Untapped potential to detect match-fixing
With the subtitle shaping the future of sports betting, this year’s “Betting on Sports”, dedicated conference for senior executives from European sports betting operators will host a session on the potential of AI to address match-fixing.
Sports integrity and match-fixing are high on the agenda for valid reasons. Markets such as Sweden are said to be especially targeted by organised crime to drive match-fixing.
“It is clear that the industry is striving towards AI across a range of areas including odds movements and match-fixing” says Chris Kronow Rasmussen, Senior Manager at FCG who will be speaking at the conference on this topic.
“Open-source data on betting forms as well as player data will be key, but there are also other factors and circumstances to consider before one can assume a ‘sure fix’” continues Kronow Rasmussen who is also an adjunct professor in sports integrity at the University of New Haven.
Sufficient data volumes will be crucial to establish objective models. Indications on match-fixing are not absolute, and the need for a human eye cannot easily be excluded. At present, the basis for this is yet to be developed. Existing data providers on odds movements can arrive at a very high certainty levels to detect a fix, but the need for investigation cannot be excluded.
Similarly to work in the AML area, one challenge is the risk of false positives, as the issue of match-fixing not being effectively regulated by law. Meanwhile, approximately one thousand matches are estimated to be fixed every year.
Kronow Rasmussen comments that one way to address match-fixing from a legal perspective is through bribery as the criminal offence.
Individual actors in the marketplace are looking at both bets movements, monitoring odds, transactions and applying standard transaction monitoring systems, largely based on AML But they do not have historical data. Current TM systems are based on AML. Alerts on odds movements are managed in parallel systems.
“The sports and betting community should be in a good position to move forward to tackle these issues, but the key will be data availability and cooperation. Transaction data in gambling companies are better than in banks. If they can predict who will make money then they should also be able to say who is placing bets on suspicious matches.”
For further information reach out to Chris at Betting on Sports in London this week or by email.
