With a system based on AI and machine learning to provide quick answers

Gaming Analytics.AI is finding new ways to improve casino profitability

Kiran Brahmandam (right) says that by incorporating machine learning into the system, the quality of the queries and results constantly improves by making subtle adjustments as new data enters the system.
2020-09-22
Reading time 4:34 min
The Casino at Gulfstream Park reports that a standard monthly slot floor analysis used to take three days, while with GA it is complete in less than a half-hour. At California's Pauma Casino, they fine-tuned their drop schedules to reduce labor, minimize disruptions, and increase efficiency. GA tools analyze the middle of machines' performance range, which could move a game that is now performing at 1.2 times house average to 2.0, or even higher.

Few industries generate as much data as casino gaming. Theoretically, an operator today has the information needed to determine which guests have been moving up or down or remaining consistent with their betting patterns, which promotions generate the greatest return, and how important free meals, free play, and complimentary hotel rooms really are for different customer types. Given our current pandemic conditions, we have the data to determine which machines are the most effective to leave in play under social distancing restrictions, and how individual guest play could be impacted by those “missing” machines.

At least that’s what we should be able to do, because the information is there. But there’s a problem: either most of our analytical tools are locked in silos, or the data are simply too complicated or too overwhelming to translate into actionable strategies. Often the hotel system can’t effectively exchange information with the slot system, neither of them talk to the showroom ticketing app, and forget about tying into the golf course, gas station or gift shop.

Casino IT departments have addressed these issues by building data warehouses that aggregate everything in one place. These warehouses are invaluable, but it seemed for a long time that only data scientists knew how to extract answers from this massive cache of information, and those scientists were seldom experts in marketing, slot performance, finance, or hotel utilization. In other words, those that make operational decisions could seldom get the information they needed.

These were just some of the perplexing issues that Kiran Brahmandam noticed over his years of developing systems products in the gaming industry, according to an analysis by CDC Gaming Reports Inc. The data has always been there, but the expertise to get it out quickly was missing.

After leaving Bally in 2015, Brahmandam worked with a non-gaming startup, where he studied artificial intelligence and machine learning. It was there that he also realized that Google, and its more talkative friends Siri, Cortana, and Alexa, were changing the way that most folks accessed data. Complex answers were suddenly available just by typing or asking simple questions.

He wondered why this technology wasn’t being used in the casino world. Could a slot manager just ask, “which slot machines should I convert or change today without jeopardizing critical guest segments?” or “Should I leave an underperforming machine on the floor?” Could an executive host ask which players were most at risk, or a general manager say aloud, “what will be the budget impact next quarter of our current operating conditions?” to a device and get actionable data?

These are exactly the kinds of questions that Brahmandam knew a system based on artificial intelligence and machine learning could answer, and answer quickly. Best of all, such a system would be easy to use for those without a computer science degree.

Assembling a team of programmers in Berkeley, CA, and gathering input from casino operators, Brahmandam formed Gaming Analytics.AI in early 2017. Within a year, this subscription-based solution was in the field and producing impressive results. And the answers did begin to come quickly.

The Casino at Gulfstream Park reports that a standard monthly slot floor analysis used to take three days to put together. With Gaming Analytics.AI, the job is complete in less than a half hour. Likewise, the marketing team at Gulfstream began using Gaming Analytics.AI to look at the demographics and geographical location of their slot customers to learn which ones had the most potential and around which areas they should target their marketing and advertising campaigns.

Gaming Analytics.AI is even being used to improve what seem like simple tasks, such as when to drop certain machines. At Pauma Casino in southern California, they fine-tuned their drop schedules to reduce labor, minimize disruptions, and increase efficiency.

Finding new players is always a good thing. But keeping the ones you have can be even better. At Casino Trilenium near Buenos Aires, Argentina, they’ve used Gaming Analytics to target a group of players who recently reduced their play, also known as player churn.

“GA’s actionable churn predictions were remarkably accurate and helped us plan our reactivation campaigns to various segments of players based on their risk and future value. The results were amazing,” said Natalia Lenzbereg, who heads up marketing at Trilenium. Using the recommendations and insights provided by GA software, they experienced a 49% increase in repeat visits, increased spend by players, and an ROI of over 97% on the targeted group.

Using just a pad and pencil, almost any slot director can identify the weakest machines and the strongest performers, but few have the motivation or tools to explore and analyze the middle of their performance range. With Gaming Analytics.AI, this overlooked dimension can provide some surprising results. A denomination, theme, or bet change might move a game that is now performing at 1.2 times house average to 2.0, or even higher. Brahmandam said: “Using traditional methods, it was rare to even consider changing machines that were performing well. But that’s the beauty of artificial intelligence – it isn’t bound by conventions or traditions. It just looks at the data and seeks the highest return. It can show you new ways to make a good machine even better.”

He adds that by incorporating machine learning into the system, the quality of the queries and results constantly improves by making subtle adjustments as new data enters the system. At one casino experimenting with gender play patterns, Brahmandam says, “they saw their prediction accuracy rise from 87% to 92% and, eventually, 95%.

Eilers & Krejcik Gaming, which is a boutique market research and consulting firm focused on servicing the gaming equipment industry, closed on a strategic investment with Gaming Analytics in mid-August. “We reviewed several AI companies focused in the gaming sector, and we believe Gaming Analytics has the best technology offering,” cofounder Todd Eilers said.

As a cloud-based system, there’s no requirement to buy expensive hardware for GA, although it can be hosted locally if required. The software can be integrated easily with all major slot and marketing systems and is sold as a monthly subscription where the longer the term, the better the rates. For most, the paybacks are immediate, with improved efficiency, stronger marketing, increased profitability, and better predictions. A demo of Gaming Analytics.AI can be scheduled on their website.

This story was first published by CDC Gaming Reports Inc.

Leave your comment
Subscribe to our newsletter
Enter your email to receive the latest news
By entering your email address, you agree to Yogonet's Condiciones de uso and Privacy Policies. You understand Yogonet may use your address to send updates and marketing emails. Use the Unsubscribe link in those emails to opt out at any time.
Unsubscribe
EVENTS CALENDAR