Machine Learning Responsible Gaming Systems Provide Industry With New Capabilities And Challenges


Responsible gaming systems have come a long way in the statewide U.S. regulated online sports betting market – but with novel capabilities come new challenges.

A recent article written by USBets author and Forbes contributor Matt Rybaltowski communicates how U.S. online sports betting responsible gaming tools – many of which are being launched by the Entain Foundation – are evolving with the times. 

Thanks to the introduction of new metrics that are aided by machine learning, online sports betting and iGaming sites are well on their way to identifying variables that can assist both companies and their customers in a preliminary form. In a way that helps spread awareness of problem gaming protocols as well as reduces problem gambling behaviors before their impact escalates. 

But with these new capabilities come a new list of variables that are necessary to consider when effectiveness is of utmost priority.

For example, a sudden increase in a deposit amount by a customer could be linked to problem gambling behaviors on one hand… but could also simply be a consequence of an increased personal budget or even a desire to take advantage of a promotional offer being extended through the regulated online sports betting portal itself.

If a customer makes three deposits during a particular month of $50 each (for a total of $150), then makes a $500 deposit during the first week of the next month… is that a human behavior that gets flagged by machine learning responsible gaming protocols?

Could the new, $500 deposit be in conjunction with a deposit reload bonus that the site is offering? Is that amount inside or outside the actual budget for entertainment expenses that the customer currently enjoys? 

And if the behavior, upon consideration of other site-owned intellectual property data/variables results in a prompt from the RG system, is it appropriate to contact the individual via email, telephone, system template communication, or human contact? Does the transaction get blocked in real time? Or is it allowed pending a timely follow-up from any one or more of the resources available to an online gaming or sports betting company?

Is the customer “chasing losses” due to a streak that resulted in all the available funds being wagered irresponsibly, or did “variance” simply catch up with the customer due to the amount of total (or types of) bets placed?

Or perhaps another scenario that may (or may not) trigger a response from an online company’s machine learning intellectual property…

What if a customer makes very small, periodic wagers during a major sport’s regular season games, but then significantly increases that amount during a professional league’s postseason and/or championship game/series?

For example, perhaps an online sports betting customer in the regulated market places a total of $200 in bets during each week of the upcoming NFL regular season, but then increases that amount to $1,000 per weekend during the playoffs, and/or to $5,000 in total bets during the Super Bowl? 

When, why, and how should such behavior be gauged when it comes to effective responsible gaming measures that are capable of both reducing the risk to problem gamblers while also ensuring the experience and enjoyment of other individuals who routinely increase their amount of risk online when professional sports contests are more important? 

Yes, a responsible gaming service provider will have many, many variables to consider (and hopefully improve upon with time) when making such determinations, but the need for human consultation and coordination still clearly exists (in this author’s opinion) even with machine learning tools. 

Rybaltowski‘s article in Forbes communicates this concept, with input from the National Council on Problem Gambling’s (NCPG) longtime executive director Keith Whyte – who warns that machine learning intellectual property will not result in a 100% panacea for problem gambling or the elimination of problem gambling behaviors.

Whyte also stresses that human interaction and consideration should remain a top operational priority for online sports betting and iGaming companies as they make the transition toward more machine learning tools and resources.

In the coming years, the landscape for how online casino and sports betting companies adapt to new capabilities regarding problem gaming behaviors will be something that both algorithms and humans examine in order to accurately gauge their true effectiveness. 

It will be, in this author’s opinion, of utmost prudence to continue prioritizing human brainstorming and talent in a way that enhances responsible gaming measures and gradually makes for a more well-rounded, coordinated product that can be exported to other online gaming platforms thanks to detailed analysis performed by both humans and machine learning algorithms.

Thank you for reading COBets.

David Huber is a seasoned writer with over 15 years of experience covering the online poker, online casino, and online sports betting sectors. He has performed various roles within the iGaming industry since 2006 including: reporter, editor, virtual assistant, podcast host, and lead forum moderator.

Comments are closed.