The power of AI in sports betting: Charles McGarraugh of Stratagem

Predictive analytics company Stratagem has tapped into the sports trading sphere with its AI-powered portfolio of products. Its flagship offering is a trading platform that leverages data collected via machine learning techniques.

A background in financial trading has allowed Stratagem CEO Charles McGarraugh to use his knowledge of risk management and the markets to build a system catered specifically to operators.

He explains to EGR Technology how he thinks the sports markets are likely to evolve, and the impact AI will have on sports betting.

EGR Technology: What is Stratagem?

Charles McGarraugh (CM): Stratagem is a predictive analytics company. We have a machine learning first approach, and that’s our core expertise. We do a couple of things with it, so we’re focused on sports and the markets around sports. At the moment that means soccer, tennis and basketball.

We look to automate trading around those three sports and we have a couple of different products that either are live now or we envisage to go live in very short order.

EGR Technology: How does your product work?

CM: The core technical background of the company is basically a big data pipe. At the one end you have a funnel that takes in a lot of sports data and market data and then data from our models and data from our team of football analysts. That’s internally generated data and externally generated data in a big funnel. In the middle of the pipe you have signal processing so we try to get all that data interacting with each other and then draw predictive insight from the data, whether that insight is around the action in the games themselves or the markets around the games.

Downstream from that we have the infrastructure for automated trading. One product is asset management, so we are proprietary traders in the sports market. The next product is price and modelling feeds, so as an operator you can access our model signals and potentially even do a managed service, so ARIA (American Risk and Insurance Association) is working on your risk management.

EGR Technology: What makes your AI-powered trading platform unique?

CM: The first thing that makes up this system are models on the action data, so predictive analytics around the games themselves. Then there are also models around how the markets move and then there is the actual trading pipes. So an ordered management system, post trade risk analytics, live risk, all that kind of thing. And algorithms that route trade executions efficiently.

We do a lot of our actual processing in-house because given how specialised what we’re doing is, it’s actually a lot cheaper that way. Most of our underlying technological backbone is open sourced and conforms to what you’d expect from a modern data science company.

Under the B2B bucket as well we provide a streaming insights product which looks at data on the game and on the markets, and tells you interesting things about them and that can either go directly to our platform or an operator can purchase it effectively like a bet-prompt style content that’s close to the bet slip.

EGR Technology: How can AI revolutionise the sports betting industry?

CM: I think machine learning can mean different things to different people. We tend to think of it in a technical sense to mean adaptive systems that learn from data in a dynamic way, over and above what a traditional stats-based approach might entail. In that context, two things can happen: you can do what you’ve always done as an industry, but do it better, and the second thing is new product can come from it. They both come from the basic concept a lot of technology businesses in lots of different industries fall under, which is that software is eating the world.

In this case we think of it as a four stage process. The first thing that you do in the context of existing products and processes is automate out cost and the next stage is optimising the top line, In terms of risk management and setting prices, after that is when it gets a bit more interesting. When you start asking what you can do that’s different with the product.

We’re just starting down that path now, and then from a business perspective as the product evolves and the consumer demands a different kind of experience with the product, what does that mean for how the market structure will change?

From our perspective we think the automating out of cost has happened across the industry with the advent of in-play and mobile already, although there’s more to do. Top line optimisation has happened already to a degree but will continue.

CEO, Charles McGarraugh

EGR Technology: In what direction is sports betting moving next?

CM: But in terms of where the product’s going, I think that’s where what’s possible can be pretty exciting. In general our view is that fundamentally sports betting is an entertainment product that enhances the entertainment value of consuming sports media. The question is, as a punter if I have money on then it’s a massively more engaging experience to watch a game. The question is how can the product deliver more entertainment value to the consumer, and help keep the wagering product competitive against other competing offerings, and equally, and maybe even more importantly, how can gambling make the sports product stay competitive and engagement against other competing entertainment offerings. We think it’s about the core of the market and bringing sports content closer to the product and bringing an augmented experience to watching the game.

The space lends itself well to machine learning and AI because there is a lot of data and there is a high sample-size of things to look at. Because sports are a relatively structured environment it lends itself well to AI. This was a great opportunity to get my hands dirty with an AI company.

The sports markets are fascinating. Liquidity is fragmented and the product has scope for improvement. At the end of the day it’s all about risk management. Data and money and a high velocity of throughput require very similar mind-sets to confidently manage the risk.

EGR Technology: How do you stay ahead in the industry?

CM: Like every start-up we get out there and talk to as many people as we can and try to find out what people might find useful and square that against what we think is actually technically feasible. Then you can make a product market fit. In terms of our internal development process, we have research pods that are sports specific, and then we have an infrastructure team that sits alongside them and is involved in the development of whatever they are trying to do. Our goal is to scale the research effort, the power of the machines. Everything we do is about scalability and infrastructure.

Anything that makes access to data easier and broader is something that will fuel innovation across the whole sector. By increasingly the pace of innovation the single best thing that can happen is just making a lot more data a lot more easily available.

EGR Technology: What are the benefits of operators integrating a variation of third-party sports betting solutions?

CM: I think at the end of the day there’s no one approach that’s right for everybody, it depends a lot on what the business looks like. But that said sports betting and sports prediction is a game of skill and if you’re on a thin margin, as the world gets more competitive and transparent, you’re going to need to have experts to help manage that risk. It’s something that lends itself well to specialisation.

EGR Technology: How do you see sports markets evolving in the future?

CM: I think what’s likely to happen is the markets are likely to become more liquid and liquidity is likely to get more centralised through time. Probably increasingly resulting in a division between retail operators who are really focused on marketing and generating volume and then specialist risk management houses which are really focused on committing the balance sheet to the risk. I think the way that technology will continue to drive product just lends itself to that kind of evolution in the market structure.