The evolution of automated sports betting: In search of the magical formula

An early example was the use of the telegraph to communicate the outcome of horse races from the track to the bookmaker, which enabled people to intercept results of the races before the bookmaker had the information, a quirk depicted in the famous film The Sting starring Robert Redford and Paul Newman.

In the modern era, the internet has enabled bettors to access a wealth of information about results, betting process and bookmakers - you are just two clicks away from reading, say, a BetBonanza review Nigeria. The technology of sports betting, however, has moved way beyond the telegraph!

Early Days

Another sector of advancing technology has been the growth of algorithms capable of producing more precise forecasts. University College of London developed, in 1911, a technique known as "multiple regression" that enables individuals, starting with multiple potentially influential factors, to evaluate how they relate to a particular result.

In 1970, two researchers named Bolton and Chapman decided to bring multiple regression to thoroughbred racing in hopes of discovering how each distinct factor determines the likelihood of victory of a given horse. Due to the difficulty of obtaining data at the time, they were not able to complete their study until 1981, and it was not published until 1986.

Bolton and Chapman were able to analyse nine distinct factors using data from 200 previous races for this study. Although their research was indeed beneficial to the creation of prospective predictive techniques, it was not yet profitable enough for betting purposes.

Internet Boost

With the advent of the Internet in the 1990s and 2000s, it became possible to collect significantly more data on all sports matches, allowing the implementation of the first simple forecasting analytics for quantitative analysis of sports betting. These early predictions were capable of producing substantial profits for their creators, which prompted sportsbooks to create similar technology.

One of these was the famous 'Midas Algorithm' which was capable of producing the most accurate industry-wide forecasts and which led to a company called Cantor Gaming gaining an edge on betting syndicates that had previously been able to defeat bookmakers. This prompted professional gamblers to seek new methods to improve their forecasts and regain their lost advantage over the betting companies.

Big Data

The rise of big data has enabled what was previously impossible through the development of AI technology. This technology needs large datasets and powerful hardware, which has become more freely available in the last few years.

AI tech has started to shape the sports world as well as the betting sector. Operators such as Wyscout and Optasports have been founded to collect sports data. The collected data permits traders and operators to analyse every aspect of a sporting event and provides football clubs with technological skills for scouting new talent and enhancing their performance.

Until recent years, it was not possible to collect these data for primarily two reasons: the required technology was too costly and the market was not mature enough. That has now changed and in soccer alone, it is possible to gather thousands of data points for each game.

The AI Algorithm

Bayes's theorem, a 1763 hypothesis, can now be applied because the computing power and data required to run a Bayesian network are available, and this theorem is one of the main learning methods generally used to train an artificial intelligence.

It is now possible to construct machines that analyse all the specific events of every match, quantify how each is related to the final result, create event probabilities and consequently market prices, and implement a betting strategy that maximises returns.

Conclusion

All of this technological development comes with a downside. Such technology is out of the reach of most sports bettors, but the insights that it produces can be useful. And as the sophistication of automated betting models grows, there may be a growing edge for those who take a less automated approach, as opposing the crowd is often a good place to start in the search for betting profit.

#FeaturedBy: Kristina Regina