The Scientist : NewsBlog Print: Game, set, matching umpire calls in tennis
The Scientist: NewsBlog:
Game, set, matching umpire calls in tennis
Posted by Elie Dolgin
[Entry posted at 16th April 2008 02:30 PM GMT]

Umpires at Wimbledon, Roland Garros, and Arthur Ashe Stadium might deserve a break, according to a new study published online this week in Proceedings of the Royal Society B. The study found that disputes over close calls during professional tennis matches arise because of double faults in the way information is processed in the brains of players and umpires. Nonetheless, both perceptions are remarkably accurate, though umpires are in the right most of the time.

At next month's French Open, tennis balls will fly across the court at speeds of up to 150 mph. In such a fast-moving sport, arguments often erupt between players and umpires as to whether a ball bounces inside or outside the boundary lines. So in 2006, major tennis tournaments introduced a high-tech, ball-tracking system called Hawk-Eye that can pinpoint the ball's position to within a tenth of an inch. Line calls are still made by umpires, but players are allowed to challenge an umpire's decision twice per set of play.

"It occurred to me that the data collected as a by-product of Hawk-Eye is a perceptual experiment that could reveal something interesting about how decisions are made," said George Mather, a psychologist at the University of Sussex in England who led the study.

Mather, who studies how humans perceive motion and is a huge Roger Federer fan, investigated 1473 challenges made by 246 players to test whether disputes occur because of perceptual uncertainty, or as a backhanded way for players to get ahead. He found that 94% of challenges occurred within about 4 inches of the line — a distance less than twice the diameter of the ball. Both players and umpires made mistakes, but only 40% of challenged line calls turned out to be incorrect. Advantage: umpires.

Mather found a sharp peak in challenges the closer the ball landed to the boundary lines, which suggested that disputes occur because of minor perceptual differences between the brains of players and umpires, rather than because of gamesmanship or lapses in concentration. So Mather created a computer model in which a ball bounces, and theoretical players and umpires judge the ball's position — either in or out — but with a small amount of uncertainty that follows a simple bell-shaped curve. If the players and judges agree, then the simulation ends. But if they disagree, the player challenges the call, and the ball's true location is used to settle the dispute, just as Hawk-Eye does in real life.

The model paralleled the data from actual matches with a high degree of accuracy, indicating that players' and umpires' brains make errors in predictable ways. "You can actually capture performance very accurately by introducing a bit of noise into the decision-making process," Mather told The Scientist.

The study "took what we do in the dark rooms and labs and put it out into the real world," said Saumil Patel of the University of Texas Medical School at Houston, who was not involved in the study. He was surprised such a simple model could capture complicated neural processes when "more than half the brain is involved in these types of decisions." By treating the brain as a "black box" in the model, Patel said the ball was still in Mather's court to show "what areas of the brain are involved or what mechanisms are used to arrive at these judgments."

Mather's model predicted that an umpire should only make a few perceptual mistakes per set; making the player's two challenges per set limit a pretty accurate one, according to the researcher. Michael Wright of Brunel University in London, England, was impressed with this result, but wondered whether umpires' performances change over the course of a match. Mather's analysis was based on more than 1,000 decisions, Wright noted, but performance might wane over time because of exhaustion, or possibly even improve if umpires enter the sweet spot of attentiveness.

Mather's analysis also grouped athletes of different skill levels together. But were top-ranking players like Roger Federer or the Williams sisters better at making line calls than others? Actually no, said Mather, adding that the grand slam explanation for the kinds of outbursts that made John McEnroe famous remains simple random fluctuations in perception.

 

Rate this article

Rating: 4.15/5 (20 votes )





Real Umpire comments
by frank bernard

[Comment posted 2008-05-26 14:33:33]

I do 10 to 15 tournaments a year as a roving umpire. In games played with a roving umpire, the players make their own calls with the rover over ruling only if they are positive an incorrect call has been made. Since the player is supposed to give his opponent the benefit of all close points, he/she should play anything that's close to the line and the ump should very rarely have to over rule. In practice, very few over rules are made because most players do adhere to the benefit of doubt rule (especially when an umpire is observing them) and because of the need to be positive an incorrect call has been made. In an average 6 hour day of observing many different tennis games I might make 1 overrule. Should more overrules be made? Not with the current guideline for umps.
I must note that I have seen college tennis matches decided by overrules. Under college rules, the first two overrules cost the player only that point, but from the 3rd overrule the point penalty system kicks in. I.e. on the 3rd overrule the player loses not only the point being played but also one additional point and so on. College players gain ranking points for matches won and some are willing to risk overrules at critical junctures to win points and matches. And, every now and then they will call a ball out that is clearly observed by the umpire to be in. I'm sure the pros are no different and they will sometimes challenge calls as a matter of gamesmanship rather than from a true belief the ball was incorrectly called.





Comment on this blog