Generating Betting Odds for Fantasy IPL using Monte-Carlo Simulation and Neural Networks

Summary

  • Developed a ball-by-ball prediction model using Neural Networks in TensorFlow for simulating an IPL match

  • Used more than 10 game states as features at any given moment in the game.

  • Used Monte Carlo Simulation leveraging the prediction model to get the distribution of more 12 game and player parameters (like …..).

  • The distributions were then used to generate odds of winning for a particular bet.

  • The generated odds were further verified by a bookmaker software and used to decide the winning amount of any bet.

Data Source - Data for 100+ IPL matches was scraped from websites like ____