Big data, huge potential: Gaming sector could see great gains from acting on rich analytics

Shutterstock.com

What would the “Moneyball” casino look like?

As a Ph.D. candidate studying big data at UNLV, Ray Cho has given some thought to this question. Cho, who has 20 years of experience in the hotel industry, works as an analyst manager at American Casino & Entertainment Properties.

With advancements in computing and machine learning, businesses across different industries are exploring how to better use data to increase performance, revenue and customer satisfaction. Oftentimes, the issues are less technical and more political. It’s one thing to crunch the numbers. It’s another to get management to act on them. This is especially true when data contradict operating procedures entrenched as industry standards.

“If a team of analysts was building a casino on its own, it would look a lot different,” Cho said.

The gaming industry has collected information on players for decades, Cho noted. But he said there is more potential to apply the data to make more specific and targeted decisions, from daily operations on casino floors to bigger-picture directions chosen by chief financial officers. It’s not necessarily about collecting more data, he said. It’s about how you use the data you have. “The key word is efficiency,” he said.

What is big data?

According to Forbes’ “jargon-free” explainer: “It all starts with the exponential explosion in the amount of data we have generated since the dawn of the digital age.” Imagine vast sets of such data. In order to analyze them for associations, patterns and trends, you need big computing power.

“Moneyball,” a 2003 book by Michael Lewis that was adapted into a critically acclaimed 2011 film, focused on how a MLB team used big data to analyze player performance and behavior to become much more competitive.

Machine learning is an important tool. It’s a kind of artificial intelligence enabling computers to learn without being specifically programmed. The idea is that when these programs are exposed to new data, they can change spontaneously.

For instance, as data is collected on players, the computer becomes better and better at predicting their behavior. Machine learning also could be used to better understand specific customer demographics or analyze how players are using slot machines. Armed with such data, casino companies could anticipate what these or similar guests might play in the future.

“This is very similar to what Netflix does when they recommend a movie for you to watch on Saturday,” said Ralph Thomas, the chief data scientist at VizExplorer, a company that helps gaming companies work to boost their efficiency and productivity with the use of big data.

With analytics for slot machines, table games, marketing and player development, hotel management can make more intelligent operational decisions, Thomas said.

“We’re starting to use (data) science to determine how we can improve our interactions with customers,” he said, adding that such information can be used to optimally target offers to players in the hope they’ll spend more.

The company, which lists Boyd Gaming as a client, uses machine learning to delve into slot behavior.

“There are all sorts of ways slot machines can behave over time,” Thomas said. Analysts for VizExplorer also model aggregated data from ATMs (minus the personal information) to give clients a sense of how gaming on a casino floor is correlated to ATM use.

The tricky part, Thomas and others said, can be getting management to use the data. “You still have operators that only want to see the daily operating report,” Thomas said. “They don’t want to see anything else.” He added that this plays out in many industries. “It’s a culture change.”

Like other industries that require customer information, the gaming industry has a formalized way of collecting data.

Alex Bumazhny, an industry analyst with Fitch Ratings, said one of the big breakthroughs in data collection came in the late-’90s, when Caesars Entertainment created its Total Rewards loyalty program. Many gaming giants have since followed Caesars’ lead by launching their own loyalty programs. Until then, companies were not taking full advantage of what they knew about their customer base — where they liked to eat, what they liked to do.

The Total Rewards program allows customers to quickly collect points that they can then spend on other amenities. This keeps guests in the casino and leaves the property’s leadership with personalized information to aid in making decisions about offerings down the road.

At the time it founded the program, Bumazhy said Caesars’ portfolio of assets in Las Vegas was older than those controlled by other gaming companies. “They were trying to do more with less.” he said.

Total Rewards was so successful that creditors saw it as Caesars’ most valuable asset during its bankruptcy. Creditors, who wanted a slice of the program, said the data was worth $1 billion.

Not all gaming companies have turned to big data, and even fewer use data to make operational decisions. “When you do data mining, you put a lot of old operating models on trial,” said Tony Lucas, who studies data as a professor at UNLV’s Harrah College of Hotel Administration.

That can make management wary.

“The challenge of big data is actually affecting change,” he said.

Lucas said that once you start digging into free-play programs, the data often show that they are not effective, given the large amount of marketing dollars they suck up. The same is true with lottery promotions, he said. But because they are ingrained in the system and because companies have sunk so much into free play, management is likely to push back on proposals to eliminate it.

“It’s political, uncomfortable and difficult to take on that existing operating theory,” Lucas said.

Cho said there often is a split between operations staff and analysts.

“Getting those two camps to work together is not always easy,” he said.

But the gap is getting smaller as more companies consider data-based decision-making.

Would the “Moneyball” casino look all that different? Maybe not. Cho said a lot of people toss around the buzzword “big data” with the misperception that it’s going to re-create the business model. In many cases, analysts work to refine existing processes — to make them more efficient.

“Good analytics happens a little behind the scenes and under the hood,” Cho said.

Gaming

Share