It would be interesting to have data to support who would win between boxers of different historical periods. There’s also lots of money to be made when you can find a way to promote a horrible addiction like gambling.
Using the following data, a program would attempt to give accurate odds to match ups:
- Basic match information (wins, losses, K.O.)
- Types of punches thrown and landed where
- Types of punches received and landed where
- Tactics used vs different styles of boxers
- Punches per round / fatigue
This would hopefully could be robust enought to predict past wins and losses.
Getting the Historical Boxing Data
Without this data, the business is useless. Currently, only win / loss stats are publicly available.
Cheap labour or video analysis software may be able to build data around video footage. However, this data would not be complete and should be treated as a sample.