Crypto Gambling: Players Who Stay Profitable Over Time

Pepper Partners reviewed a 2022 study in which British researchers compared the behavior of crypto casino players with a traditional model developed in 2012 for live sports betting, meaning bets placed while a match is in progress. The study analyzed around 14,462 players and 2.3 million transactions across three crypto gambling applications between August 2018 and July 2021.

The original model divided players into four types. When applied to crypto data, it produced five groups, and in one of them, the typical player remained profitable throughout the entire observation period.

How Players Were Segmented

🔵 Each player was described using four indicators: how many days they played, how many bets they placed per active day, how much their bet sizes varied, and whether their total betting volume increased or decreased over time
🔵 Players were grouped using the k-means algorithm, which clusters users with similar behavioral patterns
🔵 The data came from three crypto applications – Dice2Win, Etheroll, and FCK

What the Data Showed

🔴 For crypto gambling data, the optimal segmentation turned out to be five groups rather than four
🔴 A behavioral type appeared that was not present in the traditional model
🔴 This group represented around 7% of all players
🔴 The typical player in this group maintained a positive financial result throughout the entire observation period

Who Stays Profitable

🔵 These players spend more time gambling than users in the other groups
🔵 In terms of the number of bets placed, this group ranks second among all five segments
🔵 The typical player in this group remains profitable throughout their entire LTV
🔵 In other words, prolonged and active gambling in this case is not accompanied by a negative financial outcome for the player

Why Does This Happen?

🔴 The researchers do not have a definitive explanation
🔴 One possibility is that this is simply a statistical characteristic of this particular sample
🔴 Another possibility is that crypto gamblers may be more technically sophisticated or more focused on financial outcomes, which could affect how and how much they play
🔴 Both explanations remain hypotheses rather than proven conclusions

Conclusion

A model originally designed for sports betting required five groups rather than four when applied to crypto gambling data, due to a segment of players who gamble actively and over long periods while remaining profitable. The study does not explain why this behavior emerges in crypto gambling: it may be a feature of the sample or a characteristic of the audience itself.