Thursday, June 24, 2010

Left or Right? Early Detection of Soccer Penalty Kicks

Ω

In the split second before foot meets ball, a soccer player’s body betrays whether a penalty kick will go left or right, according to recent research in cognitive science at Rensselaer Polytechnic Institute. Photo Credit: Rensselaer: Daria Robbins

In the split second before foot meets ball, a soccer player’s body betrays whether a penalty kick will go left or right, according to recent research in cognitive science at Rensselaer Polytechnic Institute. The findings could explain how some top goalkeepers are able to head off a penalty kick, diving in the correct direction in advance of the kick. It could also point the way to changes in how players kick, and goalies react.

The research, performed by Rensselaer doctoral student Gabriel J. Diaz, employed motion capture technology and computer analysis to identify five early indicators of the direction a ball would ultimately be kicked. Diaz said his research stemmed from an observation of real-world penalty kicks, in which players aim for the left or right side of the goal while hiding their choice from the goalkeeper.

“When a goalkeeper is in a penalty situation, they can’t wait until the ball is in the air before choosing whether to jump left or right — a well-placed penalty kick will get past them,” Diaz said. “As a consequence, you see goalkeepers jumping before the foot hits the ball. My question is: Are they making a choice better than chance (50/50), and if so, what kind of information might they be using to make their choice?”

Diaz tested 27 potential indicators of kick direction — 12 drawn from sports literature and 15 derived from a computer analysis of the kicks — and identified five as reliable indicators of the direction the ball will go.

In the second part of his work, Diaz also showed that four of the five early indicators he identified are used by people who are able to predict the direction of the kick before the foot strikes the ball.

Diaz used motion capture technology — cameras, sensors, and software — in Rensselaer Associate Professor Brett Fajen’s Perception and Action (PandA) motion capture lab to record the movements of three college-level penalty kickers. The technology is similar to that used to create realistic movement in computer-generated graphics.

More than 40 sensors placed on 19 major joints of the body (and the ball) recorded the movements of the kickers as they stood behind the ball, took two steps, and kicked either to the left or of the right side of a goal. Diaz recorded 126 kicks, half to the left and half to the right.

Then he tested the data he collected against the suite of 27 potential indicators.

Twelve of the indicators — such as the angles of the kicking foot, kicking upper-leg, and kicking shank — were movements of a specific, or “local,” area of the body highlighted by coaches and sports psychologists. Among them he found that two — the angle at which the non-kicking foot is planted on the ground, and the angle of the hips as the kicking foot swings forward —are reliable indicators of kick direction.

The 15 indicators identified in a computer analysis of the kicks were so-called “distributed movements” — patterns of coordinated movement throughout the body. Three of the “distributed” movements proved to be reliable early indicators, none of which appears to have drawn previous attention in sports literature.

Emerging evidence in the study of motor control has pointed to a significant role for distributed movements, Diaz said. He described distributed movement as a combination of movements developed over many repeated attempts to perform a task, in this case kicking in a particular direction.

“When, for example, you shift the angle of your planted foot, perhaps in an attempt to hide the direction of the kick, you’re changing your base of support. In order to maintain stability, maybe you have to do something else like move your arm. And it just happens naturally,” Diaz said. “If this happens over and over again, over time your motor system may learn to move the arm at the same time as the foot. In this way the movement becomes one single distributed movement, rather than several sequential movements. A synergy is developed.”

A distributed movement is complex, but, as Diaz’s second experiment indicates, some people may be using it — however unconsciously — to inform their judgment as to which direction the ball will go.

In his second experiment, Diaz played an animation of the motion capture data to a group of 31 subjects, and asked the subjects to pick which direction they thought the ball would go. In the animation, each body joint is represented by a dot, and movement of the body is easily recognizable as such. The animation runs from the standing-start until the foot reaches the ball, at which point the screen goes black and subjects pressed a button to the left or right of the screen, indicating which direction they thought the ball had gone.

Among his 31 subjects, all of whom were novices to the activity, 15 were not able to score above chance (50/50), even when given one-half second after the scene to ponder the outcome. Sixteen, however, did perform better than chance.

Diaz then looked for relationships between successful judgments on ball direction and each of the “local” and “distributed” movements he had tracked. His analysis revealed strong correlations between the two “local” and two of the three “distributed movements” that were reliable indicators of kick direction.

“The question is, knowing these potential sources of reliable information, what do people actually use?” Diaz said. “I found four reliable sources that were well correlated with subjects’ judgments.”

Another finding, he said, is that the 16 successful subjects waited longer than the 15 unsuccessful subjects to make their choices (if the half-second elapsed without a response from the subject, no result was entered).

“There is a clear relationship between response timing and performance,” Diaz said.

Diaz said his findings have set the stage for further exploration. He would like to create a training regime to guide subjects’ attention toward more reliable indicators of kick direction. He also wants to know if professional goalkeepers would perform better than novices on the task.

Similar studies using video data of penalty kicks among professional Dutch goalkeepers showed that not all professional players are better than novice subjects, he said.

“Only a subset are better than average. I want to know — what is it that these successful experts are doing better than novices?”

Wednesday, June 9, 2010

Sleep preference can predict performance of Major League Baseball pitchers

Ω

Pitchers who are 'morning types' appear to have an advantage over night owls

A Major League Baseball pitcher's natural sleep preference might affect how he performs in day and night games, according to a research abstract that will be presented Wednesday, June 9, 2010, in San Antonio, Texas, at SLEEP 2010, the 24th annual meeting of the Associated Professional Sleep Societies LLC.

Results indicate that pitchers who were morning types performed statistically better overall than those who were evening types. In early games that started before 7 p.m., the earned run average (ERA) of pitchers who were morning types (3.06) was lower than the average ERA of pitchers who were evening types (3.49); however, in games that started at 7 p.m. or later, pitchers who were evening types performed slightly better (4.07 ERA) than morning types (4.15 ERA).

"We were surprised to see that chronotype did affect pitching," said principal investigator and lead author W. Christopher Winter, MD, medical director of the Martha Jefferson Hospital Sleep Medicine Center in Charlottesville, Va. "We were also surprised to see that pitchers who were more 'morning type' seemed to do better overall."

Individual pitchers showed a trend toward higher ERAs in the late games. According to Winter, this supports previous research showing that the peak performance time for most athletes is between mid-afternoon and early evening.

The study involved 18 pitchers from five MLB teams: the Los Angeles Dodgers, New York Mets, Philadelphia Phillies, San Francisco Giants and Tampa Bay Rays. Sleep preference was determined using a modified version of the Morningness-Eveningness Questionnaire (MEQ). It identifies a person's tendency to be either a morning type who prefers to go to bed and wake up early, or an evening type who prefers to stay up late at night and wake up late in the day. Ten participants were found to be evening types, and eight were morning types.

The study used the players' statistics from the 2009 season, which provided about 728 early innings and 845 late innings for analysis. Game start times were adjusted for travel using the principle that for every time zone crossed, it takes 24 hours to adjust.

"These results are important as they are potentially giving insight into an entirely new way to grade or classify an athlete, in this specific case a pitcher," said Winter. "This study may provide insight as to which pitchers would be best in a given situation based upon when the game is being played. For example, a critical game being played in the evening might be a better situation to pitch an evening-type pitcher versus a day-type pitcher."

Winter also has studied the effect of travel across time zones on the performance of MLB teams. At SLEEP 2008 he presented the initial findings of a 10-year retrospective study that was later published in the September 2009 issue of the International Journal of Sports Physiology and Performance. He found that teams traveling from Western time zones to Eastern time zones were 14 percent more likely to win than teams traveling from east to west. Teams also won more than 60 percent of the games in which they had a three-hour "circadian advantage.

Tuesday, June 8, 2010

Will the new World Cup soccer ball bend?

Ω


Physics experts at the University of Adelaide believe the new ball created for the 2010 World Cup, called the Jabulani, will play "harder and faster", bending more unpredictably than its predecessor.

But why? And what will it mean for the game?

"The Jabulani is textured with small ridges and 'aero grooves' and represents a radical departure from the ultra-smooth Teamgeist ball, which was used in the last World Cup," says Professor Derek Leinweber, Head of the School of Chemistry & Physics at the University of Adelaide, who has previously written about and lectured on the aerodynamics of cricket balls, golf balls and the 2006 World Cup soccer ball, the Teamgeist.

Along with student Adrian Kiratidis, who is studying for his Master of Philosophy (MPhil) in Physics, Professor Leinweber has been reviewing the physics behind soccer balls and what that means for the Jabulani. Adrian is also a soccer enthusiast.

"While the governing body FIFA has strict regulations on the size and weight of the balls, they have no regulations about the outside surface of the balls," Professor Leinweber says.

"The Teamgeist was a big departure at the last World Cup. Because it was very smooth – much smoother than a regular soccer ball – it had a tendency to bend more than the conventional ball and drop more suddenly at the end of its trajectory.

"By comparison, the aerodynamic ridges on the Jabulani are likely to create enough turbulence around the ball to sustain its flight longer, and be a faster, harder ball in play.

"The Jabulani is expected to 'bend' more for the players than any ball previously encountered. Players are also discovering new opportunities to move the ball in erratic ways, alarming the world's best goalkeepers. By the time the ball reaches the goalkeeper, the Jabulani will have swerved and dipped, arriving with more power and energy than the Teamgeist."

University of Adelaide students have also put the new World Cup soccer ball to the test on the soccer field. Based on Professor Leinweber's theories, they've attempted to "bend" the Jabulani and have also kicked the Teamgeist and a regular soccer ball for comparison.