Jason Jones, Sportsbook Manager at Aspire Global, examines the role of data amid the esports betting boom, looking at how AI technology can support the delivery of an optimum customer experience. 

Esports betting is still in its infancy, compared to other more established products within sportsbooks. With younger demographics having now grown up with competitive esports events as the norm, though, it is imperative that operators get on board quickly to establish a strong position within a realm of betting that is only headed in one direction – and it’s up.

The pandemic played a crucial part in accelerating the widespread appeal of esports betting and ever since, it has gone from strength to strength to move firmly into the mainstream.  While basketball and football simulations – NBA 2K and FIFA respectively – became popular overnight with fans during the pause in live sport, the scope of products and betting markets, both pre-game and in-play, has continued to blossom to become an integral part of the fabric of the overall sports betting experience. 

The potential of the more traditional esports events in drawing a different and previously hard-to-reach demographic to operators has also never been more evident, with CS:GO, LoL and Dota 2 tournaments attracting plenty of attention from bettors worldwide.

We firmly believe that the growth trajectory of esports betting will soon make it one of the biggest verticals in the near future, yet still with plenty of capacity for further growth given its huge potential within emerging markets.

Clearly, the widest possible range of esports betting opportunities needs to be successfully offered within a modern sportsbook platform in order to capitalise on the vertical’s huge growth. One of the keys to unlocking its potential, however, lies in the reliability of the data. 

Challenge

A lack of information about team members below the top tier events remains a major challenge for providers. While in the ESL and the DOTA Pro Circuit, information is credible and data widely available, tier-two events can take place anywhere in the world from Japan to the West coast of the US, and it can be difficult to keep up to date with the rosters of teams that are semi-professional. At any time, there could be a match taking place with players that traders have very little background information on, given the high turnover of participants. This is even true in tier-two events in North America – one of the more developed scenes worldwide – and it makes the prediction of outcomes very difficult.

At this level and below, official data distribution can also be lacking. That leads to a lot of fluctuations in the market and a large number of betting suspensions while traders try to figure out what is happening. Another sticking point is the broadcasts of matches, which can be difficult to locate. In the CS:GO ESEA Premier tournaments, most matches are streamed by players themselves and smaller tournaments rely on inexperienced, unknown casters, which can lead to flawed outcomes.

One particularly interesting aspect of esports trading right now is that a lot of odds providers are not necessarily using official match data. Instead, they are utilising computer vision to interpret what is happening by recording a scoreboard and inventory from a live stream, with events on-screen used as a data source to trade the match. The problem with computer vision in esports, though, is that those traders using official data are able to offer better uptime in-play, as the data is more stable. That means a superior product offering for the operator using that official data and a better, less frustrating experience for the end-user.

Invaluable

The ideal scenario is to pair an AI solution with official data to deliver greater uptime, particularly towards the end of matches. For a lot of providers using computer vision within DOTA 2 or League of Legends, when the action gets very complex near the climax of a match they cannot maintain late-game markets. Premium data providers are now able to calculate what is happening using their model alongside official data to provide greater uptime and a superior in-play betting experience. None of this is to suggest that the human touch of the trader is not relevant. In fact, this can be invaluable in protecting the operator if and when certain integrity issues might arise that would not necessarily be flagged by machines.

We expect the reliability of the data behind esports markets to improve as investment increases throughout the sector. In this developing vertical, maintaining the interest of bettors is invaluable for operators seeking to gain market share. Esports is now mainstream and the delivery of the best possible product, backed by the best data sources, is a non-negotiable factor, just as it is with any other betting vertical. 

At Aspire Global, we believe passionately that fans of esports deserve a range of betting products and markets every bit as wide-ranging and sophisticated as those served up to real-life sports enthusiasts, and we will continue to do everything we can to push the industry forward in providing exactly this.