Institutional Repository
Technical University of Crete
EN  |  EL

Search

Browse

My Space

Competing with humans at fantasy football: Team formation in large partially-observable domains

Chalkiadakis Georgios, Tim Matthews, Sarvapali D. Ramchurn

Full record


URI: http://purl.tuc.gr/dl/dias/FEB49255-6D6A-430F-BC85-CB8CFD6AA608
Year 2012
Type of Item Conference Full Paper
License
Details
Bibliographic Citation T. Matthews , S. D. Ramchurn ,G. Chalkiadakis.(2012).Competing with Humans at Fantasy Football: Team Formation in Large Partially-Observable Domains.Presented at the Twenty-Sixth Conference of the Association for the Advancement for Artificial Intelligence, Toronto, CA, 22 - 26 Jul 2012.[online]. Available: http://www.intelligence.tuc.gr/~gehalk/Papers/fantasyFootball2012cr.pdf
Appears in Collections

Summary

We present the first real-world benchmark for sequentially- optimal team formation, working within the framework of a class of online football prediction games known as Fantasy Football. We model the problem as a Bayesian reinforce- ment learning one, where the action space is exponential in the number of players and where the decision maker’s be- liefs are over multiple characteristics of each footballer. We then exploit domain knowledge to construct computationally tractable solution techniques in order to build a competitive automated Fantasy Football manager. Thus, we are able to es- tablish the baseline performance in this domain, even without complete information on footballers’ performances (accessi- ble to human managers), showing that our agent is able to rank at around the top percentile when pitched against 2.5M human players.

Services

Statistics