision capabilities for ball and player tracking. It would be trivial to add things like neutral zone and illegal formation identification to this system, but they need to get teams comfortable with a robot throwing flags.
That sounds very plausible, I’m guessing that you’re right on this. It’s making a minor change to the game and if it goes wrong it’s not going to be a big deal, but if it works then they can use similar technologies that have a bigger impact on the game.
Yup. A big part of any vision technology will be getting the cameras in place uniformly across stadiums and then accumulating a bunch of labeled training data to feed the models. They’ll probably want to run this for a few seasons before they start asking it more nuanced questions.
Maybe they start using it to grade officials privately before they consider anything public.
Browns sign Joe Flacco to a one-year deal for $4M. Kenny Pickett is still the supposed starter, depending on what the Browns do in the draft. And depending on whatever dumb decision the Browns make next.
As that article indicates, it looks like his success rate (or lack thereof) at 50+ yards was a big reason why. Last season, he was perfect on XPs, and on field goals inside of 50 yards (28 out of 28), but only 3 out of 9 from 50 yards or longer.
This is some interesting news. Most of the discussion is about how this could potentially alter the draft. Personally, I’m hoping the Saints step into the Rodgers sweepstakes and snatch him from the Steelers. That would be peak comedy.
He had been projected to be picked in the first round until his recent legal trouble. Sources say it was a suicide.
Editing since I read a different source right after I posted. Apparently he got into a family argument, pulled out a gun and fired a round into the ground, got into a police chase and shot himself while still being pursued. He was found dead when they approached the crashed car.