TOKYO -- When double world champion judoka Chizuru Arai on Wednesday bested her Austrian opponent Michaela Polleres in a blistering final for the women's 70 kg category, Japan won its sixth gold medal in the martial art for the Tokyo games, putting it just two short of its record haul tallied in Athens in 2004.
But this time around, the coaches, athletes and sponsors are discovering new ways to throw their opponents to the ground by tapping the insights gained from data technology.
Since the 2016 Rio Olympics, Japan has sent analysts to almost every major international judo tournament to record and analyze video of more than 4,000 fighters. Findings are available to both athletes and coaches.
Referees in the last two world championships were found to reward corner drops and their many variations over bigger, flashier moves like shoulder throws. Naohisa Takato, who won the gold in men's 60 kg judo, and coach Minoru Konegawa adjusted their fighting strategy in response to predictions that more opponents would attempt drops.
For Shohei Ono, gold medalist in men's 73 kg judo, coach Yusuke Kanamaru tracked the number of shido , or warnings, by referees in roughly 13,000 matches since 2019. Ono, armed with the knowledge that referees were issuing about 15% fewer shido than in the past, remained calm even when receiving two in his final match Monday, eventually prevailing in sudden-death overtime.
The All Japan Judo Federation also ranked foreign competitors using a formula widely used in go and soccer tournaments instead of relying on the official world rankings, which are calculated solely on their performance in competition.
"We're able to better measure the athletes' actual ability and identify those we need to pay special attention to, even if they are technically ranked low," said Takanori Ishii, a member of the federation's science and research department.
In addition, the federation has since 2019 been monitoring athletes' body composition daily, alerting them to any unusual readings through the Line instant messaging app.
The framework came in response to the Rio Olympics, when cramping led to the loss of a critical match. "We monitor fatigue and other issues that aren't visible to the physical eye to plan rests and training programs," Ishii said.
Japan and Canada in Olympic women's soccer at the Sapporo Dome on July 21. © Kyodo Data has also played a critical role for Japanese women's soccer team, which beat Chile 1-0 on Tuesday to secure a place in the quarterfinals.
For assistant coach Morinao Imaizumi, the key metric to watch is expected goals -- the likelihood of a successful shot based on such information as distance, angle and the number of defensive players in the path.
"Is it the right call to shoot many shots that only have a 5% chance of making it in the goal?" said Imaizumi, who is also an assistant coach at Florida State University in the U.S. He tells players to always consider the best course of attack and to pass the ball rather than bet on an unlikely shot.
His tool of choice is Sportscode, software that allows him to analyze each player's performance at specific points in a game and offer detailed pointers, including on how to kick the ball.
"It's important to increase our ability to re-create good plays, which will lead to better decision-making in different scenarios," Imaizumi said.
Analyzing past data can also highlight what a team needs to work on. Japan was able to recover less than 50% of loose balls at the Women's World Cup in 2019, which it lost in the round of 16 -- a significantly lower percentage than any team that made the quarterfinals.
"Looking at the data made me more aware of our weaknesses in contact play," defender Risa Shimizu said. "Now, I make sure to check the numbers against my impressions of a match."
"The coaches' personal insights are the most important thing," Imaizumi said. "With objective data to back it up, players walk onto the pitch with confidence."
Climbers Akiyo Noguchi, left, and Tomoa Narasaki use a tablet. Karate athletes started using the same video analysis program employed in judo in 2019. In the just-added BMX freestyle park competition, Rim Nakamura is working with Tokyo-based sponsor WingArc1st at a smart training facility completed just last year to analyze his speed and movement and help him master harder maneuvers.
Sports climbers Tomoa Narasaki and Akiyo Noguchi, both members of Team au, are receiving backing from the KDDI group. The athletes use KDDI's artificial intelligence technology, originally for tracking things like customer traffic, to identify their center of gravity based on the position of their joints, or to visualize how they look from various angles. This information lets them climb faster and straighter.
Such valuable data did not always make its way to coaches in the past. But stakeholders are now making greater efforts to ensure that athletes benefit from these insights.
At the All Japan Judo Federation, the science and research department frequently consult with coaches and nutritionists on how to translate the data into results.
"All of this is just information when we gather it," Ishii said. "The key is turning it into intelligence."