There’s no doubt the role of the traditional odds compiler or sports trader has shifted significantly in recent years with automation and algorithm-led data becoming more pivotal to pricing. Arithmetic, knowledge and foresight, although still important, are taking more of a back seat to the needs of today’s sportsbooks.
These tend to be increasingly focused on liability management and customer profiling, ensuring consistent margins across a variety of sports, key to stakeholders and accountants. There’s nothing wrong with this, and there’s also no reason why the trader shouldn’t be able to do both with the tools now available.
With customer demands for a greater number of markets, and more personalised pricing mechanisms, it’s clear that these tools need to embrace the latest technology to ensure both quality, accurate pricing and functionality for the user.
One area we’ve seen take off in recent months at Sportsbet.io has been request-a-bet style betting. To manually trade this is impractical and, labour costs aside, accuracy would be erratic at best. Therefore, it’s important to find a solution, utilising trader knowledge, with supplier data and algorithms, to create a quality betting proposition.
This type of bespoke market, where bettors can decide whether Mo Salah will make have more completed passes than Juan Mata, for example, is so data-thirsty that they can only be driven by automated methods.
However, they are also highly volatile. This means they require a second set of eyes from traders to ensure there isn’t data error or information not factored in that’s impacting on the prices. Only in a relationship where the limitations of automation and trader skill is understood can a market like this be offered profitably. How this partnership develops in the coming years will dictate the industry’s bet offering.
Machine learning may also have a part to play in this. With customer profiling becoming more in-depth, automated pricing is increasingly possible, particularly for high-liquidity markets. There’s been a trend for traders being asked to follow larger operator prices and reduce arbitrage risk, but these prices have been led by liquidity elsewhere, and customer profiles likely very different to their own.
All this does is bring bookmakers to the centre ground and mitigates any notable risk-taking. Machine learning, led by in-house developed quantitative teams and trade input, can allow for bespoke prices led by algorithms unique to the individual.
This may be riskier, but the margins available to those who succeed are significant. Customers are becoming increasingly apathetic to which bookmaker they choose as they see odds comparison sites converging prices and reducing choice. Perhaps machine learning is the answer to this.
It will be interesting to see whether liability-based pricing like we see in Asian and North American markets maintains its appeal in the coming years, particularly as bets become increasingly quickfire and more personalised to the individual.
It’s likely we’ll see 1X2s, over/unders and handicaps in major sports still led by the vast money traded in those regions for the foreseeable future. However, we’re seeing individual, personalised bets, such as bet requests, coming to the fore elsewhere, predominantly led by the UK market.
This adds a new dimension to both the role of the trader and their operator employers. Whether this is a regional preference or a continued trend to carry into the rest of Europe remains to be seen. It’s certainly a challenge, but also an exciting time to be part of a trading team at a progressive operator.
Dominic Hardcastle, Head of Sportsbook, Sportsbet.io
Founded in 2016 as part of the Coingaming Group, Sportsbet.io is a leading bitcoin-led sportsbook operator: www.sportsbet.io