Could hawkeye be deceptive?

ac3111

Professional
Chair umpire says to Andy Roddick in UsOpen "Shadow is bigger than the mark... hawkeye makes it look bigger"...
And I was thinking that a part of the ball could be over the line but not actually touching it. Will hawkeye show it as in?
 

jelle v

Hall of Fame
I have always wondered how the Hawk Eye system translates the bounce of the ball to a shadow, how it was even possible to do it accurately. :-?
 

naylor

Semi-Pro
The only way that kind of comment makes sense is if the mark (the little visible imprint the ball leaves on the court when it bounces) only represents part of the ball the touches the ground - the inner circle, so to speak. Whereas the shadow is the whole half of the ball that does touch the ground as the ball flattens on impact. In other words, when the ball bounces it squashes flat against the ground, and hawkeye reads the bounce as a complete hemisphere touching the ground - and hence catching the line (or not) - whereas when you look at the visible imprint it's smaller than that.

You get the same idea if you soak a tennis ball and then you bounce it. The wet mark you can see on the ground is smaller than the half-section of the ball.

The question then becomes, how much does a ball actually squash flat, and therefore really touches the ground. I can believe that if a ball hits the netcord hard and just trickles over the net, the real bounce / squash mark will be quite small - and yet when you see the shadow on hawkeye it still looks like the full hemisphere. Whereas the squash mark produced by a full blown smash will indeed be close to the shadow of a "half ball" as the ball will truly flatten.
 

roysid

Hall of Fame
Well said. The Hawkeye system is wrong. Sometimes you see a line of the shadow touching and that's counted in. Whereas it clearly bounced outside.

They should count the shadow atleast 10% in before marking it in.
 
I think the problem is everybody looks at the moment the ball is at it's deepest point, thus when the ball bounces back up. Yet hawkeye looks at where the balls makes first contact, thus touching the line a little before the ball is at it's deepest point. In other words the ball skids on the court before bouncing. It depends on how you look at it, to me this way of looking at the point is wrong. I think linesmen always have looked at the bounce not the few hairs on the ball that touch the line. However this doesn't mean Hawkeye makes mistakes, the system only gives the representation asked for by the programmer. In the end it all depends on what you choose.
I can live with this system it is a lot better than some of the dubious line calls we have seen in the past. My only criticism is that it should always work, and not fail during a match. All the players should be able to use their challenges on all points. Otherwise it is not fair.
 

shadows

Legend
Chair umpire says to Andy Roddick in UsOpen "Shadow is bigger than the mark... hawkeye makes it look bigger"...
And I was thinking that a part of the ball could be over the line but not actually touching it. Will hawkeye show it as in?

You misunderstand the meaning of what the umpire said. The shadow is the whole ballmark, the issue is that on a hardcourt only the areas of the ball that hit the surface hardest leave some kind of a mark. It's why on soft clay the ballmarks are bigger.
 

DNShade

Hall of Fame
Not to mention guys, that Hawkeye isn't really "seeing" anything. It's just guessing - an educated guess, yes - but a guess none the less. And yes, the mark left on a hard court is smaller than the amount of ball that actually touches ground.

But hey, it's much better than just people guessing.
 
Last edited:

naylor

Semi-Pro
... hawkeye looks at where the balls makes first contact... the ball skids on the court before bouncing.... linesmen always have looked at the bounce not the few hairs on the ball that touch the line. However this doesn't mean Hawkeye makes mistakes... I can live with this system it is a lot better than some of the dubious line calls we have seen in the past. My only criticism is that it should always work, and not fail during a match...

Agreed entirely, provided the system operates consistently, it's a lot easier to live with than linesmen's calls - particularly, if they just "watch the line", rather than the ball trajectory / pace / spin through to its ultimate landing point.
 

West Coast Ace

G.O.A.T.
I believe it's plus or minus 2 mm...

The players, and what a shock (not!) that Roddick complained, need to just accept it. The system isn't perfect - but it's better than the alternative. I wish before a match the chair would tell the players (they should already know): "I will only overrule if I'm sure there was a mistake. Otherwise it's up to you to use a challenge. Don't ask me what I thought." They also need to go back to the 'challenge immediately' rule and not let guys wander up to the net and try to shop for a mark.
 

MizunoMX20

Rookie
The only way that kind of comment makes sense is if the mark (the little visible imprint the ball leaves on the court when it bounces) only represents part of the ball the touches the ground - the inner circle, so to speak. Whereas the shadow is the whole half of the ball that does touch the ground as the ball flattens on impact. In other words, when the ball bounces it squashes flat against the ground, and hawkeye reads the bounce as a complete hemisphere touching the ground - and hence catching the line (or not) - whereas when you look at the visible imprint it's smaller than that.

You get the same idea if you soak a tennis ball and then you bounce it. The wet mark you can see on the ground is smaller than the half-section of the ball.

The question then becomes, how much does a ball actually squash flat, and therefore really touches the ground. I can believe that if a ball hits the netcord hard and just trickles over the net, the real bounce / squash mark will be quite small - and yet when you see the shadow on hawkeye it still looks like the full hemisphere. Whereas the squash mark produced by a full blown smash will indeed be close to the shadow of a "half ball" as the ball will truly flatten.

I'm not sure if I understood what you meant but I think you are wrong. I don't think the ball mark has got anything to do with the half-section of the ball, and neither has Hawkeye. It's clear that when you for instance hit a hard, flat serve, the ball mark will be elongated. This means that if any part of this elongated mark hits the line, the ball is in. On clay this can be seen best:

istockphoto_6384734-tennis-ball-mark.jpg


Contrarily, when you hit a really soft shot, the ball mark will be really small and round (not nearly as big as the half-section of the ball, even though it may still have the same shape as the 'half ball'). It's the same when a ball is just lying on the ground, the ball is touching the ground, but in a very small spot. To put this in perspective, Hawkeye would show this very small contact spot as the 'shadow', rather than the shadow of the half ball.
 

Ratsoup

New User
I'm not sure how the hawkeye works exactly, but I'm pretty sure, that it follows the balls bottom trajectory. Meaning, that even if the ball bounce mark on the screen is wider and kind of "manipulated via the computer" and not as the the "real" mark, as it would leave on clay court for example, it doesn't make much difference because the place where the bottom trajectory crosses the ground, is the point what defines if the ball was in or out. Wow, that was a long sentence.

Another thing is the calls which the hawkeye makes for the sidelines tho. And that is more tricky, because the ball's elastic properties. I'm not sure actually how to the rules define exactly when the ball is out or in, (Meaning, if the mark of impact - the area that the ball touches the ground) marks the spot, or lets say that the "in-area" of the court would be higher, then if the ball touches the court with a piece of fur, it's still in?) But I'm quite sure that the hawkeye system is accurate enough to be in use, and players should accept the outcome. Yeah, it can be wrong a little, but since the computer doesn't have a personality, it's always doing "mistakes" equally to both players.
 

MizunoMX20

Rookie
I'm not sure actually how to the rules define exactly when the ball is out or in, (Meaning, if the mark of impact - the area that the ball touches the ground) marks the spot, or lets say that the "in-area" of the court would be higher, then if the ball touches the court with a piece of fur, it's still in?)

A ball is called in if it hits the court (+lines) with ANY part of the ball, so in your case, the would be in.
 

Gimmick

Semi-Pro
Of course hawkeye is wrong. The most obvious example is the failure to calibrate hawkeye to the barometric pressure's effect on the ball; not to mention the more dramatic change in ball pressure (compression when it strikes) when new balls are put in; nor does it compensate for balls sometimes fluffing up and changing diameter.

When hawkeye starts adapting to atmospheric conditions, new balls, humidity fluffing of felt, when it works all the time and doesn't fail in the middle of a match like it did this week, and when they show the margin of error as part of the shadow............then hawkeye may be mildly accurate.
 

jrod

Hall of Fame
Of course hawkeye is wrong. The most obvious example is the failure to calibrate hawkeye to the barometric pressure's effect on the ball; not to mention the more dramatic change in ball pressure (compression when it strikes) when new balls are put in; nor does it compensate for balls sometimes fluffing up and changing diameter.

When hawkeye starts adapting to atmospheric conditions, new balls, humidity fluffing of felt, when it works all the time and doesn't fail in the middle of a match like it did this week, and when they show the margin of error as part of the shadow............then hawkeye may be mildly accurate.

Like humans have a shot at doing this???

Do you have any idea as to how much reduction in error would be achieved if the Hawkeye system actually did what you are suggesting?
 

raiden031

Legend
Hawkeye is the best option available. Until something better comes along, no point in nitpicking that it is less than 100% accurate.
 

sureshs

Bionic Poster
I would just use some sensor chips inside each ball to detect impact and report its coordinates wirelessly to a base station.
 

THESEXPISTOL

Hall of Fame
I would just use some sensor chips inside each ball to detect impact and report its coordinates wirelessly to a base station.

That's not a good ideia..
Look at then number of balls spend during the USOPEN on the HawkEye courts and try to imagine the cousts of that.
And then think about the change in the balls specs.
 

Cindysphinx

G.O.A.T.
Hawkeye is the best option available. Until something better comes along, no point in nitpicking that it is less than 100% accurate.

Yeah. We gotta use the best technology we have until something better comes along.

I mean, can you imagine if you hurt yourself and your doctor said, "Well, MRI and X ray aren't 100% accurate and can't see everything, so we're not going to bother with them. Now bite on this stick while I poke around with my fingers to take an educated guess."
 

raiden031

Legend
Correct...how is a 2mm reduction in error possible by taking into account atmospheric conditions?

I don't know if it is or not. The system is already accurate with a small enough margin of error as it is. So really there is no point to bother with atmospheric conditions.
 

jrod

Hall of Fame
I don't know if it is or not. The system is already accurate with a small enough margin of error as it is. So really there is no point to bother with atmospheric conditions.

Agree, it is quite accurate now. I was impressed with it's unofficial (broadcasters only) use at the French Open this year, even though they were using less cameras causing the average error and standard deviation of the error to increase.

As for the atmospheric conditions, the only significant impact I can think of that could be included is when it rains the fuzz on the ball is reduced somewhat(your 2 mm guess was probably pretty close!). I suppose when they start playing tennis in the rain on a regular basis Hawkeye could factor this compression in...
 

pmerk34

Legend
Yeah. We gotta use the best technology we have until something better comes along.

I mean, can you imagine if you hurt yourself and your doctor said, "Well, MRI and X ray aren't 100% accurate and can't see everything, so we're not going to bother with them. Now bite on this stick while I poke around with my fingers to take an educated guess."

The players have accepted the system (well Federer still hates it LOL) and I have seen a marked decrease of pointless arguing with the umpire over line calls the umpire doesn't even make. Plus we (the fans) like it. It's close to perfect as we have now.
 

mtr1

Professional
Surely if a ball has loads of topspin on it, it will touch a tiny amount of court. If it has slice on it, then it might "squash" and look how hawkeye shows almost every shot. The ball marks shown by hawkeye are far too large, and there have been occasions when repleys have shown this. A point in the Williams-Dementiava Wimbledon semi final comes to mind.
 

jrod

Hall of Fame
Surely if a ball has loads of topspin on it, it will touch a tiny amount of court. If it has slice on it, then it might "squash" and look how hawkeye shows almost every shot. The ball marks shown by hawkeye are far too large, and there have been occasions when repleys have shown this. A point in the Williams-Dementiava Wimbledon semi final comes to mind.

I seriously doubt any human can account for the area of the ball that touches the ground in real-time or even slo-mo. On the other hand, the image projected by Hawkeye does apparently account for trajectory and predicts an impact region based on the assumed properties about the ball's construction, speed, trajectory, etc. The quoted average accuracy is about 5% of the ball's diameter, or ~3.5 mm and presumably accounts for the compression variation among other things.

That said, I still am waiting to see a distribution of the error as quoting the average error is not nearly as informative as providing it's distribution.
 

naylor

Semi-Pro
... I don't think the ball mark has got anything to do with the half-section of the ball, and neither has Hawkeye... when you ... hit a hard, flat serve, the ball mark will be elongated. This means that if any part of this elongated mark hits the line, the ball is in. On clay this can be seen best:
istockphoto_6384734-tennis-ball-mark.jpg

... when you hit a really soft shot, the ball mark will be really small and round (not nearly as big as the half-section of the ball, even though it may still have the same shape as the 'half ball'). It's the same when a ball is just lying on the ground, the ball is touching the ground, but in a very small spot. To put this in perspective, Hawkeye would show this very small contact spot as the 'shadow', rather than the shadow of the half ball.

I agree entirely with your points, namely, a hard-hit ball should show a much larger, elongated mark than a soft dropshot (the shadow for the soft drop would only be fractionally larger than that of a ball lying at rest on the place the dropshot landed). My point is that you don't see that when they show the hawkeye pictures (particularly, I have never seen a hawkeye shadow looking like the huge clay smugde you included in your post) - whether they show you a hard serve or a soft dropper, the area of the shadow looks very similar on both (a bit more elongated for the serve, quite round for the drop, but that's all). But, as you yourself put it (which I agree with), the shadow of a soft ball ought to be a lot smaller also.
 
W

woodrow1029

Guest
I agree entirely with your points, namely, a hard-hit ball should show a much larger, elongated mark than a soft dropshot (the shadow for the soft drop would only be fractionally larger than that of a ball lying at rest on the place the dropshot landed). My point is that you don't see that when they show the hawkeye pictures (particularly, I have never seen a hawkeye shadow looking like the huge clay smugde you included in your post) - whether they show you a hard serve or a soft dropper, the area of the shadow looks very similar on both (a bit more elongated for the serve, quite round for the drop, but that's all). But, as you yourself put it (which I agree with), the shadow of a soft ball ought to be a lot smaller also.
The image on the Hawkeye screen is only for a visual that the ball is actually in or out. It is not a true representation of the compression of the ball on the court. Sometimes when it is only 3 mm out, the space will look much larger that. Sometimes when the ball is 2 inches out, the space will look much smaller. The point they are getting across is that it is in or out.
 

socal29

Rookie
Hawkeye is like a paternity test, it is ALWAYS right. Commentators who argue even after the result comes in are so annoying...they just can't admit they were wrong.
 

DNShade

Hall of Fame
That's not a good ideia..
Look at then number of balls spend during the USOPEN on the HawkEye courts and try to imagine the cousts of that.
And then think about the change in the balls specs.

Actually not that that expensive at all. Casinos are trying this out with each playing card on the table to track play,betting and payouts and to make sure they match up...think about the amount cards a major joint on the strip goes through. And Hawkeye isn't all that reliable. So someone will come along here with something that actually tells us where the ball really lands as opposed to just a good guess.
 
Last edited:
Chair umpire says to Andy Roddick in UsOpen "Shadow is bigger than the mark... hawkeye makes it look bigger"...
And I was thinking that a part of the ball could be over the line but not actually touching it. Will hawkeye show it as in?

really dumb comment by roddick. the ball flattens on impact, but the outer part of the ball does not touch the court with the same force, thus leaving no marks.
 

crash1929

Hall of Fame
op good question never thought about this one...

i was wondering why the players have to challenge though...cant the umpire just watch all the shots on hawkeye and then over rule as needed. is the tech that slow?
 

Deuce

Banned
I don't know if it is or not. The system is already accurate with a small enough margin of error as it is.
^ Says who, exactly? The makers of HawkEye? By the ATP or WTA?
Kind of like Ford saying that their cars are better than the competition - it's marketing spin.
I'll start considering the results of tests when, and only when, they are done by an objective third party.

The image on the Hawkeye screen is only for a visual that the ball is actually in or out. It is not a true representation of the compression of the ball on the court. Sometimes when it is only 3 mm out, the space will look much larger that. Sometimes when the ball is 2 inches out, the space will look much smaller. The point they are getting across is that it is in or out.
^ If this is true, then it is deliberately deceiving people. Otherwise, instead of the nice little cartoon, they would show only the word 'in' or 'out'.
If what you say is true, then they are just playing with people - deliberately manipulating the situation for the sake of 'entertainment'.
If all they are doing is conveying the message that the ball was in or out, that's what they should do - without the artificial cartoon theatrics.

crash1929 said:
i was wondering why the players have to challenge though...cant the umpire just watch all the shots on hawkeye and then over rule as needed. is the tech that slow?
^ It's part of the 'entertainment package' that HawkEye/ShotSpot brings. The challenges create a 'drama' (artificially, of course), and even affect the game.
It would be considerably more boring if the umpire simply had a private indication that the ball was in or out, and then announced it. There ain't no 'entertainment' in that... so they created a little 'entertainment package' (the challenge system) which they hope will keep the fans more interested.

HawkEye/ShotSpot is simply a slick marketing tool that the ATP and WTA are using to take advantage of people's attraction (addiction?) to 'slick technology' and anything involving moving video images, with the intention of increasing interest in the game as a result.

It's a toy. A video game. That's why people are attracted to it.
Well - that and the over-the-top marketing that it has received - starting a few years ago when virtually all the tennis commentators on all the TV channels were praising HawkEye/ShotSpot as if it were the second coming of Christ.
Anyone who doesn't think that all that marketing was all prearranged and very rehearsed is quite gullible.
 
Last edited:

LanceStern

Professional
It's not always right.

Sometimes you see a shot with the regular camera and hawkeye says that same shot is a few inches away from that point. Especially on serves or shots close to the wide baselines.
 

lawrence

Hall of Fame
solution? implement hawkeye at the FO, just for the sake of comparing ball marks to hawkeye results

the most practical future solution i see happening, would be for hardcourts to be re-created with a new surface that is purposely easily marked by impact.. you'd have a lot of shoe marks but you'd have the ball marks to confirm ins and outs too. kind of like clay
as for grass? nfi.
 

nomie

New User
Why change the whole court surface? Only the lines needs to be pressure sensitive / easily marked.
 

jrod

Hall of Fame
^ Says who, exactly? The makers of HawkEye? By the ATP or WTA?
Kind of like Ford saying that their cars are better than the competition - it's marketing spin.
I'll start considering the results of tests when, and only when, they are done by an objective third party....blah blah blah........


You obviously have absolutely no idea as to what you are talking about. Before you sound off like this do some research. Try reading the original patent filed by the inventors of the system. Assuming you can even comprehend the basis for the technology, then we can start to have a meaningful conversation.

The systems accuracy has been established and is generally accepted. In addition, 3rd party assessments have been conducted by the ITF prior to its adoption. They also used it at the FO this year for the broadcasters and employed less cameras than when used for line call challenges. Even with less cameras employed the accuracy was quite impressive.

As for the marketing hype, what did you expect?
 

jrod

Hall of Fame
Actually not that that expensive at all. Casinos are trying this out with each playing card on the table to track play,betting and payouts and to make sure they match up...think about the amount cards a major joint on the strip goes through.

The technology used in casinos is in no way related to hawkeye, aside from the use of cameras. The requirements are entirely different.

And Hawkeye isn't all that reliable. So someone will come along here with something that actually tells us where the ball really lands as opposed to just a good guess.

What do you consider to be reliable?
 

jrod

Hall of Fame
The image on the Hawkeye screen is only for a visual that the ball is actually in or out. It is not a true representation of the compression of the ball on the court. Sometimes when it is only 3 mm out, the space will look much larger that. Sometimes when the ball is 2 inches out, the space will look much smaller. The point they are getting across is that it is in or out.

Precisely.
 
Top