The gambling industry has embraced many new technologies over the years, from the advent of the internet right through to mobile gambling and, in some instances, cryptocurrency.
One such phenomenon that is becoming more common in the market is machine learning. This is a type of artificial intelligence (AI) that uses data to make predictions and decisions without being explicitly programmed to do so.
There are various different algorithms that can be used for machine learning, and they all work to automatically improve themselves through the collection of more data.
The idea of machine learning has been around since the late 1950s and was originally a subset of AI, however the two fields diverged in later years as AI research and technology began to focus more on symbolism, while machine learning was concerned with statistics.
It wasn’t until the 1990s that machine learning began to properly grow in its own right and started to be used to solve practical issues. The key way it differs from AI as a whole is that machine learning makes passive observations, whereas AI is more active.
There are several ways machine learning could be applied to gambling both by the customers themselves and betting companies. This range of casino games at DraftKings, for example, could one day be optimised to become as engaging as possible for consumers through the use of machine learning.
These suppliers could use this AI to monitor the behaviours of their customers and gain a better understanding of their betting practices and preferences.
This can help the betting company more effectively target its marketing efforts, for example through offering bonuses relevant to a particular type of customer or sending emails and notifications at certain times of the day when they know they’re more likely to make a bet.
It can also help these suppliers provide a better service, as machine learning can predict the needs of customers beforehand, allowing the company to pre-emptively make changes and improvements to their platform.
A more impactful use of machine learning could be to tackle the issue of problem gambling. Research and studies into this have been going on for a few years now, but the theory is that this technology could look at the data of a person’s gambling activity and predict whether it is going to become a problem for them.
This would help suppliers take more of a responsibility in protecting their customers and to reduce the chances of more people developing an addiction to gambling.
In fact, casinos and gambling companies use machine learning to try and stamp out instances of fraud and cheating by picking up on anomalies and flagging irregular instances. These same algorithms and the data they collect could also be used to protect the wellbeing of customers and determine whether they are in full control of their gambling.
Unsurprisingly, some people have tried to find out how they could use machine learning to win more money from betting suppliers, though the results have been far from encouraging.
Rather than trying to just correctly predict the outcome of sporting games and events, these studies instead tried to implement machine learning to predict instances where a bookmaker has offered some erroneous odds; i.e. they’ve made a mistake and priced a comfortable favourite with generous odds.
While this algorithm did find success in correctly predicting such eventualities, the team in charge also found that these occurrences of the bookmakers getting it wrong were rare and that using such a machine learning model would take a long time to make significant profits.
As outlined by others, this is in large part because the bookmakers set the odds and so determine the payouts, meaning they are always at an advantage. While these odds are decided by teams of experts, many companies do also use some form of machine learning to influence the prices they put out.
That being said, there is plenty of potential for people to use machine learning to stack the odds slightly more in their favour. Major sports like soccer and American football have stacks of data available online, which could then be plugged into a machine learning algorithm.
This could, in theory, help punters get a much clearer idea of how games are likely to play out, thus providing them with the chance to make some decent returns from gambling.
If this became a widespread tactic, though, it’s highly likely the gambling industry would need to respond and limitations could be set on how and when machine learning can be used.
So far, the impact of machine learning on gambling has not been huge, though there is certainly room for it to grow. Suppliers could implement algorithms to accurately predict how their customers are going to act and prepare accordingly.
Likewise, there is scope for people to use machine learning when placing their bets by highlighting where bookies may have misjudged some odds, or by trying to work out how certain events will go based on available data.
The developments of machine learning in gambling could be fascinating, particularly if they are put to good use in the fight against gambling addiction, and by improving the overall experience for bettors.