ha can you imagine replays on the big screen as both sets of fans wait to see if the ball crossed the line in slow motion, may work with tennis fans, yet I think the nature of football fans would end in riots!
They said that 5 years ago in regards to computer facial recognition. Today, an off-the-shelf computer can recognise 300,000 faces a second. Body parts are recognised by computers today using toys - think Microsoft Kinect. Forget the chips on body parts; that's a dead end. It's all fascinating stuff and I welcome it all if it means more fairness.
One of the players last night was given a yellow card for diving so and rightly so. Some penalty decisions are pathetic though and maybe some sort of technology should be adopted for that too.
Fai enough Inda, but I stick by my prediction you won't see it in the Prem before 2025, if you ever see it. I could list so many objections as to why it won't happen any time soon, but I'm content with my prediction.
Could happen, but highly unlikely to happen any time soon. Almost certainly it would require video replays, and that's still a big no-no, for many reasons, some of which have a lot of credibility behind them, even if they're not deal breakers eventually. But in the short-term they certainly are.
Disagree. Technology in matches should begin and end at goal line technology as it's by a long way the most fundamental problem we have at the moment. Dives should be dealt with retrospectively though and a ban brought in for those who are caught red handed.
Well said YV, this technological fix solution for complex decisions like offside and pdiving etc, it really is a long way away, there are so many reasons why it won't happen. Chip-centred solutions, forget them for any short-term timescale. VRs, there are so many reasons against them, one obvious point being many grounds round the world won't have the facilities that Sky provide for their featured Prem games. FIFA don't want such a fundamental change to the game to come in for certain games only, ignoring the vast majority of professional matches played upon this globe on a weekly basis..
It's not just the level of technology needed, I neither want a "perfect" game or want any unnecessary stoppages and delays to the game.
Yep, I agree, that's part of the the point I'm making why chip-centred solutions won't work. It's not just to do with the technology, it's to do with FIFA, football politics, crowd psychology, the football experience, the fact that players cheat in a way that robots haven't learnt to, yet, and so on and so on. When people start talking about technology in these matters they often forget the human side. It's taken decades to get near a technological solution for goal line decisions that satisfies all the interested parties, or enough to get it into our game, and BTW, it still hasn't happened yet. But goal line decisions are simple compared to the complexities of offside and diving. Yes there massive technology issues to resolve, but massive human issues too. For example, in some football cultures, diving is seen differently to other cultures, though there seems to be an evening out of this diversity as TV takes the game global and players move across continents.
On the subject of dives, has anyone else noticed that Ivanovic's one against us went virtually unnoticed, despite him being booked for it? Why no Bale/Suarez-style inquest in the media, especially as he's done it more than both of those players already this season?
There have been lots of interesting and well-informed comments on this thread. My last one was too hasty, though. Having given it a bit more thought (on further review), offsides detection is not a trivial issue (though goal detection should be), and may or may not be quickly solvable, from what I know. If we go with my idea, which is a computer vision solution, you'd have to train the system to recognize a "pass" which is not on its face straightforward, particularly since it would have to distinguish it from a "shot". The fact that categorization isn't always clear to human observers (or even, on occasion to the player making the pass or shot) may mean the human judge and the review booth are stuck in the loop, and if that's the case, there may be no great advantage for an automated system, and perhaps no way to avoid delaying the game. Bottom line: an automated offsides detection system would be an interesting challenge for computer vision specialists to study. If I had to guess, I think they would be able to come up with a good, workable system, but I'm far from certain about this. I also should make clear that I'm not a software engineer, I'm a designer. While I write code for our website, and developed code teaching curricula, I'm nowhere near the level of expertise of the computer vision specialists who would have to tackle the problem.
Kinect is remarkably imprecise, though - surprising, that, Microsoft releasing a shonky product with a lot of hype and an inflated pricetag...
Hi RWB, as I said no chance in the foreseeable future. 2025 maybe , just maybe, still doubt it then. When Spurs sign their first robot, then it's a goer, but I expect i'll be dead by then. As long as football is a human game, such techno-centric solutions about offside, diving, etc, have a long, long, way to go yet.
I would never try to detect diving--it just doesn't seem doable. I agree that's a perfect candidate for retroactive punishment. Hand out a few retroactive red cards and you won't see so many going down faster than Juarez hooker. I also have to say a team full of robots would be a little short on appeal for me. But offsides is really difficult for a person to detect, and my guess is that machines can do a better job--if not this year, than within five years. This particular branch of tech is moving ahead fast right now. The key element which makes things like this (though not necessarily this particular thing) doable in the near future is machine learning. Cameras and ladar receive tons of data. If enough people create enough training sets which distinguish one thing from another (say, a shot from a pass), the computer can be programmed to fine tune, by process of elimination, the thing or things it needs to look for to make the same distinction. Compared to driving a car, which computers can now do, it seems relatively straightforward.
Well we just have to disagree on this RWB. To me there is no chance whatsoever you will see a chip/computer/robotic-driven solution to offside in the Prem within the next five years, nor ten years. I very much doubt within 15 either, for all the reasons given by me and others in this thread. In the next ten years the only realistic chance is Video Replays, but again for all the reasons given I doubt that will happen in the Prem in that time-frame either.
Have you guys seen the robots playing football on YouTube? They're not full size (shame), and it gets a little boring after 5 minutes, but it's something every football fan should watch for the amusement factor. I really don't want to get deep into this; there are other forums for that but... If you fall to the ground without being touched, that's a dive and that's something a computer could decide. "Collision detection" is something a student of video game programming learns early in their career. It's been sloved. http://en.wikipedia.org/wiki/Collision_detection You talk about cars being driven and Google is leading the way here. There's also their Project Glass project. http://en.wikipedia.org/wiki/Project_Glass Maybe the referee could wear these. The glasses could overlay a big red line showing the offside line, just like we see on our replays, but in real-time. He could even use them for video replays. We could get rid of linesmen altogether... Todays dreams are tomorrow's future.
Offside needs tech in the players kit and gear around the pitch. Their position can be pinpointed exactly and an offside called within 5 or 6 seconds by someone talking to the ref over a mic. Better than a team arguing with the ref for 60 seconds. Implementing this across the board is one of the biggest problems facing this. Clubs just can't afford this tech
Hi Inda 'todays dreams are tomorrow's future' mmmmm, I would say a lot of what you've written is the stuff of nightmare not dreams, but like dreams strongly divorced from reality. If Spurs start signing robots I expect that's the day a lot will stop caring about football quite as much. But none of the stuff you or RWB are talking about will be happening in the Prem for at least a decade, and probably a lot longer,. I think you're both mistaking technological possibilities for what people want from football, and it ain't robots running out at White Hart Lane or Old Trafford, and all the rest of the chip/computer/robotic-centred stuff you're talking about.
That's fine, Ghoddle. Incidentally I'm not sure I was saying "we will see it," so much as "It could be done if you had a certain amount of money." I doubt we'll see it in the next five years. Inda, so I gues I'm the moderate: yes for goal line and offsides detection, no for diving detection. While it's true collision detection could play a part in a dive detection system, I think the complexities of dive vs. non-dive are too much for a person to handle, let alone a computer. I'm amazed at how equivocal some cases look, even in good focus very slow motion. Think about it: you're playing for fun, someone sticks their leg in, you hop over and fall, because you think you might just twist your ankle if you try to keep your feet. If you do exactly that in a match, is that okay? It would look like you're falling when there was no contact. What if you try to stay on your feet and slip? At the least, there have to be a lot of cases where you take no action because you can't figure out what happened.
