Optimising long-term outcomes using real-world fluent objectives: an application to football
Optimising long-term outcomes using real-world fluent objectives: an application to football
In this paper, we present a novel approach for optimising long-term tactical and strategic decision-making in football (soccer) by encapsulating events in a league environment across a given time frame.
We model the teams' objectives for a season and track how these evolve as games unfold to give a fluent objective that can aid in decision-making games. We develop Markov chain Monte Carlo and deep learning-based algorithms that make use of the fluent objectives in order to learn from prior games and other games in the environment and increase the teams’ long-term performance. Simulations of our approach using real-world datasets from 760 matches shows that by using optimised tactics with our fluent objective and prior games, we can on average increase teams mean expected finishing distribution in the league by up to 35.6%.
196-204
Beal, Ryan, James
d9874cb0-bd92-4a16-8576-78d769b41ff7
Chalkiadakis, Georgios
06a66d27-705a-46a2-beeb-b3b7d443ffdb
Norman, Timothy
663e522f-807c-4569-9201-dc141c8eb50d
Ramchurn, Sarvapali
1d62ae2a-a498-444e-912d-a6082d3aaea3
2021
Beal, Ryan, James
d9874cb0-bd92-4a16-8576-78d769b41ff7
Chalkiadakis, Georgios
06a66d27-705a-46a2-beeb-b3b7d443ffdb
Norman, Timothy
663e522f-807c-4569-9201-dc141c8eb50d
Ramchurn, Sarvapali
1d62ae2a-a498-444e-912d-a6082d3aaea3
Beal, Ryan, James, Chalkiadakis, Georgios, Norman, Timothy and Ramchurn, Sarvapali
(2021)
Optimising long-term outcomes using real-world fluent objectives: an application to football.
20th International Conference on Autonomous Agents and Multiagent Systems, London (Virtual), Virtual, United Kingdom.
03 - 07 May 2021.
.
(doi:10.48448/em7e-gb49).
Record type:
Conference or Workshop Item
(Paper)
Abstract
In this paper, we present a novel approach for optimising long-term tactical and strategic decision-making in football (soccer) by encapsulating events in a league environment across a given time frame.
We model the teams' objectives for a season and track how these evolve as games unfold to give a fluent objective that can aid in decision-making games. We develop Markov chain Monte Carlo and deep learning-based algorithms that make use of the fluent objectives in order to learn from prior games and other games in the environment and increase the teams’ long-term performance. Simulations of our approach using real-world datasets from 760 matches shows that by using optimised tactics with our fluent objective and prior games, we can on average increase teams mean expected finishing distribution in the league by up to 35.6%.
Text
Football_Objectives_AAMAS_21
More information
Published date: 2021
Venue - Dates:
20th International Conference on Autonomous Agents and Multiagent Systems, London (Virtual), Virtual, United Kingdom, 2021-05-03 - 2021-05-07
Identifiers
Local EPrints ID: 449655
URI: http://eprints.soton.ac.uk/id/eprint/449655
PURE UUID: c613cae5-9cc0-4465-8093-50e868e0fa62
Catalogue record
Date deposited: 10 Jun 2021 16:31
Last modified: 17 Mar 2024 03:41
Export record
Altmetrics
Contributors
Author:
Ryan, James Beal
Author:
Georgios Chalkiadakis
Author:
Sarvapali Ramchurn
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics