Bayes Esports and ESL Gaming have recently expanded their partnership, adding more titles to the existing ones, including Dota 2 and Counter-Strike: Global Offensive, and vowing to “drive esports forward.”
Both companies have played a distinct part in achieving just that, but they are looking to further enhance their collaboration and set “uniform standards” that allow esports to thrive and develop. Uniform standards are precisely one of the things that needs changing, argues Bayes Esports CEO Martin Dachselt to see the esports ecosystem thrive.
Bayes is making it possible for sportsbooks to access accurate and correct, real-time live data and replace classical statistical models that no longer work in esports, and there is a lot the company is already doing to change ossified data collection.
Today, we get a chance to hear from the other side and speak with Martin, who can better explain what these uniform standards are, how media and set standards play a part in the further successful development of esports and where the limits of the industry are if any.
Q: Martin, how important is big data in setting proper standards in esports, and what standards are we talking about specifically: engagement, player and teams' chances of success, something different, or all of the above?
You might be wondering why something that is available for free on YouTube or Twitch can still be worth money to the media and betting industries. There are two reasons for this: First, you cannot directly translate data from video streams into the numbers that many media partners will want to display as statistics or on live tickers or that many bookmakers need to generate their odds.
Second, remember that the video broadcast is usually delayed by anything from 45 seconds to five minutes. For the betting industry, having access to undelayed data is crucial to minimizing risk. As a result, your event data can be valuable if it is marketed correctly.
The betting industry still widely uses classical statistical models to predict the outcomes of traditional sports. This is no longer possible in esports, where the amount of information per time is infinitely larger. This is why Bayes Esports uses AI algorithms like neural networks and gradient boosted trees for our predictions.
Q: How important is media presence for esports? Is esports still a niche form of entertainment despite the huge surge in popularity we have witnessed in recent years?
Currently, media dealing with esports is not yet able to address and inspire a large group of casual gamers, despite the pre-existing fan community. For an entertaining live experience, different visualizations and perspectives on-site and within the live streams are essential, not only to understand and enjoy but to feel like part of the game – especially in complex 3D games. In order to provide these advanced features, accurate and correct, real-time live data must be available – especially in esports, which are much more agile and fast-moving than traditional sports.
Our parser CARP, developed in close collaboration with ESL, allows the extraction of data for the games CS: GO and Dota 2 in real-time and with unprecedented quantity and quality of data points. Thus, the media can display complete 2D visualizations (replays) live and during the game, with the representation of player positions and trajectories of objects. This was previously only possible after the game, i.e., not in real-time.
Q: Blizzard recently scaled down their live events in light of a difficult 2020. Do you think live events are important to the success of esports as a continually developing ecosystem?
Short answer: Yes. Hardcore esports at the very top end relies on physical events and international competition – where the world's best teams from each region play against each other. This is the pinnacle of competition and obviously was not possible because most of the largest competitions did not take place last year.
While the competitions were converted to online setup, it still had a negative impact on the esports industry as a whole. The fans want to visit such events in real-life to exchange with other fans and feel part of a larger movement. League of Legends Worlds last October was the first such major event since March 2020.
This year we might see fewer injections of investment into esports as many economies attempt to rebuild out of a recession. This might threaten startups, especially those that are not yet self-sufficient.
When it comes to the crisis content, this is almost exclusively relevant due to a lack of normal sport. As more events come back to normal, the interest in this content will wane, and sometime in the future, the product will likely land on a kind of always-on product where there is something available during usual downtimes for traditional sports content.
Q: How can esports hosts benefit from reaching out to Bayes Esports and rely on your expertise in the sector?
I'd say there are three big buckets. First of all, we're trying to introduce standards to the industry where there are none. That goes from establishing standards in league operations to building the systems that keep “the clock synced” between matches (and their delays) and the various data consumers in the space.
Maintaining the order for data delays to make sure betting odds change before the score changes on Google and before viewers (and in some cases competing players) see it on Twitch sounds very easy but is an incredibly sensitive process. Especially in a world where systems aren't natively built to be connected, and standards are not properly defined. We, as Bayes, are absolutely unique with our approach to building long-term partnerships exclusively. This allows us to establish standards and provide our customers with reliable long-term data delivery.
Secondly, we're trying to move away from legacy technology where possible. Last year, we established BEDEX as the market-leading in-game data delivery platform. This year, we're expanding in all directions. We're building a brand new betting data product from scratch, we're building a new booking and reporting platform that automates lots of work for our clients, and we're working on a set of smarta data infrastructure pilots with our key data partners. Most parts are “top” secret, but I will hear about them in the near future.
Together with our sister company Shadow.GG, we're expanding the product line up for data visualization and tools. We're talking live data score widgets that allow for a more immersive media consumer experience, and we're talking visual data tools for the betting industry, as well as an expansion of our esports team training tools. Visualizing game data, making it accessible and consumable is a big driver of engagement with the fan base, with professionals, etc. We've started working with global media companies, e. g. L'Equipe in France, to attract new target groups.
Q: What do you think are the biggest challenges in the further development of the esports ecosystem?
What has to happen above all is regulation and standardization. Esports lack a uniform structure and format. The processes, rules, and data of most are not standardized, as its, e.g., the case for soccer. The industry needs game publishers like Riot and top league operators like ESL that show strong involvement and have a long-term view of the ecosystem.
Only with that kind of perspective, we'll create a level playing field where better skills win, and exceptional players make their way to the top through consistent performance, and where esports is presented as a sport rather than just games. Changing this perception in the public eye has started but will take a while, as we will see teams training and being managed professionally to be successful. In Asia and the US, this is much more well known already than in Europe, for example.
Q: What are Bayes Esports set goals for 2021?
We are the market leader in esports data, period, not by subjective perception but by any whatsoever metric. Achieving this within just 1.5 years is an unreal achievement that certainly neither side could have accomplished by themselves.
By the end of 2021, Bayes will be one of the top companies in esports and (still) one of the coolest places to work in the industry. We will make revenue in the double-digit million range with a team of 50 staff. Bayes Esports will be the definition of data services in the esports industry, synonymous with professionalism, reliability, scale, and customer focus. Staying in the number one spot can be harder than getting there, especially in an industry where everyone claims to be the leader.