SAN JOSE, Calif. – Four Indian American students at San Jose State University, Shivam Tyagi, Sarath Chandra Makkena, Rashmi Kumari and Swayam Swaroop Mishra, under the guidance of their professor, Dr. Vishnu Pendyala, have created an application that can predict the number of balls or the duration of a cricket match before it starts using historical data and, when the match is in progress, using historical and real-time data.
Using the techniques learned in their data mining class taught by Pendyala, the students developed a prototype tool. They used data analysis and machine learning techniques to predict the duration of a cricket game in terms of the number of balls. The prototype uses historical data to create a profile for the teams and their players.
Based on the profiles of the players playing, the tool predicts how many balls the player will play, and based on each player’s performance, the number of balls that the whole team will play is predicted. The prediction takes into account the bowlers and batsmen playing in the team and their combinations.
The tool also gives an option to make a real-time prediction for a batsman when a particular bowler is bowling.
Pendyala told India-West that the application has become a huge hit among the students and can potentially make the cricket fan base more thrilled about the game.
“Nothing brings people together as much as a sport can… Cricket is one such sport which is the second most popular sport after football in the world,” he said. “Players are worshipped and treated like gods. The sports industry and the advertisement industry work together to complement and benefit each other. Different product companies try to grab the advertisement slots to display and introduce their merchandise to people…Companies pay a lot of fees for purchasing advertisement slots. This enables the team management to pay their players well…A relationship thrives when both parties benefit from it. The inherent unpredictability of sports event is a huge hindrance and a major reason for contention.”
Pendyala noted that since there is no control of anyone over the duration of a match, it become a risky business for companies buying advertisement slots because if the match ends before their slot arrives, they not only lose money but also lose an opportunity to make customers.
Using the fully developed tool and taking into account other parameters like weather and location, better predictions can be made which can then be used while selling advertisement slots, he stated.
An idea about the duration of the game will benefit both the team management as well as companies by giving them an idea of what they will get for their money and with what probability, he said.
Credit: Google News