top of page
Vegas Joker Banner Gamomat

Q&A: AI Powered Odds With Jordan Peltier, Chief Data Officer At PandaScore

Q&A: AI Powered Odds With Jordan Peltier, Chief Data Officer At PandaScore
Q&A: AI Powered Odds With Jordan Peltier, Chief Data Officer At PandaScore

What role does AI play in generating odds for esports?

AI plays a crucial role in generating odds for esports. This is because it is incredibly effective at handling complex, high-dimensional and unstructured data and finding patterns within it.

These patterns are needed for us to be able to predict and forecast the likely outcome of events, and this is the foundation for determining the odds for each market we offer. AI can also adapt dynamically to new data, something that’s not possible with old statistical models like parametric distributions as these often struggle to capture the non-linear relationship between variables and so require more manual engineering which takes time and valuable resources.

For us, leveraging both AI and specialist human traders leads to the greatest result – you have models that can capture the complexity of esports but that are not black boxes with overfitting, supported by human traders with a deep understanding of each discipline.

Can AI alone generate accurate odds that deliver value to both operators and players?

Not at the moment. The best odds are based on true probabilities which simply means determining the actual probability the event will happen.

This is what allows operators to offer high-value odds to bettors while ensuring a strong and stable margin for the esportsbook. And while there have been major breakthroughs in AI, especially the transformer-based models such as the Large Language Models being pioneered by OpenAI – they still make errors, especially when it comes to forecasting.

If you look at it as a percentage, AI can get you 90% of the way to determining the true probability of an outcome and that’s why you need human specialist traders to get you the final 10% of the way to perfection. It must be remembered that AI models are only as good as the data they feed from and having 100% coverage and depth is easier said than done.

For PandaScore, perfection is a must. We need to be the best we can be everywhere at all times. It only takes one small error and sharp, savvy bettors will quickly exploit it. This is not what our operator partners expect from us, nor what we expect from ourselves.

By combining AI, solid data and the best esports traders in the business, we can deliver odds that are statistically sound while providing tremendous value to bettors and our operator partners.

Why is it important to combine AI/machine learning/etc with human traders?

Manual action is still required to deliver the most accurate odds to bettors and operators. In

particular, human traders provide a security layer for core actions such as settlement and odds-checking when the model goes off-market. In some instances, the trader has access to a key data point the model does not see or was not trained on.

This could be something like a last-minute change to the roster. Our human traders are also responsible for picking the tournament and market coverage that will be most appealing to our operator partners, and our traders are on hand 24/7/365 to support our operators in their efforts to offer the best esports betting experience to their players. These are things that AI simply can’t do at this stage.

How do your traders work with AI and other tools to generate odds for your partners?

Our trading and data science teams work hand in hand for the entire lifecycle of each AI model we build. This covers all steps from designing new models, testing (both new and existing models) and interacting with the models in production to ensure they deliver. This is a virtuous cycle, with our data scientists training the best models that prove to be the easiest for our traders to manipulate.

Our traders also give a lot of deep and detailed feedback for our data scientists to use to improve models. Ultimately, this means our trading team has little input once the model is rolled out – they tend to make one input per match and then all markets and odds are calculated by our AI models.

This is crucial to delivering our innovative, market-leading BetBuilder and PropBet products.

How important is it to have specialist traders for specific esports disciplines?

It’s really important. Each video game is like its own sport, and while there are similarities between some titles, to really deliver the top performing markets and odds to operators and bettors, you need specialist traders for each discipline. This means they can get super granular when it comes to things like patches and the impact they will have on our models. If it is significant, they can request for the model to be retrained.

Of course, the entire PandaScore team – including our data scientists – are esports fans which brings a deep understanding of the different esports titles we offer odds for to the organisation. Ultimately, we all work together to ensure our odds are highly accurate while delivering a superior, stable margin for our operator partners.

How does trading for one discipline differ from another?

There are lots of differences between games, especially between shooter-round based games such as CS2 and Valorant and MOBA games like Dota2 and LoL. With the former, there’s a lot of repetition in the gameplay which takes place round over round so betting markets are focused on rounds and kills for teams and individual players.

Our pricing team must master this key dynamic of these games, which they have absolutely done, and we have lots of data that we use to back our model and support our trading team.

MOBA games like Dota2 and LoL are a little more challenging as there are more objectives within the game that have a direct influence on who wins the match – kills, towers, nashors, inhibitors, dragons and so on. This kind of complexity doesn’t exist in traditional sports but that’s what makes this such an interesting space in which to create betting opportunities. It’s also why it’s so important to have specialist traders for each discipline as the differences between games are vast.

Anything else to add?

AI is a key pillar of our business. The revenue we generate depends directly on the performance of our models and that is why we are so committed to perfection and ensuring they are flawless.

As part of this, we always need to stay ahead of the game and that’s why we have built a team of the brightest minds, smartest traders and true esports fans with a genuine passion for video games.

This, combined with AI and other cutting-edge technologies, has allowed us to become the go-to provider for esports odds, stats and betting products.


Rated 0 out of 5 stars.
No ratings yet

Add a rating

iGaming News Delivered Daily

iGaming Gazette is your go-to source for the latest and most comprehensive iGaming news. Stay up-to-date with all the happenings in the vibrant world of iGaming through our carefully curated articles, breaking news updates, and exclusive interviews.

As a leading platform dedicated to iGaming, we pride ourselves on delivering accurate, insightful, and timely news that covers the entire spectrum of this exciting industry. Whether you're interested in online casinos, sports betting, eSports, or the latest gaming technologies, iGaming Gazette has you covered.

Join our community of passionate iGaming enthusiasts and never miss a beat with our in-depth coverage of iGaming news.

bottom of page