Ultimate Bet COO Discusses Player Refunds
I wanted to start blogging more frequently in order to answer any questions that the poker community has. So I am going to start blogging bi-weekly and do my best to cover topics the poker community would like to hear my thoughts on.
For this post I wanted to discuss the UltimateBet refunds as I have received a lot of questions recently regarding some concerns with the methodology we used.
How much money was refunded?
We initially refunded $22,054,351.91 and then later we refunded an additional $227,956.66. The additional refund was because of an error that was made with split pots. Later on we refunded an additional $494,932.20 to several accounts that had their refunds on hold because they were still under investigation at the time we did the refunds but were eventually cleared and processed. So the total amount that was refunded to players was $22,777,240.77 USD.
Why did we use the "Net Loss" method to determine how much to refund players?
When investigating which players were cheating we ran into many issues. We were limited by the information available to us and we had legacy data sources and applications from the previous owners that we learned we could not trust. We also were initially relying on the old UB software developers to help us with the analysis, which proved very problematic.
We eventually hired independent database analysts and a poker mathematician to help us redo the analysis because of the frustrations we had dealing with the problems I mentioned above.
We tried many different ways to calculate the refunds but with the information that was available to us, we could not isolate the hands where the cheating accounts had an unfair advantage. We also tried isolating sessions during which the suspects were cheating but we were unable to do so with a degree of confidence that we were comfortable with. The short-term variance of poker unfortunately does not allow it. The cheating accounts also displayed abnormal winning statistics for their entire lifetime. Some of the cheating accounts did have losing sessions. However, these few losing sessions appeared to be an attempt to avoid detection. All of the cheating accounts won abnormally high amounts for their entire existence. Because of these facts, the fairest solution to the players was to assume that the suspects were always cheating and to refund all the money lost by the players. This really was the one fair way to determine refunds, and that is what we did. In order to determine how much a player lost we used the "Net Loss" calculation.
"Net Loss" Explained
"Net Loss" is the sum of the player's pot contributions where one of the cheating accounts won minus any money the player won from the cheating accounts.
"Net Loss" Example:
Cheating Accounts: BeachBumAA, Broke_In_LA, NioNio
First Pot
Player A contributes $500
Player B contributes $50
BeachBumAA wins first pot
Second Pot
Broke_In_LA contributes $300
Player B contributes $200
Player A wins the second pot
Third Pot
Player B contributes $1000
NioNio wins the third pot
Refund Calculation using Net Loss
Player A: $500 (lost to BeachBumAA) - $300 (won from Broke_In_LA) + $0 (no win/loss to NioNio) = $200 Net Loss
Player B: $50 (lost to BeachBumAA) + $0 (no win/loss to Broke_In_LA) + $1000 (lost to NioNio) = $1050 Net Loss
What is the significance of Uri Kozai's refund analysis that was done for Excapsa?
The previous owners of the UltimateBet.com business were a group of companies. For the purpose of this post I will call them the "Excapsa Group" of companies. They were the owners of the business whom we inherited the cheating scandal from. We eventually settled with them for $15million in order to get players their money back. Excapsa is currently in liquidation and required court approval to settle with us. It is my understanding that the Excapsa liquidator employed Uri Kozai to do some independent review of our refund analysis in order to confirm that the amounts of money we claimed were stolen from players was accurate. It is understandable that they needed to confirm this independently because we could have inflated the refund amounts in order to try to obtain a larger settlement. His analysis was nowhere near as thorough as ours. He simply reviewed the work we did and was satisfied that our analysis was accurate. It is untrue that he developed the refund methodology that we used.
I already received several other questions that the community would like me to comment on, so I will blog about them in future posts and I will keep collecting requests and try to address them as quickly as possible.
