Thanks!
Speaking of which, watch this rally in table tennis.
The table is clearly too small for their game.
Speaking of which, watch this rally in table tennis.
The table is clearly too small for their game.
-
-
So some contary information. Just got back from speaking at Vegas at USPTA National. My friend Dave Ramos from USTA was there. They have complete access to all shot spot data and a lot of inside info on the system. Dave does not believe it is accurate above 115mph or 3000rpm.Comment
-
Very interesting. So, we're supposed to accept that ELC is accurate to millimeters -- but can't measure velocity and spin at pro levels?
Bring back the lines people ! < g >
Again, I have not seen any, independent verification of Hawkeye, Foxtenn only their own manufacturers' claims. Might exist but if so, one would expect the vendors to tout it. For serves, at least, hasn't it been correlated with the radar guns? Surely radar works above 115 mph.Comment
-
And the posters on the forum like to agree with each other a lot. My partner at the club that I used to work at here in Sweden was in love with his speed measuring devise and I absolutely hated it. His idea of spending constructive time was measuring how fast the group could register their serve on the device. I thought that this was an absolute waste of time and would rather try to iron out kinks in motions...which ironically would produce more speed with less effort. The students thought that the speed gun device was a gas as well as they knocked themselves out trying to out do themselves.
stroke likes to tally the rpm's on the forehands as if this is some sort of indicator of a player's capacity. I guess it is some sort of indicator but of what I am not certain. Fyodor Dostoyevsky had an interesting comment on the science of his day that I find eerily true today in modern times. He called it the "half-science" and I leave it up to you to figure out what he meant by it.
"Reason has never been able to define good and evil, or even to separate good from evil, not even approximately; on the contrary, it had always mixed them up in a most pitiful and disgraceful fashion; as for science, its solutions have always been based on brute force. This was particularly true of the half-science, that most terrible scourge of mankind, worse than pestilence, famine, or war, and quite unknown till our present century. Half-science is a depot such as has never been known before. A despot that has its own priests and slaves, a despot before whom everybody prostrates himself with love and superstitious dread, such as has been quite inconceivable till now, before whom science itself trembles and surrenders in a shameful way."
Doesn't this remind you a bit of what we see to day when you have the masses of idiots reminding everyone to "follow the science" when in reality all you are witnessing is a stupid game of follow the leader (idiot)?
Statistics can be all fun and games for some. I used to work in quality control in a testing lab. One element of statistics that is not discussed very often is "fudging the numbers". Not that anyone would fudge the numbers in your article and analysis of 400 million points in tennis competitive play. Statistics have been used in excess trying to analyse the phenomenon of "climate change" or what is it they are calling it now. It used to be "global warming". The statistical models used to play the COVID-19 game has been absolutely diabolical. You cannot seem to trust a single number that they are throwing out there. None of it. So, I am in league with Dostoyevsky here. I'm skeptical. Experts? That's another good one. Who's an expert? Most experts are self declared until they can muster a following and then it almost becomes a cult. Take Shaun for instance. A professor of ethics at the local university here told me that the "global warming" crowd had nearly become a cult.
I understand that science plays a role in the game and sport of tennis. I guess I am more of a romantic than scientist when it comes to tennis. I think the scientists got a hold of her somehow and look how they have worked her over. Just following the science, they will tell you.
The one "scientific breakthrough" that I really think was a game changer is video. The johnyandell advocacy of "visualisation" is a valuable teaching process.
don_budge
Performance Analysthttps://forum.tennisplayer.net/images/smilies/cool.pngComment
-
I've written too long a post here, but I'm trying to think through the implications.
I don't know if this is too far off target for this discussion board, but it's perplexing to me.
Count me a Hawkeye skeptic for line calling, so part of me feels validated by that.
Still I find it hard to believe tennis would be moving to electronic line calling (ELC) if it knew about inaccuracies as large as this implies.
While stats are nice to have, the job of line calling is critical to the sport.
Line calling by the tennis version of Hawkeye relies on creating a simulated impact mark. It's not measured {as competitor Foxtenn apparently tries to} it is generated by the computer from models for the given angle of impact, speed & spin. {I do NOT claim to know Hawkeyes' set up nor physics in depth.}
Modeling the ball impact seems more difficult, and has far more variables, than simply measuring speed.
If -- Big IF-- Hawkeye can't accurately measure speed and spin at pro levels, I can't see how it can generate an accurate simulation of the impact mark. As one example, a much faster moving and more rapidly spinning ball is highly likely to create a different shape (and larger) impact mark than a slow moving one.
As John' TPN slow motion video shows, the movement of a tennis ball is far more complex than that of a solid sphere; it's more like a wax blob in a lava lamp
As I posted here, the differential between the highest groundstroke velocity numbers at Roland Garros and the years of data from the Australian Open was roughly 10% for many players' data.
It seems improbable Hawkeye could consistently over-estimate velocity by that much for an entire major and still call the lines accurately anywhere. Even a simple radar gun for baseball pitchers is said to be accurate to +/- 1%. It would be socking if the ITF/ ATP converted to ELC while it knew there was a 10% difference between their radar guns' measurements and Hawkeye visuals.
Major League Baseball is far more transparent with its metrics and how they work than pro tennis is. MLB has moved to a more complex version of Hawkeye for it's "Statcast" systems, which are installed in every ballpark. From the leagues' posts, they can locate the center of a baseball to within 0.25 inches, or about the width of a pencil. That seems roughly comparable to the claimed 3.6 millimeter accuracy of Hawkeye line calling in tennis, although that is, again, more complex to do.
https://technology.mlblogs.com/intro...d-a5f5c20321b8
MLB is sharing data with television that includes home run speed down to tenths of a MPH as in "the Statcast record for exit velocity is 122.2 mph by Stanton on Oct. 1, 2017". Would MLB risk its credibility if it knew that number could be off by 5-10%?
There is also the issue of setting up and operating Hawkeye properly. Players have complained loudly about Hawkeye at smaller events. But Roland Garros is a major. So, presumably they had a competent tech team.
I'd love to ask someone with an overview what's up, but as we all know tennis a headless beast.
Comment
-
Who knows? The ball flight impact error for line calls is supposed to be about 3mm. But measuring speed and spin are different from the impact. My friend at the USTA accepts the 3mm bounce error. It's speed and spin at higher levels he doubts.Comment
-
Who knows, indeed, which is part of the issue.
Perhaps it's my lack of imagination, but I don't see a way to calculate a computer-simulated impact 'mark' correctly while relying on incorrect speed and spin. Get those wrong, the shape is wrong. The mark could be inaccurate even when the speed and spin are fairly accurately measured. But the reverse seems near impossible.
Remember, the Hawkeye cameras are well above the ball, looking down and can't measure the impact mark visually -- unlike Foxtenn, which puts cameras right on the lines.
One thing I like about the competitive Foxtenn system is that fans get to see the actual video of impact with the replay. Occasionally, the eye says the Foxtenn call is clearly wrong. If multiple times her match is your version of "occasionally".
Regardless, I'd prefer the lines people back with all the tension and drama they bring.Last edited by jimlosaltos; 09-22-2021, 10:28 AM.Comment
-
It never ceased to amaze me how many times that the electronic gizmos confirmed the call of the linesperson. I might be interested in that particular statistic if it were available. Linespeople are probably not coming back. Once a step in this direction takes place it rarely if ever reverses. Replacing people with robots. With AI. Trending.
don_budge
Performance Analysthttps://forum.tennisplayer.net/images/smilies/cool.pngComment
-
Comment
-
I have never trusted Hawkeye because I have seen balls land out that have been called in by the wretched machine and vice versa. I know because I have sat there at SW19 right next to a line and seen the mistakes with my own eyes...and I do have good eyesight. Roger was very reluctant to accept Hawkeye when it was first introduced but he eventually came to do so. On balance, however, Hawkeye is considerably better a human at calling lines. Linesman are under pressure calling lines in big matches and that's where the problem lies rather than human eyesight not being good enough.
As for measuring the speed of a serve, I believe the equipment used to do that is sometimes inaccurate. I have seen fast serves register a low mph and slower serves given a fast mph. Again, on balance, there is some accuracy but it is by no means infallible.
The bottom line? Never trust a machine...not a 100% anyway.
StottyComment
-
During the recent US Open, when the crowd was noisy, like on a big point with "Oohs" and "Aaahs", I definitely missed the confirmatory "In" hand signal from a linesperson. Just the absence of an automatic "Out" call (which may have been drowned out) led to several occasions where noone was sure who had won the point.Comment
-
"Never trust a machine". I think that's been the theme of more than a few Sci Fi books and movies
Since you mentioned Fed's mistrust. I recall Fed-Nadal Wimbledon, early days for Hawkeye, an example of how that system creates a "simulated ballmark" Both players were at the net and Nadal hit a soft floater that was just out of Fed's reach on top of the net, and was called on the line. Fed challenged then had a snit when Hawkeye confirmed the call.
The "ball mark" from Hawkeye was a long, thin line as if the ball came in hard and at an extremely low angle, skidding along the court -- when it actually fell almost straight down near the net. Since the long, thin front skid mark touched the line HE ruled it in. It's almost dishonest to show those in video as if they are real ball marks.
That's one psychological reason I prefer Foxtenn as a viewer. You see the low camera angle along the line along with the call. I've been in long threads elsewhere, which I won't repeat, but I've seen four captures of those videos where the naked eye can see the ball clearly out in the Foxtenn review screen, while being called in. One of those was on clay in a final with Ash Barty. Ash didn't swing, was looking straight down at her toe as the ball landed in front of it past the line and immediately pointed at that mark. It was called out but over-ruled by Foxtenn when Sabalenka challenged.
In her Foxtenn challenge, you can see, in slow motion, the ball traveling over the baseline, which is obscured by clay, then landing right in front of her toe, a few inches out.
But at least we got to see reality even if the call stood.Last edited by jimlosaltos; 09-24-2021, 08:11 AM.Comment
-
Players were sometimes confused because they couldn't hear the call and there was no hand signal.
The crowd response to a lines call, and the instant of tension as you wonder if the chair will overrule, then if the player will challenge adds drama. Eliminating that is a mistake.
Pro tennis needs to be speeded up, but by eliminating the boring parts, not the ones that create drama.Comment
-
don-budge, searched but couldn't find an overall figure. Perhaps this helps. From 2016, "David Goffin is the most successful player to have called for Hawk-Eye challenges according to a study conducted by ATP. The world no. 15 has a success rate of 44% . The study includes tennis players, who have used Hawk-Eye at least 100 times in the last 10 years... As for the top players, Novak Djokovic is sixth with 35%, Roger Federer is ninth (218 won and 404 lost). Rafael Nadal 22th with 32%."
https://www.tennisworldusa.org/tenni...vic-and-nadal/Comment
Comment