Use of technology in figure skating tech calls

[jinhamasaki]

I'm not a seasoned tennis watcher at all, but after watching the exciting match yesterday between Djokovic and Federer, I was enamored by the technology used in the game. When shots were called to question, they replays were available to everyone to view! Although its not a true replay, more of a predictive approach, I do believe it is better than the human eye.

I looked into what they were using:
https://en.wikipedia.org/wiki/Hawk-Eye

In regards to Serena Williams:
2004 US Open
In her 2004 U.S. Open quarterfinal match against Jennifer Capriati, an overrule was made by chair umpire Mariana Alves in Capriati's favor, even though later video review showed this to be an error (as Williams's shot was inside the court). There were several calls that incorrectly went against Williams throughout the match, including late in the third set which prevented her from rallying from a break down. Williams attempted to argue a couple of calls, but was not successful. Capriati won the match, but tournament officials dismissed the umpire from the tournament, and she was suspended. The controversy renewed calls for the adoption of technology like the MacCam and Hawk-Eye systems. (wikipedia)

And it made me think, even though I am not a seasoned tennis watcher, I appreciated being able to see and make my own judgement on calls. It also puts pressure on the ref to make correct calls. Wouldn't a system like that be valuable to Figure Skating? We have the super slo-mo, but that seems to only be shown at events held in Japan. During the K&C, they should show the analysis for the jumps that the judges are bringing up to question right away for the whole audience, tv viewers, AND the skaters. Right now, they instead of the silly backstage K&C and the on stage K&C which frankly has no importance to me.

Who knows, it might actually make the technical sport side of Figure Skating more relevant to today's short attention span public and gain more fans.

 

[Mathman]

I think one problem is that it is really hard, at least for me, to see under-rotations and edge calls, no matter how many times I watch in super slo-motion and even in stop frame (which fans of aggrieved skaters indignantly post on You Tube for our edification).

On U.S. television broadcasts, the commentators often say things like, "Oh. that looked like it might have been short -- we'll have to see how the panel calls it." And then they show the slo-motion replay and sometimes say, "yes, that was clearly cheated," or, "it looked OK to me, within 45 degree leeway." However, the angle that the television camera shoots from is different from what the technical panel and judges see, so it is not as cut and dried as "in or out" in tennis.

 

[drivingmissdaisy]

For me, under rotations are easier to see than edge calls because the former can be seen from most camera angles. The blade has to be nearly perpendicular to the camera to accurately determine whether a skater rolled onto the wrong edge at the last moment.

 

[NaVi]

I was going to create a another site(actually a board on http://voat.co/ which is a another site cloan) and a small wiki dedicated to edge and rotation calls. It's one thing to grip about the judging system and judges, but there is no proper "culture of critique" in figure skating... no system in place that documents disputable calls... and that's the fans fault themselves.

I think tech panels should document the video frames they use to make their calls.

 

[LiamForeman]

I remember that Serena Williams match like it was yesterday. HawkEye clearly showed the shots were on the line or even worse inside the court not even close to the line. So is it possible for skating to do this? I don't think so. Too many different angles. And it's all selective anyway. When Meissner was favored for gold at USN the year she won, she clearly UR'd her combo, no question, yet Rachael Flatt got hosed with an UR and it looked within 90 degrees.

 

[CanadianSkaterGuy]

Here's the thing: whenever technology gets involved, "purists" of the sport say that it interferes with the play/tradition.

Look at baseball with the pitch visualizer that shows where each pitch goes. So if the ump miscalls a strike, it still stands as a strike, even though the technology pointed out it was a ball. But you think they'd ever replace an ump who's prone to subjectivity/error with a computer that accurately calls balls or strikes?

Same with Hawkeye; unless the player challenges it doesn't matter if the linesman calls out and Hawkeye knows it's in. If anything, Hawkeye should be a confirmation for linesman/chair umpire, so that if they make a miscall, immediately Hawkeye notifies everyone that it was a miscall.In baseball, you think there would be a reversal if an umpire called something incorrectly? They're not even afforded "challenges", except replays for stolen bases/slides etc.

Same with figure skating... disputed calls that are clearly incorrect won't even get those tech specialists a slap on the wrists. Anonymous judging certainly doesn't call out poor judges for the same reason in baseball that they don't want to disprove the referees with technology for the sake of traditionalism (and also "protecting" those judges/referees in their poor judgment calls).

 

[Layback11]

I think one problem is that it is really hard, at least for me, to see under-rotations and edge calls, no matter how many times I watch in super slo-motion and even in stop frame (which fans of aggrieved skaters indignantly post on You Tube for our edification).

On U.S. television broadcasts, the commentators often say things like, "Oh. that looked like it might have been short -- we'll have to see how the panel calls it." And then they show the slo-motion replay and sometimes say, "yes, that was clearly cheated," or, "it looked OK to me, within 45 degree leeway." However, the angle that the television camera shoots from is different from what the technical panel and judges see, so it is not as cut and dried as "in or out" in tennis.

Edge calls are very tricky to spot. URs are a little easier-if you can see their feet, you can usually see if the blade wasn't backwards-but can still be tricky. as a rule of thumb, of there's any kind of turn out of a jump, it was UR.

 

[Mathman]

URs are a little easier-if you can see their feet, you can usually see if the blade wasn't backwards-but can still be tricky. as a rule of thumb, if there's any kind of turn out of a jump, it was UR.

My problem is that the blade is constantly turning throughout the landing, and it is not always clear exactly when the blade touched down. Some skaters (like Mirai) sort of reach down with their toe-pick and others have a somewhat flatter blade on contact. To tell the truth, when I watch closely it seems like almost every jump is under-rotated, and it is just a question of how many degrees.

There is also the question of the take-off. Exactly where do you start counting the in-air rotation.

 

[gkelly]

My problem is that the blade is constantly turning throughout the landing,

It should stop the quick rotation by the time the blade touches down -- the skater should be checking the rotation as soon as the blade touches the ice, even better if they can start to check it in the air. Although that is rarely the case for ladies' triples. Watch some intentional doubles by good jumpers who can do triples, or intentional solo triple toe by men who can do quads -- you won't see much blade turning on the landings.

and it is not always clear exactly when the blade touched

That is a concern. The camera angle makes a difference in what we or the tech panel see. And the quality of the video also makes a difference -- much of what's on youtube is pretty useless for pinpointing the exact point of landing.

down. Some skaters (like Mirai) sort of reach down with their toe-pick and others have a somewhat flatter blade on contact. To tell the truth, when I watch closely it seems like almost every jump is under-rotated, and it is just a question of how many degrees.

Well, every jump except a lutz will have less than a full 360 (plus additional 360 for each additional rotation) because of the curve of both takeoff and landing edges in the same direction as the in-air rotation. On edge jumps the takeoff edge will get tighter or start to turn around just before the takeoff, and on toe loops the body starts to turn with the weight transfer from the edge to the toepick.

Lutz should be a hair more than 360(+additional) because of the change of direction.

There is also the question of the take-off. Exactly where do you start counting the in-air rotation.

That isn't clearly spelled out in the documentation.

Maybe someday we will have better technology available than video and the human eye for measuring the amount of rotation. To do something like that Hawk-Eye technology for balls, would it require cameras under the ice? sensors on the blades and/or the ice that don't interfere with the actual skating?

 

[jinhamasaki]

I remember that Serena Williams match like it was yesterday. HawkEye clearly showed the shots were on the line or even worse inside the court not even close to the line. So is it possible for skating to do this? I don't think so. Too many different angles. And it's all selective anyway. When Meissner was favored for gold at USN the year she won, she clearly UR'd her combo, no question, yet Rachael Flatt got hosed with an UR and it looked within 90 degrees.

These days programs are known well in advance. Skaters can submit to the competition around which area which jumps will take place and how. The slow-mo cameras can prepare and video the those specific jumps. Many skaters jump around the same area for all their jumps...but that's a different discussion

I remember seeing Asada's 3A in the SP at the Olympics 2010 and Kurt Browning went over the take-off and landing and calling it rotated. If EVERYONE saw on slow-mo SkaterX URed a combo and didn't receive a deduction, then that judge should be put on probation. It will force judges to do their job well! On a side note, you must remember that we here on GS can try to spot URs and edge calls, but the majority of the general public cannot.

Here's the thing: whenever technology gets involved, "purists" of the sport say that it interferes with the play/tradition.

Same with figure skating... disputed calls that are clearly incorrect won't even get those tech specialists a slap on the wrists. Anonymous judging certainly doesn't call out poor judges for the same reason in baseball that they don't want to disprove the referees with technology for the sake of traditionalism (and also "protecting" those judges/referees in their poor judgment calls).

But COP was supposed to "standardize" everything. There's hardly a standard in measuring jump rotations and takeoffs. Although the judge GOEs are anonymous, the technical callers are definitely not. I believe they are the ones with the most influence on the score. There's only 7 or 8 jump sequences compared to the hundreds in tennis and baseball.

 

[solani]

I think it wouldn't be easy to use something like Hawk-Eye (which I really appreciate) in figure skating. I think that it's using 10 cameras in the rink and only with those ten it's possible to predict and display the 3D position of the ball. We know the tennis field with it's lines and the size of the ball but we don't know exactly where on the ice rink the skater's take of and landing position is. I guess a lot of cameras would be needed and then the question would always be if the system is faultless in any situation.

 

[CanadianSkaterGuy]

But COP was supposed to "standardize" everything. There's hardly a standard in measuring jump rotations and takeoffs. Although the judge GOEs are anonymous, the technical callers are definitely not. I believe they are the ones with the most influence on the score. There's only 7 or 8 jump sequences compared to the hundreds in tennis and baseball.

COP was designed to give skaters technical credit for what they execute as well as more accurately factor in grade of execution (whereas in the past, a judge wouldn't really award a skater credit for a particular jump or spin; more like overall jumps and spins for one score out of 6.0). COP was also decided to categorize artistic impression, likely to encourage better choreography and more transitions (i.e. program content) as well as award skaters who had better interpretation or basic skating skills.

As far as measuring jump rotations and takeoffs, like baseball and tennis, if they reviewed every single thing by computer, it would take ages. Technical callers definitely are not anonymous but you hardly see backlash when they get it wrong - even in big events (like Sotnikova's 3Z(e)+3T< in Sochi that was ratified as clean).

There's no accountability in figure skating and when cheaters are actually legitimately caught (on camera even!) they come back again, judging another Olympics, a la Yuri Balkov.

 

[karne]

There are multiple factors at play here.

First: the computer can get it wrong. Very, very easily. There have been weird Hawkeye calls in tennis. The ball-tracker (formerly called Hawkeye as well) in cricket is notorious for being unreliable. I can only begin to imagine the reactions the first time a computer got it wrong - oh, especially if it happened at a Russian event, or in favour of a Russian skater!

Second: the Hawkeye-style equipment is very, very expensive. One of the reasons it took so long to have the Decision Review System (DRS) implemented in cricket was because the grounds/associations had to spend a lot of money to make sure the grounds were equipped properly with the special cameras and instruments required. And even then the DRS is not universal; India refuse to use it and won't pay the money to have it installed in their grounds.

Third: If you want the tech panel to go through slo-mos of every single element then competitions will last weeks, rather than days. Remember, if the next skater has to wait longer than 10 minutes, the whole rest of the group has to go on again for a six minute warmup. Bam, sixteen minutes lost right there on one skater. Now imagine having to go through every single element from multiple angles with some kind of computer system - especially for a long program, you're probably looking at 15-20 minutes just of reviews if it was a shoddy skate! Not every element is reviewed. The caller calls it as they see it. If the other tech panel members disagree, they call review. If none of the other tech panel members disagree in real time - and remember, we are talking about highly-trained judges who have to meet very strict requirements! - then the element will just go in as it is and not get reviewed.

 

[leoncorazon]

I actually just heard an amazing story on NPR just this am about an apartment/lab that they constructed with sensors in the floor that can detect the changes in balance and the corresponding corrections made by the ankle so that they can predict when someone (an older person) might be in danger of falling. I don't think it would be too difficult to put similar sensors under the ice. Theoretically a computer would be able to detect when the skate blade left the ice and when ice contact was re-established. Adding bars similar to what they do in the NFL to define the angle at the time of take off and landing would also be helpful. Not that we will ever see anything like this in skating but I'm sure if the powers that be wanted it, it would be possible.

 

[ice coverage]

Jump rotations were one of the parameters measured by the VeriSkate analytics software from Craig Buntin.

 

[gkelly]

Not that we will ever see anything like this in skating but I'm sure if the powers that be wanted it, it would be possible.

How much would it cost?

Keeping in mind that some competitions are held in permanent rinks that may be used primarily for hockey, and public skating, and figure skating practices but only host one or two competitions a year. Major championships are held in arenas that may put down ice only for that event and take it out the following week. And an Olympic-sized ice surface is about 20,000 square feet, significantly more than a typical apartment or even an average-sized house.

If it becomes hardly more expensive to install the sensors than to install ice without them, and if they could stay there and stay operational during months of use between championship, then I think we'll start seeing that become standard practice. If it would add hundreds of thousands of dollars to the cost of every competition, it's not going to happen.

If the relevant technology gets developed and widely adopted for other purposes, that would eventually bring down the cost for skating purposes.

 

[Sam-Skwantch]

It would have to be beneficial to hockey games first and foremost. I'm sorry but Figure skating brings in very little money to ice rinks and the only technology they will develop will be one that benefits them. Besides this could only work if someone like the NHL or Europeon and Canadian Elite leagues could find value in it. I've seen ice rinks built from zero ice to 2inches thick and I just don't think there is any possible way.

It is however a nice thought but I think the ISU could invest in camera's that utilize 3d technology and it would be cheaper, transportable and just as efficient. I also don't think we really need to be a sport of inches and millimeters. Let the judges judge!

 

[ice coverage]

How much would it cost? ...

If the relevant technology gets developed and widely for other purposes, that would eventually bring down the cost for skating purposes.

It would have to be beneficial to hockey games first and foremost. ...

Agree.

Not sure if I have all the details 100% straight, but my understanding is that Craig Buntin created two companies based on much the same technology: VeriSkate was for figure skating analytics, and his other one was for hockey analytics.

Mark Cuban, no less, invested big bucks in Craig's company for hockey analytics, which seems to be thriving.

In contrast, VeriSkate seems to have gone dormant, after what seemed like a successful introduction at 2014 Skate Canada. I hope that some day it will get sufficient financial backing to be revived.

 

[Sam-Skwantch]

Just an out loud thought but with the next Olympics slated for Korea maybe Samsung would find a benefit in developing such technology and using the Games as a platform to advertise not only it but their brand.

I'm sure there is something beneficial to Figure Skating replays that could be beneficial to other Winter Games events as well.

#BRUGHTtoYOUbySAMSUNG

 

[Barb]

Hawk Eye is different than Phantom camera, right?, because I think it is not counting the lines, distance, etc but you can see perfectly everything, I guess it is not so costly like Hawk eye https://www.youtube.com/watch?v=vCZhiYp6sEM