Esports Gambling: How Skins, Money, and Biases Shape the Market

Hi everyone,

Today, I would like to discuss a rather unusual topic: esports gambling. We will learn about the structure of esports gambling, how people tend to bet, and what causes them to do so. Happy reading! 🙂

Esports has long been one of the fastest-growing entertainment types in the world. As a consequence of this growth, the gambling industry has set its sights on esports as well. If you’ve ever watched an esports competition, you’ve probably noticed the many different gambling sponsors prominently displayed, particularly in Counter-Strike.

"What is... surprising is the fact that esport gambling transactions far exceed the revenues generated by the competitions [tournaments] themselves [1]."

When looking at gambling markets in traditional sports, biases and inefficiencies exist. A bias is a tendency toward or against something. In an efficient betting market, if a player has a 2:1 odd of winning a match, it would be expected that the latter player wins about 33% of the time in repeated games. Consequently, over the long run, a bettor would break even (excluding the bookmaker's commission). In an inefficient market, one side may end up better or worse.

These two aspects, biases and potential inefficiencies, may exist in the esports betting market. Furthermore, to date, we don't fully understand how the professional esports gambling market is structured. So let's find out.

🏦 How is the Esports Gambling Market Structured?

Let’s start with the esports gambling market structure, which has multiple forms. The most common types of bets include money line bets and side bets. Money line bets are placed on the winner of a specific match and are paid out according to spread betting. Side bets are placed on specific occurrences within the game, e.g., betting on the team in DotA 2 that will reach 10 kills first.

"While esports gambling markets offer the opportunity for bettors to bet real currency on the outcome of matches, there is a type of betting that is unique to esports gambling: skins betting [1]."

These in-game cosmetic items (acquirable by purchasing or randomly through gameplay) usually grant no advantage but are a display of status. They can typically be sold and bought on open marketplaces for money. In esports betting, skins can be used instead of money for placing bets. For instance, Valve's open marketplace through Steam allows for trading between players and bots.

"These “bots,”... collect digital items from players on behalf of the gambling website using Steam’s trading function and API (application programming interface), then credit the user’s account for that item which can then be used to make wagers on the gambling website.

On a side note, another way to gamble in gaming is loot boxes, where people buy in-game boxes that contain items of different rarity with varying odds. More rare items usually have lower odds of being in the boxes. Famous examples include Counter-Strike and FIFA.

📊 Where do the Data come from?

The study used data from two online gambling websites: one using real money betting, and the other using skin betting. In total, 5,171 DotA 2 (2,396) and Counter-Strike (2,775) esports matches were analyzed.

🪙 Can We Turn a Pretty Penny?

The first result to spotlight is the subjective vs. actual probability of an underdog winning. While people, on average, think underdogs win 30.6% of the time, they actually win 26.9%.

"While this difference in actual win percentage versus subjective win percentage may seem small, there are large consequences for bettors as a result of this disparity: The average actual return on bets placed on underdogs... was... a loss of 25.8 cents per dollar bet [1]."

Consequently, bettors have a bias: the tendency to overestimate underdogs, found in both games, DotA 2 and Counter-Strike. Interestingly, it appears that the greater the discrepancy between the favorite and the underdogs, the more people are willing to bet on the underdog. This, according to the data, leads to heavy losses for bettors.

"In other words, less pronounced underdogs do not seem to be overbet to the extent that heavy underdogs are [1]."

Second (and this likely confirms common knowledge) is that both platforms have or are close to negative expected returns. This means that on average, bettors are expected to lose money over the long run, especially when including the marketplace's commission. On one of the two websites, the loss was 1.9 cents per dollar bet.

Third, on the platform using skins to bet, no bias was found. This means that the actual expected return of a bet matches the subjective expectation of a player or team to win.

🇪🇺 The EU-Biases

The authors argue that the bias mainly comes from Europeans having much more exposure to the esports ecosystem. This increased exposure may lead to bettors overestimating their abilities relative to non-European teams. Therefore, betting on the underdogs (e.g., non-European players/teams) seems more likely.

Furthermore, the study highlights that informed bettors (like you and me, knowing that people tend to overbet on underdogs) could turn a profit by betting against the heavy European underdogs' bias (betting on the likely team to win). (And no, this is not a call to gamble on esports matches. Please, just don't!)

Hope you guys enjoyed this episode. See you next Sunday. Best,

Christian 😃

References

[1] Sweeney et al., 2019