GAMBLING FOR ROBOTS

By

Dr. Steve Bright, a particle physicist by training, heads data analytics at OPTX and has been charged with implementing artificial intelligence elements into the casino gambling experience. Gaming America asks him about the use to which this cutting edge technology can be put.

Could you describe to our readership what OPTX does and also your overall mission as the VP of data science?

OPTX is a casino analytics software platform, a delivery method for information findings and recommendations to casino operators. This information will help them operate more efficiently in every aspect of the business. We base our value around ingesting data from all sources in a casino environment. Our long-term objective is to get data from player and gaming systems, the machines themselves, from the hotel and the food and beverage side – everywhere a property interacts with a customer. In most casinos now, thosesystems don’t really talk to each other and are very siloed off from one another. What we do is take in all of that data and unifyit into one consistent schema across thesevarious functions of the casino and all thesevarious touchpoints with the player. This allows us to report back to the client, the casino operators, and gIve them a holistic view of their players and their property. We also give a host of reports that the operators need to understand their business at any state of time.

How does artificial intelligence fit into your overall mission as a company?

So right now we are looking to add AI features on top of all of this data analytics work I have just been describing. In fact, I was brought onboard to oversee our development in this area. By way of definition: AI is just any kind of computer that tries to mimic human cognitive function. That can be something really hard – like understanding and interpreting human speech or driving a car – or it can be something less complex, like recommending a movie on a screening platform. In all ofthese cases there is a computer algorithm that is just doing some math under the hood and, in doing so, providing actionable recommendations and insights.

We are trying to add AI features on everything. Basically, instead of just reporting a player’s past value, we’re looking to tell the operator that “based on what we know about this player, here’s what we believe their activity is going to be like in the future.” For instance, we can figure out how likely a player is going to be in responding to a particular offer or the type of offer that will be best for this player, or here is how we can automatically process this data and learn from it. The algorithms we will write will provide that curated and hosted experience.

I want to be clear: I don’t view AI as a replacement for humans, I view it as asupplement – a co-pilot not an autopilot.
We want AI to provide additional facts and contacts to the provider, but ultimately the human operator is going to do what the human operator is going to do. I think it’s wise to have an AI that is more of an assistant to the human.

It’s also important that our AI features collect feedback from the users. The most important piece of information that you get when you’re building AI features is knowing how the human you’re intending to help is reacting to your AI recommendation. Sometimes the human is going to take those recommendations and implement them, sometimes they’re not going to. You want to know the cases when they are not because that’s really some useful feedback as the human might have some information that your AI does not. Ultimately you want the AI to help the property get smarter, and the human to help the AI get smarter.

OPTX is currently working more with land-based casinos rather than in the online world, but I’m wondering if AI looks differently for land-based operators versus online ones?

The underlying data should probably be richer for online. I would imagine that, for an online operator – and I haven’t yet seen the data, so I am speculating here – you would insist on collecting much more detailed information about gameplay. That means things like: how much is a player betting each time? Do they change their bets with regards to individual actions like winning or losing? I think that sort of information about how humans behave at the micro-transactional level is data that we don’t really have for the most part in land-based casinos.

In a land-based casino most of the data wehave is at the session level: where you know that this person came in, stuck their card in, played the game for 20 minutes, pushed the button this many times, won this amount, lost this amount. But you generally don’t get the individual handle/pull-level data. I think having modelling, understanding, of that human behavior would be super useful for the folks who design those games.

In addition, with land-based operations you are going to also have hotel data and food and beverage data, something that’s not a thing for online casinos. And it helps the bottom line for these businesses. Casinos sometimes have offers that are food and beverage related. So, it would be useful to know the dining preference of your slot machine players or their musical preferences for certain events.

As a person with one foot in the present and another in the future, what does the casino of the future look like?

That’s a great question. Is the trend goingto be more toward online rather than traditional gaming? Maybe. I know that there are places where online gaming and online sports betting is widely adopted. I think, though, that there will always be a need for a Las Vegas-type destination. I suppose you should distinguish a Las Vegas of the future from a casino of the future. Gambling destinations like Vegas have people arrive and almost pretend to be somebody else for a week. I think people find that very fun and liberating, that they can cut loose from their everyday life. I think there’s an innate human need for that sort of vacation. As far as casinos go, I think it’s a fun group activity, and people will want to go and spend time with their friends and gamble.

If you had asked me this question five orten years ago, I would have said that esportswould be absolutely huge and dominateeverything. It hasn’t turned out to be the case. I don’t think the industry at large has really figured out the alchemy of games of chance combined with video-type games of skill; participation versus spectatorship and how to monetize that. There’s some really great business in there somewhere, some really great experience, I just don’t think we’ve really figured it out.

Latest News
Advertisement