Regulators Leverage Machine Learning to Combat Match Fixing

Ahead of the commencement of the Rugby World Cup, rumors of teams engaging in espionage have surfaced. While this may be seen as expected tactics, it is undeniable that combating cheating in sports remains a challenge for authorities.

Our new machine learning model could be a game-changer in detecting questionable behavior and unusual results – especially in matchmaking practice.

Currently, changing the results of a match for personal or team benefit is largely due to the anomalies of the sports betting market. When bookmakers notice unusual odds or changes in a betting line, they alert regulators.

However, this approach is limited and often fails to identify all collusions, especially in less popular sports or leagues. This is where machine learning can help.

Essentially, a subset of artificial intelligence (AI), machine learning acts as a digital sonar: mining sports data, uncovering hidden patterns, and reporting unusual events.

Machines can delve into team performance and unexpected fluctuations by examining all aspects of sporting events.

Using artificial intelligence to detect unusual activity

As part of our research, we introduced the concept of “anomaly match detection,” which involved identifying irregular game results regardless of their underlying causes.

There can be a number of factors at play, from strategic losses for future benefit – like the US National Basketball League (NBA) “bank” – to marketing tactics to increase ticket sales or just a bad day of performance.

Our research model allows us to flag unusual game results and hand them over to regulators for further investigation.

By utilizing machine learning, we can detect abnormal matches by comparing our predictions with actual game results.

When we talk about sports anomalies, we are talking about matches that stand out from the norm.

While match-fixing – the deliberate manipulation of results to gain profit – is one possible explanation for unusual game results, it is not the only one.

Identifying the many causes of unusual match results can also help improve our understanding of the complexity of sports.

Faced with an unusual or unexpected result, spectators and officials can ask themselves: was this the result of an unexpected strategy or are there other influences at play?

Learning about basketball

Our research methodology involved training machine learning algorithms to find patterns between certain past events and subsequent game outcomes.

Once these relationships are established, algorithms can predict likely future match outcomes.

Differences between these predictions and actual results may indicate potentially anomalous hits.

To test our model, we looked at whether there were outliers in the 2022 NBA playoffs.

We built models based on data from 2004 to 2020 to predict match outcomes and then compared the machine’s predictions to actual game results.

We found several anomalies in the 2022 playoffs, especially the matchups between the Phoenix Suns and the Dallas Mavericks.

In their seven games played in May 2022, Dallas won four games and Phoenix three.

According to the data, the odds for the 2022 playoffs included a 0.0000064 chance that the Suns and Mavericks will actually play each other in the NBA’s Western Conference semifinals — which includes 15 teams.

We also identified several players with performances in the playoffs that were particularly abnormal based on their previous game data.

This does not mean that it was a match. Rather, our results report games and players, which moderators could then monitor if match-fixing was a concern – which it wasn’t, this was just an example of testing the model.

This approach to detecting anomalies in a series of matches can be applied to many sports.

Examining a significant number of anomalies can provide valuable insights into unusual match events, helping regulatory bodies and sports organizations conduct thorough investigations and maintain fair competition.

Increase confidence in sports

Although our research focuses on specific sports, the principles and techniques can be extended to other arenas.

Research shows that machine learning can be used to safeguard the integrity of sports competitions and help regulatory bodies, sports organizations and law enforcement agencies maintain fairness and public trust.

But as we embrace the potential of machine learning, we must also navigate the ethical implications and ensure its open use.

The future of sports may well see AI become an ally to fans, helping to ensure a level playing field where talent excels and viewers enjoy the authenticity of sporting events.