With a background in physics, the new head of OPTX’s Data Science team hopes to take artificial intelligence (AI) in the gaming industry to another level in how it predicts the impact of changes to the casino floor.
Steve Bright, vice president of Data Science for the Las Vegas-based data-platform company, is leading a team that designs, develops, tests, and monitors new AI features for OPTX’s data modules. The algorithms developed and used by OPTX create actionable recommendations for casino operators, provide real-time individualized insights to ensure that no actionable player is missed, and empower teams to spend less time compiling the data and more time implementing strategies that increase guest visitation, revenue, and profitability.
Bright said he wants to build out OPTX platforms in the future to make them even stronger. Going forward, Bright said the goal is to use AI everywhere it makes sense, including player AI.
“The non-AI portion of the player module gives you insights and useful information about a player’s history,” Bright said. “The AI version uses the player’s history, plus all of the other information about the player’s habits and all the interaction points OPTX gets. Instead of looking backwards, we want the AI version to make forward-looking projections, predictions, and recommendations.”
Oftentimes, properties make decisions about segmentation based on a player’s previous spending, Bright said. But with AI, they can now make predictions about future spending and how a player will respond to an offer.
“AI allows you to skate where the puck is going to be, rather than where the puck was. On the slot side, AI can tell an operator how a proposed change to the floor configuration is likely to affect overall performance, by creating a model of the overall floor. Our objective is to supplement and augment that backwards-looking reporting with forward-looking predictions and recommendations.”
Bright said he ultimately wants an AI feature that will pick out the most important information from reports provided to operators, flash those highlights, and provide context and direction.
“There are a lot of reports in there, and a client is not going to necessarily look at every single one. But with an AI script, we can help the operator read through their own reports and point out the highlights and the impactful insights, then boil those up to the top,” Bright said. “As we develop that feature, it’s critical that we get feedback on it. When I’m writing an AI script that will tell the client what I think are the most important things, I’m not right in my assumptions. The AI is not right 100% of the time. It’s important that we have a way for the user to tell the AI when it’s wrong or when the AI has missed something or doesn’t have access to the data that the human being does. Having that human in the loop is crucial and we continue to bake it into every OPTX AI feature. We want the OPTX product to make the client more effective and efficient. We want to take advantage of their knowledge and their feedback to make our AI even smarter.”
The future of AI, not only for gaming, but for society at large, means a more thorough integration of players’ touchpoints everywhere they react with a product or brand, Bright said. Data scientists will have a more complete picture of a person.
“That can sound a bit scary and I think the industry needs to be mindful of not making it too creepy,” Bright acknowledged. “You do that with rules around data governance and privacy. Civilization is long overdue for a frank conversation on how far we want to go with this. What guardrails do we think there need to be and what constraints do we want to impose on AI and data science as they evolve?”
Prior to joining OPTX, Bright”¯worked as a data scientist in many different industries and has published more than 30 papers in fields ranging from experimental high-energy physics to pharmacovigilance (drug safety). He has several years of experience applying data science to problems in the gaming industry, including machine- learning algorithms to forecast and optimize slot-floor performance.
“I worked for a quantitative hedge fund based in Atlanta for five years and that’s where I got my first taste of data science and applying math to business problems in a structured way,” Bright said.
In 2012, Bright went to work for IGT in Las Vegas, his first exposure in the gaming industry.
“We tried to do some of the optimization work, but the data infrastructure wasn’t there,” Bright said. “This notion of an end-to-end data pipeline and presenting AI findings to a user was a technology that didn’t exist at the time. Now it does and that’s what pulled me into OPTX. All that fun stuff I wanted to do when I first started working as a data scientist in gaming, I can now do. The data, the platform, and the data-delivery method are all there.”
Bright said it’s not about trying to predict what an individual will do, but by applying data science, you can predict how a large group of people will behave.
“For as much data as we can get on them, individuals are so unpredictable. A poker player, Mike Caro, invented the ‘Law of Loose Wiring,’ which refers to people’s choices not being clearly connected to their outcomes. I think that’s the case with gambling. There’s so much randomness that individuals tend to behave in unpredictable ways.”
Bright holds a B.S. in physics from the”¯Georgia Institute of Technology”¯and earned his M.S. and Ph.D. in physics from the”¯University of Chicago. That background helps him apply the scientific method in approaching problems. “People forget about the science part of data science,” he noted.
“That means understanding the domain and forming some hypothesis – understanding the question the client wants to ask and translating that into an experiment that can be answered by data. That means collecting data, analyzing it, and faithfully, truthfully, and rigorously examining that data, questioning your assumptions, and thinking critically about your conclusions. Something scientists have to be aware of is that you’re the easiest person to fool. We want to double- and triple-check our assumptions before leaping to any conclusions. That’s how science works and what I did when I was working as a scientist. That fundamental approach is what makes data science a real field and a real science.”
Data science doesn’t mean drudgery. It can be fun, Bright insisted. “It’s exciting to see something no one has seen before, communicate that to your colleagues, put those findings into action, and make an impact in the world.”
Bright is a recreational poker player and gambler, an avid runner, and a long-time resident of”¯Las Vegas.
