Friday, July 14, 2006

Technology in Tennis: Hawk-eye

With the advent of technology, its integration into sports has come at an astonishing pace. We have seen technology used to aid in sports training (heart rate monitors, wattmeters, training programs etc), improve the field of play (artificial playing surfaces, covered/indoor stadiums, etc), enhance sporting equipment (tennis racquets, synthetic material for basketballs, hockey sticks, etc) and expand media coverage (newspaper, radio, television, internet etc).

One other area where technology has played a large role is that of officiating. Different sports have embraced technology at various levels. This difference can be seen between sports like baseball or soccer, where there is minimal technology used to aid in officiating, to sports like football and hockey, where video replay can be used to aid officials.

Over the last few years, tennis has been one of the sports that faced pressure from the public to adopt technological aid to assist in line calling. This pressure has been growing as a result of the increasing speed in the game, along with the increasing number of controversial line calls by officials. Perhaps the breaking point in this came in the form of the 2004 women's U.S. Open quarterfinals match between Serena Williams and Jennifer Capriati. In the third set of that match, there was an incorrect chair reversal of a call made by the linesperson which eventually cost Williams the match against Capriati. The mistake was acknowledged when the tournament's head of officiating removed that official from further matches and personally called Serena Williams to apologize for the bad call.

In March of 2006, tennis introduced technological aid in officiating in the form of instant replay technology at the Nasdaq-100 Open in Miami. Following the great feedback from players and fans, the use of electronic line calling will be expanded to include Sony Ericsson WTA Tour and ATP events participating in the U.S. Open Series. This will culminate in the debut of electronic line calling at the U.S. Open, a first for a Grand Slam event.

How does Hawk-eye work?

Hawk-eye uses multiple cameras placed around the court to track the players and the ball. These movements are then processed by computers. Hawk-eye takes ball skid and ball compression into account and is accurate to 2-3 mm.

A visual description of the Hawk-eye system (number of cameras may vary depending on setup)

It is important to note that Hawk-eye isn't entirely new; however, it has been refined over the years to such a degree that the confidence level in the system is high enough for use as an officiating aid.

The use of this tracking technology also allows for a new range of statistics to be captured and presented to the viewing audience; something that many tennis viewers may already be familiar with. In addition to calling lines, statistics that Hawk-eye can measure include speed of the ball at any point of a rally, service comparisons (i.e. service patterns, direction and depth of aces, placement of 1st and 2nd serves etc), bounce points of the ball, percentage of time a player spends in a region of the court or an approximate measure of distance that a player has run throughout the match.

Statistics from the 2005 Wimbledon Final between Roger Federer and Andy Roddick. We can see Federer targeting the Roddick backhand side on second serves. Also note the placement of Federer's aces on Roddick.

As part of this system, new video boards will be a large part of the in-stadium atmosphere on centre court for instant replay coverage, statistics, fan interaction, contests and corporate partner recognition. You may even see players sneak a peak at their stats during changeovers, in order to make in-game adjustments. The television broadcast is also enhanced by this system, primarily through the usage of the various camera vantage points around the court, and the display of statistics during the course of the match.

With all this technology in place, there is always an associated cost. The system costs approximately $20,000 to $25,000 USD for a court per week (excluding the video screens in the stadium). As a result, in Toronto's Rexall Centre and Montreal's Uniprix Stadium, this system will only be set up on the main court. Advertising opportunities will help to defray these costs, as partners can associate themselves with this technology (think of the Hawk-eye brought to you by "..." or these Stats are brought to you by "...").

Changes to the game

At all U.S. Open Series tournaments, the on-court instant replay system can be used by the players in the form of a challenge system. The challenge system will work in the following manner:

  • Each player will receive two challenges per set to review line calls.
  • If the player is correct with a challenge, then the player retains the same number of challenges.
  • If the player is incorrect with a challenge, then one of his/her challenges is lost.
  • During a tie-break game in any set, each player will receive one additional challenge.
  • Challenges may not be carried over from one set to another.

Once a player challenges a line call, an official replay will be provided simultaneously to the television broadcast and in-stadium video boards, allowing players, officials, on-site fans and television viewers the opportunity to see the live results of a player challenge.

This serve is out

Reaction to this system

Most players have embraced this system. Perhaps the most notable exception is Roger Federer. "What is happening is madness," Federer said of the decision to use the Hawk-Eye computer system on the main court for the Nasdaq-100 Open. "A pure waste of money." Perhaps with time and/or further refinement of the system, he will come around.

Here are reactions from other players and broadcasters.

Andre Agassi - tennis player

In my 20 years in professional tennis, this is one of the most exciting things to happen for players, fans and television viewers. This new technology will add a whole new dimension to the game.

James Blake - tennis player

The ball's moving so fast these days that sometimes it's impossible for anyone to see, even a trained official. With instant replay we can take advantage of technology and eliminate human error. Having just a few challenges will make it both fun and dramatic for fans at the same time.

Jamea Jackson - tennis player (first to make a challenge)

It takes a lot of pressure off. You don't get so angry. If you think a call is incorrect, you don't spend extra games thinking about it. It's really quick. I remember people were complaining about maybe it throwing off the timing and rhythm of the match, but it didn't do anything like that at all.

John McEnroe - former tennis player, current broadcaster

If anyone's been listening to my commentary the past year then they know I'm in favour of using replay. I think it will make tennis more interesting.

Cliff Drysdale - former tennis player, current broadcaster

I'm thrilled about it because tennis needs it for a variety of reasons, one of which, the main one being the viewing public, the second being the live audience and thirdly just because we don't need more repetitions of what we've seen over the past year with people getting hooked out of matches.

I'm looking forward to seeing this system work at this year's Rogers Cup tournament in Toronto. In the future, I'd like to see them go away from a challenge system. Instead, all close calls should be reviewed, since the results are available to the chair umpire pretty quickly. Regardless, this is definitely another positive step forward for tennis.

Technorati Tags:
, , , , ,


Anonymous said...

looks interesting..

Anonymous said...

Are you sure this is a good idea?
I'm surprised at McEnroe's comments. If we had replay back then, we would never have seen his memorable tirades against the umpires. I think I can safely say that those tirades are more exciting than having a computer spit out a result. This just takes the emotion out of the game. Maybe Federer is on to something. For someone that shows little outward emotion on the court, perhaps he understands this better than the rest of us.

Anonymous said...

Good job. Very comprehensive.

Anonymous said...

Hawkeye was originally developed for cricket and it's been used in the (proper) world series cricket for 5 years now. When used properly, it should inform the viewers only whether the umpire's decision was correct or not.
In my opinion, the umpires decision is final, regardless of whether he is correct or not, otherwise you are going to end up in court as to whether the ball was 2 atoms away from the line or not..

Anonymous said...

awesome article. everything i needed to know

Jason said...

Anonymous #2: I'm going to have to disagree with you. What if the umpire is wrong with their call?

We saw what happened in the Williams/Capriati match when it was obvious to everyone that the chair ump blew the call.

Given the technology available to verify the call, wouldn't you rather have the correct call made, rather than leave it up to an umpire?

Lets put it this way.. when has a correct call been detrimental to a sport? Would you rather an obviously wrong call stand?

Anonymous said...

Could you say more about how "HawkEye" takes ball compression and skid into account? Ever since last summer, I've been wondering if the computer-generated images are exaggerating the size (image) of the area-of-contact between the tennis ball and the court. The images look as though the ball's entire diameter always comes in contact with the court ... and I have my doubts that that is what actually happens. If the ball does NOT often compress that much, then the extremely close calls (for example, where they have to zoom in to see if there's a sliver of court showing between the ball's image and the line) are being MISrepresented ... or so it seems to me anyway. Why hasn't anyone been talking/complaining about this???
PS: I don't think my question applies (as much) to the skid factor.

-- dhsh

Anonymous said...

Yesterday the same query raised by "anonymous" occurred to me. I would be interested to hear a response, especially on the compression aspect but also on the skid aspect. The width of the image on the screen certainly appears to be wider than I would have expected to represent the area of impact of the ball, even allowing for some compression. If compression is allowed for I wouldn't expect the shape to remain circular as it appears to be.

Anonymous said...

Thanks, Geoff, for seconding my query of a few days ago. Jason, if you have nothing to say about this, please tell us how to contact appropriate other parties for discussion of this topic (the possibly-erroneous instant-replay images of the ball's contact with the court) ... Thanks in advance.

Another angle on instant replay that has come to interest me is the question of player sportsmanship: When a player is the closest pair of eyes to a ball that's on or near the line, perhaps that player should/could take on some of the responsibility to challenge a wrong call by the officials, even if it will go against that player's "self interest" ... I am especially interested in this since the RF/AR semi-final in Melbourne the other day, in which the broadcasters showed the TV audience three instant replays of UNchallenged calls, and all three of them would have gone AR's way (had he challenged ... too bad for Andy)! But on one of those calls (at least), the ball landed just about at RF's feet, so with his good vision, he must have known it was IN, but he did not challenge the OUT line call. If RF really wants to demonstrate his dislike of the new challenge system, and in a match where he clearly could "afford" to let the point go to his opponent, wouldn't it be the gentlemanly and "politically effective" way to go? Ah, but just in case he was wrong, he would lose one of his challenges ... there's the rub!

BTW, I am a fan of Roger Federer, but I also like Andy Roddick, and really felt for him, going through such a drubbing. Along with lots of other fans, I sure wish the match had been a lot closer!

-- dhsh

Anonymous said...

ok, my question is quite basic, yet i don't find the answer to it anywhere. what happens when the hawk-eye proves that the call was incorrect? is the point awarded to the other person regardless of the player position, or is the point replayed?

Anonymous said...

Well, I think it's a great system. It certainly adds suspense when about the entire stadium erupts in 'oooooohh!' when the challenge is made, and it's another aspect for commentators to talk about.

And for 'dd's question: say if (in that oh-so-sad Roddick - Federer game) Roddick got a correct challenge on an IN call on his serve, the point would be replayed if Federer had been at the ball, or about to, had the the call not been made. If Federer had no play on the ball, it would be an ace.

Great article, by the way, Jason.

Anonymous said...

Thanks Jason - answers most of the questions I had but I haven't noticed the eliptical shape that I would have expected to see to take into consideration angle and skid, partilarly low long drive just over the net landing near the baselin. I particularly liked the collection of player comments at the end - Very interesting!

Jason Tsang said...

dhsh - I'm not sure about the answer to your question. I will try to find out and follow up.

dd - essentially, if it is a point ending shot, the call by hawk-eye would determine who gets the point. if it is not a point ending shot, then the point is replayed. the chair ump would decide whether the point is awarded or replayed.

Anonymous said...

Jason, have you had any luck yet, looking (further) into the questions about how the images produced by Hawk-Eye compare to "actual" ball-on-court contact points? I'd be glad to pursue this some more myself, but I don't know who (else) exactly to ask! Any suggestions?
-- dhsh

Anonymous said...

In tennis, what happens if: Player 1 hits it to player 2; player 2 then returns, but the ball is called out from Player 1's initial shot. Player 1 then challenges with Hawk Eye, and this shows the ball was atually in. Does the point go to Player 1 because the ball was in? If so, this is not fair on Player 2 because not only did they get to the ball, but also returned it successfully.

Jason Tsang said...

They would call a let, and the point would be replayed.

Anonymous said...

Partial electronic calling is a bad idea. It brings a completely NEW aspect to the game, an aspect that is inappropriate and benefits no one.

Either go all electronic (take out the human element altogether) or leave as-is.

Anonymous said...

Well done Jason... I searched the Web for the actual scientific explanation for this and couldn't find it (can't find it on "how stuff works" either). Thanks for taking the time to inform us tennis fans of this exciting, new development!

Jerry's Dallas said...

There seem to be some doubt and dispute of the accuracy of this contraption. It seems when the machine makes 'an error' or its accuracy is called in question, the replay would never be re-shown to clear up the controversy. In football, the replay is shown over and over to settle all questions.

Mark said...

Hi Jason. I really enjoyed the quality of writing in this blog post about hawk-eye. No surprise that you have been linked by big time news companies.

I want to share with you my new site, I think it will be right up your alley. It is a nice alternative to the standard fare available on the main forums. LOL.

Anyway, have a looksee.


Anonymous said...

Small correction regarding, "accurate within 2 - 3mm".
This is not necessarily true.

During ITF testing in 2006 Hawk-Eye made the correct call in 100% of all tests, showing an average error of only 3.6mm.

Anonymous said...

great article

Anonymous said...

The hawkeye is wrong. A tennis ball is a sphere, not a disc or hockey puck. When a replay shows the edge of the ball just over a line, the ball is out by just under the radius of the ball. The center of the ball is always the first part of the ball to touch the court. No matter how the ball is distorted by spin, the edge can never be the first part of the ball to contact the court regardless of what direction the ball is coming from. Please explain why no one seems to recognize this obvious fact.

David Nielsen said...

I was upset when the umpire over-ruled the line official declaring a ball from Murray was out in the Murray/Nadal match July 27. Murray had used up his final challenge and was left helpless. I think it should have been resolved by the Hawkeye system in that situation to give a third opinion so to speak. Who knows how that great match would have gone.
Thanks for the article. Good stuff.

Unknown said...

Anyone out there know where to get ranking statistics on which players are the most successful at challenges? In addition to all the info about win record in 5-set matches, tie-breaks, first serve percenage, etc., there should be numbers on who is most successful (or not) at challenging line calls.

Anonymous said...

I am not sure of the system personally. Some of the Hawk-eye reviews seem incorrect. While it is difficult to determine whether the line call was correct, sometimes the review seems to place the ball in a different place on the line. For this reason I would question it's credibility in accuracy. This article also clearly states that it is accurate to 3mm which is not a problem. If the call was closer, the point could simply be replayed however calls are often made within 1mm by the system.

Apart from these issues I have raised, I am aware that it is only used on center court in many tournaments (the Australian Open for example) and this is unfair as it advantages top players who have their matches on center court for the purposes of television.

I think that the technology should be available to any tennis player in any match of any tournament with the rule in place if it is going to be used at all. The technology should also be improved in order to remove the problems that I raised in the first paragraph.

DARKRAIN said...


I'm a bit sceptical on the whole thing.
I think the Hawk eye makes no sense. One can easily animate a ball going
in or going out. Thus can play massive roles with regards to betting/bookies
and outright winning championships.

Say for instance Murray was playing a Tsonga at Wimbledon and was loosing.. Tsonga calls for 3rd Empire (Hawk Eye) and wants to know whether his serve was in or our.. Would it not be possible for Wimbledon to animate the ball ‘just in’?

Or if a the odds where in Favour of Murray, wouldn't the controllers/editors of Hawk Eye be able to effect the outcome?

After all that’s what tennis on that level is all about.

- Digi -

Old Codger said...

I have a few questions.
1. What is the camera shutter speed and are the 6 cameras synchronised? The frame rate (60/sec) is only the sample rate; if it was the exposure the ball image would be about a yard long.
2. How is the system calibrated and how often?
3. The algorithms used for the data capture and trajectory calculation are amazing but how is the bounce position identified? I do hope that my guess is right. (Could it be that the XY coordinates are selected when the Z coordinate change changes sign?)
4. As the samle rate for the ball position is known for every metre (approx)then the bounce is a linear interpolation between two samples. I can't imagine how heavy spin is dealt with.

Anonymous said...

Hi Jason, Macmillan Education Australia is interested in reproducing the table All About Hawk-Eye in our textbook VCE Physical Education 1. Could you please contact me asap to discuss this? jessicakateboland AT

ravi said...

how the speed and angle of the cricket boll is measured bye using HAWK EYE