21–23 Mar 2018
Nordita, Stockholm
Europe/Stockholm timezone

Can machines learn to solve football? A look at how we model complex systems

22 Mar 2018, 15:15
45m
122:026 (Nordita, Stockholm)

122:026

Nordita, Stockholm

Speaker

David JT Sumpter (Uppsala University, Sweden)

Description

Football is the most mathematical of sports. Passing networks, the geometry of player positioning and randomness are all important. There are lots of people—coaches and players---who understand football intuitively. Can we train a computer to understand the game and generate insights these people have missed? I argue that we can, but we can’t just throw away these people’s knowledge and start from scratch. I explain how studies of collective behaviour can be used to help improve our analysis and understanding of the beautiful game, and how the problem of modelling football should inform the way we approach all complex systems.

Presentation materials

There are no materials yet.