Science intersects with our daily lives in more ways than we can imagine. The most mundane acts such as thinking, eating, and gazing at the sky relate to one branch of science or another. However, science doesn’t only make our lives better, it can make it tougher too.
Let’s take sports betting as an example. On the one hand, scientific-based predictions let sports bettors fine-tune their bets, but on the other hand, the mass amounts of information, data, and statistics that sports betting operators have access to, along with modern computers, have made it much more difficult for sports bettors to earn a profit.
Today, a single bookmaker's site can offer a dozen features and options for users' bets: you can take out your bet until the end of the match on cash out betting sites, watch matches live and analyse statistics with bookies offering live streaming, and so on. And even then, there's no way to turn bookmakers' odds against them? Well, we’re not quite there yet, but a quick glance at some studies and experiments conducted in recent years shows that there is hope.
A Statistical Theory of Optimal Decision-Making in Sports Betting
Jacek Dmochowski is an Associate Professor with the Grove School of Engineering in New York. He called his study A Statistical Theory of Optimal Decision-Making in Sports Betting. Dmochowski’s theory was that estimating the median outcome of a match is more successful than estimating the average outcome. This goes against the logic of a lot of average bettors who focus on averages over medians.
Dmochowski used an NFL match between the Philadelphia Eagles and Kansas City Chiefs to illustrate. In his scenario, the Chiefs had won the three previous games by margins of 3, 7, and 35. This means that the average margin is 15 points while the median margin is 7 points. The bookmakers have listed the Chiefs as 10-point favourites. According to his study, the optimal decision is to take the Eagles at +10 because the median margin is 7.
As an aside, several other scholars including Fabian Wunderlich and Daniel Memmert of Germany’s Sports University of Cologne have done similar studies and generally agree with Dmochowski’s take.
Team Tokyo
A few years ago, Lisandro Kaunitz of the University of Tokyo along with several colleagues wanted to take a scientific approach to turning the tables on bookies. The team pored over extensive data from around 500,000 football matches that had taken place over the previous decade. They also examined the odds that 32 bookmakers posted for those games. When they looked at a particular match, they searched for the best odds. As such, successful bets with the higher odds would net a higher return.
When they put their theory to the test in a simulator, it earned a small profit of 3.5% as opposed to the simulated random bets that lost 3.32%. Incredibly, the team successfully applied this strategy in real life. They placed $50 bets about 30 times per week for five months and ended up making over $900 for an impressive gain of 8.5%. The funny thing is that they wanted to continue testing their method, but several bookmakers limited their accounts, effectively ending the experiment.
Predicting the Future
Data science, statistics, and artificial intelligence are playing a starring role in science’s attempts to reliably predict the future. Right now, a bunch of brains from some professional sports team are feeding a plethora of information and video clips into a high-powered computer. The computer will then tell them exactly how their team and players should attack and defend.
It takes a split second for these programs to identify individual and team strengths and weaknesses. The bookmakers also use volumes of data to predict results so that their odds and spreads are accurate. Meanwhile, scientists and punters alike are scrambling to turn the odds against the bookie. It’s almost certain that these approaches are also being used in medicine, finance, and lots of other areas including war.
Will They Succeed?
They might level the playing field a bit more, but it is currently impossible to consistently and accurately predict sports outcomes. It doesn’t matter how much information you have or how powerful the computers are. This is because sports are rife with random events and chaos. Balls take weird bounces, refs make odd calls, and players sometimes take an untimely and unexpected tumble. A lot of unpredictable things can happen. Science can explain those oddities, but it can’t predict them. Well, not yet.
