I have never placed a bet on a football match in my life. Well, once. When in high school I bought a Pro Line ticket, the Ontario Lottery and Gaming corporation’s legal betting means, and placed some final score bets, and lost.

My lack of interest in the global betting/match-fixing scandals comes in part from the fact I find betting on sports boring at best, lame at worst. In fact, that’s my view of gambling in general; everyone thinks they know something the system doesn’t. They don’t, and so they’re essentially throwing money away.

Unless they rig the results.

Anyhoo, the Guardian posted this neat little video of how some companies track fixed matches through betting trending patterns, and it got me thinking about another reason to keep improving the state of analytics, particularly in regard to developing reliable match outcome probabilities based on a host of available team metrics.

Let’s say Sportradar sends information about a game they believed based on betting patterns is fixed in some way. In lieu of on-the-ground investigatory work, what would constitute possible evidence of a fixed match?

Analytics might help provide an answer. Let’s say team X fixed a match against a lower league opponent, perhaps by conceding a loss. Sportradar notices some fishy betting patterns. From an advanced stats perspective, what you could look for? Currently, not much, but not nothing either. Presumably a team conceding a game would likely have an irregularly low save percentage. They also wouldn’t be shooting much, so the game would likely show a low PDO (save % + shooting %).

That wouldn’t mean much in itself, but normally a team expected to win will have a reasonably good TSR. Now in order to look like they’re trying, a team might attempt to rack up way-ward shots to avoid suspicion, so TSR might not be the best metric to use, but a shots on target ration (SoTR) might be, particularly if one side is actively trying to concede a goal.

This hypothetical is extremely crude, but you see what I’m getting at here. Advanced analytics might make determining a true footballing outlier much easier, and could help aid the fight against match fixing by focusing investigations by sorting out signal from noise. Nothing will eliminate the practice altogether, but limiting the scale would go a long way, and analytics could be an important tool in that fight.