2006 FIFA World Cup, Germany vs Argentina, Quarterfinal penalties: German goalkeeper, Jens Lehmann, reads a note from his sock and readies himself. He exactly knows what to do. What was written in the note? Instructions—what to look out for and which side to dive. Lehmann saved two goals and correctly anticipated the other two, helping Germany win the match. The notes were not a result of guesswork but a collation of information on who plays Argentina’s penalties and how they should be taken. This is big data for you.
Technology has always answered questions that we humans cannot. Therefore, it should be no surprise when one says that data analytics is doing its best in scripting victories for sports teams. Gathering data and analyzing patterns of players and matches, among the least, is big data’s big contribution toward sports.
An article on sports analytics will not be complete without referring to the famous Oscar-nominated movie, Moneyball. It is based on Oakland Athletic’s 20 consecutive wins in 2002 because of the sophisticated sabermetric approach. This showed that data analytics has hit a home run in baseball. It’s no secret that the extensive data analytics by SAP was instrumental for Germany’s success at the 2014 FIFA World Cup. Also, IBM’s in-depth analysis on backhands and dropshots of tennis players has shown them what is needed for the ball to be in their court.
The Pitch for Big Data in India: Suitable?
“Sports analytics in India is quite niche,” says Sandeep Kannambadi, co-founder and chief technology officer of Sportingmindz Technology, a Bangalore-based sports analytics company. India houses only a handful of these companies that have ventured in this field, which amalgamates sports and information technology.
About a decade ago, cricket analytics was not explored to its core, with only a few international products in the market. Looking at this scenario and seeing the advancement in technology, Javagal Srinath took the initiative to collect data on a fairly moderate level, however without much analysis.
This is where companies like Sportingmindz, whose founders were sports enthusiasts with a technology background, came into being. Sportingmindz’s clientele include several IPL teams such as Kolkata Knight Riders, Kings XI Punjab and Delhi Daredevils; the Indian Hockey team; English cricket county teams; and cricket boards of South Africa, Sri Lanka, Bangladesh and others.
When you examine the depth of cricket analytics, you will realize that the entire match has been converted into data. This data is all about decisions—batting and bowling orders, environmental factors like pitch conditions and humidity, and responsiveness of players toward crowd and fans. Analytics does not let anything go away without capturing it—even Virat Kohli’s temperament will be converted to data.
“We understand that each bowler and batsman has a pattern that he follows, consciously or unconsciously—that’s the way he plays,” Kannambadi says. “We then start analyzing the behavioral patterns of the player.”
He said this has helped them come up with various strategies such as how to get a batsman out quickly, how to face a ball, and how to enhance the player and team performance. With this data, SWAT analysis is also done for each player to further enhance the analytics.
“Every ball has around 15-16 parameters,” Kannambadi says. “We have data for all the balls bowled in the world. The dataset is huge.”
During matches such as the IPL, the team has a dedicated data analyst, constantly monitoring the players, either by being physically present with the team or working remotely in the office. The analyst feeds data in their algorithms, which then process the data fed and perform numerical calculations. And there you have your prescriptive analyses for the next match. The analyst then updates the team and the coach about all the things that went right and wrong in the current match and what the player ought to do in the next match.
Most of the strategies adopted by the team are conceived during the team meetings. “How the analysts prepares the team is important. The analyst tells the bowler this is exactly how you should bowl,” Kannambadi says. “If the bowler is not convinced or does not believe it, it is then up to the captain to convince the player.”
As with every algorithm, there is always a margin of error. However, Kannambadi says keeping all other factors constant, the only percentage of error that may occur would be the timing of the ball and the time the ball is hit—the urgency of the player to hit the ball a few milliseconds before or the latency to hit the ball a few milliseconds later. This is where analytics has no control.
According to Kannambadi, cricket analytics in India is very advanced, but this is not the case with other countries. “They need some thoughts on how to approach the game. Getting data is one thing, but it is important for somebody to tell them how to use the data.”
What About Other Sports?
Cricket in India is worshipped by all. Hence, it is obvious that technology would prefer to join hands with cricket and just give a fleeting glance at other sports such as tennis, table tennis, hockey and swimming.
“Cricket is bombarded with data,” said Santhosh Patil, founder and director of Gamatics, a sports analytics company incubated in Bangalore. “However, people now want to look at other sports. We feel this is the right time to tap in to other sports as well.”
Currently, Gamatics is focusing on swimming and other individual sports. “Right now we have data for swimming competitions at the state and national levels over the past two years. We also started looking at the training and practice sessions of players,” says Patil.
It is common knowledge that if you’re into the sports business that does not revolve around cricket, it is very difficult for you to make money. Data analytics for other sports is also facing similar troubles. “We approached a couple of investors. They said analytics is very good, but this will not make money. This is the mindset of Indians,” says Patil.
Will Technology Help Mould the Next Saina?
There is more to data than just increasing the winning chances of your favorite team. Sports analytics is also being currently used in sports academies, where our future sports heroes are being nurtured.
In addition to undergoing regular training, the entire training cycle of sports amateurs can be recorded, either in the form of videos or hundreds of data points, which can later be used for enhancing their performance.
Kannambadi says after the coach feeds data in the software, various patterns about the player will be revealed. What diet should be given to the player? What kind of muscles should be developed? How should the player improve his attitude on and off field? Is he ready for the next level? Big data will help answer all such questions.
Skill fitness, injury, and mental fitness are the factors that analytics will help calculate. “Once you know a player has gone through the entire training cycle, it is then easy to replicate the same for other players who show the same caliber,” says Kannambadi.
Constantly following analytics will also help academies plan tournaments for the students. “If you know that a tournament is in September, you can estimate when you need to start training and what kind of training is required,” says Kannambadi. “For example, three months later, the coach can do a split screen of the videos of how the player was before and after, and tell him this is how you were and this is how you are.”
However, when it comes to training budding sportspersons, for example, in badminton, historical data is unavailable. Such analytics can be developed only by collecting information from the scratch. “Unfortunately for us, we don’t have any data regarding Saina’s training under Gopichand,” says Kannambadi.
But he is confident that big data will be instrumental in bringing out sports players with international standards. “Rather than having one Saina, we can have a hundred Sainas—may be not as talented as her, but at least they’ll be able to break in the top circle,” he says.
So the next time you question why Team India decided to bat first on a day where you thought bowling first should have been better, just think that this was the result of some extensive data crunching.