Goal Line Technology in NHL

[tarheelhockey]

there's already rinkball

That looks like a lot of fun.

Hockey was originally played this way. They cut the ends off the sphere to make a puck, because a ball was quite dangerous for spectators in the days before there were boards around the rink.

 

[Kahvi]

That looks like a lot of fun.

Hockey was originally played this way. They cut the ends off the sphere to make a puck, because a ball was quite dangerous for spectators in the days before there were boards around the rink.

That was fun when I played it for fun in the early 2000's university student. At least in the 80's, 90's and up to early 2000's that was the equivalent of NA Beer League hockey in Finland. No contact sport, so it was cheaper because all you needed was skates, stick and helmet to play, and maybe something else just in case. In the 80's and 90's it was also a really popular winter sport in so-colled company sunday leagues, meaning most local companies even in smaller towns had at least 6-7 guys who could skate and ice a team, just for fun and exercise. Did I compare it to beer league hockey already?

At its peak in the 90's Finland had structured system of a 3 tier national league system, maybe some teams in the top league were semi-pro and rest was fully amateur.

Not sure about the history, but I'd guess it originated as a mixture of bandy and ice hockey, looking a lot like bandy played in the rink instead of soccer field.

 

[Stephen]

I work with high accuracy GPS systems, and it's quite involved to get down to sub 1cm accuracy.

I'm not even sure how you would implant this stuff into a puck, and it could require additional base station nearby (maybe attached to the glass behind the net?).

Then on top of all of that, the people bringing up the fact that the puck isn't spherical are right on. It means you need the exact angle, as a slight difference in tilt in the puck could be a difference of a few centimeters.

Overall, it's not nearly as easy as people in this thread are making it out to be. Maybe possible, but certainly not simple.

Fair enough! Yeah it’s frustrating to think all the pieces are there to get this right but from a practical sense maybe it doesn’t need a whole research and development wing to try and solve a problem that’s still pretty intermittent (tell that to the team and fanbase that gets screwed).

 

[qcal1427]

IIRC in addition to all the other problems mentioned I think they did some tests with chips in a puck and it totally threw off the weighting/balance of the pucks to where players definitely noticed. I think this was on a 32T?

 

[Kahvi]

Some reasons why they might not want to use chips in puck for this

• Pucks get a lot more beating that soccer balls. You need to make sure they work even when a frozen puck hits the crossbar after a 100mph slapshot and falls to the goal line
• They work already in puck tracking in advanced stats, but if something gets broken, its only the stats that are affected. What if the chip malfunctions when you have a critical goal review
• Chip alone is not enough, you need some kind of gyro or something to monitor the position of the puck, like others have said. More technology inside the rubber that might break and make the monitoring more unstable

 

[tarheelhockey]

Not sure about the history, but I'd guess it originated as a mixture of bandy and ice hockey, looking a lot like bandy played in the rink instead of soccer field.

I was thinking the same thing, it looks like bandy with hockey sticks on a hockey rink. Which IMO is a big improvement over bandy, which has too much field hockey influence.

 

[Kahvi]

I was thinking the same thing, it looks like bandy with hockey sticks on a hockey rink. Which IMO is a big improvement over bandy, which has too much field hockey influence.

If I remember correctly, bandy originated as a "what the hell should we do with the soccer fields in the winter" in Sweden. And bandy gave the NHL the godlike skating of Reijo Ruotsalainen.

 

[Summer Rose]

If I remember correctly, bandy originated as a "what the hell should we do with the soccer fields in the winter" in Sweden. And bandy gave the NHL the godlike skating of Reijo Ruotsalainen.

Sounds like Australian football. A very basic summary is that it was invented in order to give cricketers something to do in the winter in order to keep fit.

 

[kingsholygrail]

I just don't think it's a common enough problem to warrant the investment.

 

[HockeyVirus]

Here is why I think they are hesitant.

Imagine a scenario a clear save is made and it signals it as a goal. It questions all the other rulings of the tech and questions the credibility of the sport. Soccer is in the position to take those risks.

 

[hatterson]

Soccer is able to do it because they have a bunch of cameras. The issue here was that the puck was in the glove and thus there was no unobstructed camera shot.

And there are examples of it not working when the cameras have an obstructed view (like would happen on virtually every relevant play in the NHL). There's usually one every year or two where GLT fails because a keeper was smothering a ball or a defender was trying to squeeze it against the post and the cameras didn't have their usual angles.

 

[Blueline Bomber]

IIRC in addition to all the other problems mentioned I think they did some tests with chips in a puck and it totally threw off the weighting/balance of the pucks to where players definitely noticed. I think this was on a 32T?

It was an issue about 5 years ago when this first started. It's not been an issue since then.

 

[A Real Barn Burner]

I work with high accuracy GPS systems, and it's quite involved to get down to sub 1cm accuracy.

I'm not even sure how you would implant this stuff into a puck, and it could require additional base station nearby (maybe attached to the glass behind the net?).

Then on top of all of that, the people bringing up the fact that the puck isn't spherical are right on. It means you need the exact angle, as a slight difference in tilt in the puck could be a difference of a few centimeters.

Overall, it's not nearly as easy as people in this thread are making it out to be. Maybe possible, but certainly not simple.

You are absolutely right! Achieving a sub-1cm accuracy with GPS is an intricate and challenging process. It becomes even more complex and potentially impossible when you have a building with a roof, which introduces variations in material refraction and absorption of the signal from a constantly moving constellation of satellites due to the Earth’s rotation. If you can get a signal at all….Accurately calculating the ellipsoid’s shape at the speed of a NHL game would require more than a few additional base stations or cell tower triangulations, along with an excessive amount of processing power. Not to mention the chip and power needs inside the puck to do such a thing.

That’s why the NHL doesn’t use GPS. The way the NHL currently tracks the puck hasn’t changed since the 1990s. Both the NHL’s current EDGE system and the 1990s FOX track system use infrared cameras and sensors embedded in the puck.

Since infrared light necessitates a clear line of sight, it’s ineffective for the Jets’ probable objective Goal from yesterday’s game. Once the puck is inside a goalie’s glove, beneath some pads, or bodies, only the sensor on one side of the puck is visible or transmitting, which disrupts the telemetry of the signal and the precision of the ellipse, leading to incorrect triangulation. When the ellipse elongates in one area because 17 out of the 20 cameras can’t see the puck, you’ll receive signals indicating a goal even though the puck never crosses the line. This is why the NHL doesn’t use it for goal calls. Any decision that required a referee more than a minute to make is not likely visible to these infrared cameras either.

The best bet for something like what soccer has would be RFID technology or a magnet based technology. Infrared or lidar might work….but whatever system they use would probably require sensors be put into the ice itself to make goal calls. This would make any light based system difficult. If they did this along with the cost it would make the setup times for NHL game even longer.

Sure we will get there eventually, I can’t wait to see what they come up with!

 

[hatterson]

You are absolutely right! Achieving a sub-1cm accuracy with GPS is an intricate and challenging process. It becomes even more complex and potentially impossible when you have a building with a roof, which introduces variations in material refraction and absorption of the signal from a constantly moving constellation of satellites due to the Earth’s rotation. If you can get a signal at all….Accurately calculating the ellipsoid’s shape at the speed of a NHL game would require more than a few additional base stations or cell tower triangulations, along with an excessive amount of processing power. Not to mention the chip and power needs inside the puck to do such a thing.

That’s why the NHL doesn’t use GPS. The way the NHL currently tracks the puck hasn’t changed since the 1990s. Both the NHL’s current EDGE system and the 1990s FOX track system use infrared cameras and sensors embedded in the puck.

View attachment 1022654

Since infrared light necessitates a clear line of sight, it’s ineffective for the Jets’ probable objective Goal from yesterday’s game. Once the puck is inside a goalie’s glove, beneath some pads, or bodies, only the sensor on one side of the puck is visible or transmitting, which disrupts the telemetry of the signal and the precision of the ellipse, leading to incorrect triangulation. When the ellipse elongates in one area because 17 out of the 20 cameras can’t see the puck, you’ll receive signals indicating a goal even though the puck never crosses the line. This is why the NHL doesn’t use it for goal calls. Any decision that required a referee more than a minute to make is not likely visible to these infrared cameras either.

The best bet for something like what soccer has would be RFID technology or a magnet based technology. Infrared or lidar might work….but whatever system they use would probably require sensors be put into the ice itself to make goal calls. This would make any light based system difficult. If they did this along with the cost it would make the setup times for NHL game even longer.

Sure we will get there eventually, I can’t wait to see what they come up with!

Yea that’s the key. It’s not that no possible solution could work, it’s that to get a solution that actually meaningfully improves on just dudes watching a screen with multiple camera angles and trying to guess, you would have to pay a lot of money.

The NHL has made the call (rightly so IMO) that the marginal benefit of a highly accurate goal decision system is not worth the cost it would take to install and maintain such a system.

 

[Anglesmith]

In geometry class when kids ask the teacher when they're going to need this in real life, the correct response is "to avoid looking silly in an Hfboards discussion."

 

[DudeWhereIsMakar]

That was totally a goal! And I'm not saying that because I'm a Jets fan. It was in the white.

Probably doing it because they didn't want Binnington to throw a tantrum.

 

[dabeechman]

It's really not that difficult of a problem to solve, and as an electrical engineer I can say that with a fair amount of confidence. Are there challenges? Yes, of course. But nothing that isn't insurmountable.

I believe the desire of the NHL though is to not phase out the relevance of the officials and to create some controversy and drama. Same for why they don't allow replays on most penalties, despite the growing number of phantom calls, self inflicted fouls (i.e. clamping sticks under arm to draw hooking calls) and dives that are plaguing the game today.

Puck over glass, high sticks, etc. would also benefit from positional technology because we've seen far too many times that the calls are wrong, and now that there is betting involved, everybody needs to be very cognizant of the financial ramifications for missed calls and claims of game fixing (whether warranted or not).

 

[algernon]

I work with high accuracy GPS systems, and it's quite involved to get down to sub 1cm accuracy.

I'm not even sure how you would implant this stuff into a puck, and it could require additional base station nearby (maybe attached to the glass behind the net?).

Then on top of all of that, the people bringing up the fact that the puck isn't spherical are right on. It means you need the exact angle, as a slight difference in tilt in the puck could be a difference of a few centimeters.

Overall, it's not nearly as easy as people in this thread are making it out to be. Maybe possible, but certainly not simple.

Why not coat the pucks in a radar-reflective coating while putting 6 sensors around the perimeter of the puck?

 

[andora]

A major part of why it's easy for them in soccer is the ball is spherical and the distance from the center point to any other point on the ball is equal distance. Thus using sensors and cameras it can determine/triangulate the center point of the ball it's easy to calculate its radius and if it full crossed the goal line. The shape of the puck in hockey is what makes this a challenge.

Math exists

 

[hatterson]

It's really not that difficult of a problem to solve, and as an electrical engineer I can say that with a fair amount of confidence. Are there challenges? Yes, of course. But nothing that isn't insurmountable.

I believe the desire of the NHL though is to not phase out the relevance of the officials and to create some controversy and drama. Same for why they don't allow replays on most penalties, despite the growing number of phantom calls, self inflicted fouls (i.e. clamping sticks under arm to draw hooking calls) and dives that are plaguing the game today.

Puck over glass, high sticks, etc. would also benefit from positional technology because we've seen far too many times that the calls are wrong, and now that there is betting involved, everybody needs to be very cognizant of the financial ramifications for missed calls and claims of game fixing (whether warranted or not).

If it’s so easy, then design an economical system and sell it to the league.