Update, October 18, 2026: QQPK subsequently announced 724 permanent bans for unauthorized automated play. Our follow-up on QQPK’s account bans, published October 13, explains what that announcement did and did not establish about the original 2,587 restricted accounts and possible refunds. The report below records QQPK’s initial response and its stated October 17 reporting deadline.
QQPK says it has temporarily stopped withdrawals from 2,587 accounts while it investigates allegations of automated play and unusual all-in results.
The poker platform says it will publish its findings by October 17, giving players and the report’s authors a week to see whether the review produces answers they can check.
The account restrictions are precautionary, QQPK said, and do not mean that every account on the list has been found to be operating a bot. The company says it will assess each account individually, reassess restrictions as those reviews are completed and close accounts where bot use is confirmed.
QQPK responds to a large tournament data analysis
The review follows a community investigation into QQPK tournament hands played from July 2 to October 2. The report’s authors say they examined approximately 4.18 million hand histories covering about 26 million hands across 27,936 tournaments, then identified 2,587 accounts they suspect were automated.
The authors report 193,757 qualifying all-ins between those accounts and players outside the group. They calculate that the flagged accounts finished around 138,705 big blinds above equity expectation in those hands, a deviation of 14.6 standard deviations under their model. A separate set of 33,958 exact flop and turn calculations was reported to be about 19,006 big blinds above expectation.
Those figures are the investigators’ findings, not independently confirmed results. They say that two parsing tools and two EV engines produced similar outputs, but the full hand-history dataset is not publicly available because it contains information about real players. The authors say they are willing to share it with relevant parties for independent review, and they have described their analysis as focused on multi-table tournaments rather than cash games.
The report does not claim to prove that QQPK manipulated its RNG or that the operator was involved in any alleged bot activity. Its authors say the account patterns and statistical results justify a detailed investigation. Their full account of the analysis is available in the community investigation on Reddit.
What QQPK says it will check
QQPK says it will rerun the all-in EV analysis using its complete server-side records for the period. It argues that its records include hands third-party tracking can miss, such as a player’s final hand before elimination, and says it will compare the results against the same 2,587 accounts named in the report.
The company also says its security team is reviewing those accounts for unauthorized automated play and other integrity violations. Confirmed bot accounts will be closed, QQPK says, and the company intends to publish the actions taken along with an explanation of how it will identify affected players and determine any remediation.
“Players deserve specific, verifiable answers,” QQPK wrote in its response. It described the withdrawal holds as “precautionary measures” and said it would “publish our findings by October 17,” including methodology and enough detail for the report’s authors and the wider community to check the work.
QQPK says it has invited the report’s authors to arrange a hand-by-hand comparison, asking them to provide hand IDs, the account list and their calculation rules. It also says the investigation will continue while that comparison is arranged, and that it will identify any unresolved work and set another reporting date if the October 17 review is incomplete.
What the results will need to explain
The central test is whether QQPK can show how its server-side replay addresses the report’s main claims, and whether the authors or another independent reviewer can reproduce the key calculations. A useful report will need to explain the sample, the EV method, the bot-account review and the way any affected players will be identified.
QQPK also points to a GLI-certified random number generator and an Anjouan licence. The company’s public site identifies QQPOKER Technology Limited as the operator and lists licence number ALSI-202603009-FI1, while its post links to a GLI certificate. Certification of an RNG is relevant to card dealing, but it does not by itself determine whether accounts used automated play or settle the specific statistical claims in this report.
There is a further detail QQPK could clarify alongside its findings. A review of the platform’s public documents says the GLI certificate names BTS Ltda. doing business as QQPK and lists qqpk.poker, while QQPK’s site names QQPOKER Technology Limited and qqpk.com. Different legal entities can operate related brands and domains, so that distinction alone establishes no wrongdoing; a clear explanation of how the certificate applies to the platform under review would help readers assess the claim.
The timing matters because online poker has recently faced separate concerns over account integrity and player security. A high-stakes poker superuser investigation and subsequent reporting that around 30 players may have been affected have kept questions about evidence, operator reviews and player protection in view. QQPK’s case involves different allegations, but the promised public report now gives the company a clear opportunity to answer this one directly.
For players whose withdrawals are held, the most immediate questions are practical: how long the review will take, how QQPK will communicate individual decisions and what happens to balances if an account is cleared. The company says restrictions will be reassessed as reviews are completed, but its October 17 update will need to explain how that process works.
Readers looking at the wider issue can also see what happens when a poker site investigates suspected cheating. In QQPK’s case, the difference between a promise and a credible answer will be the detail in the results, the clarity of any account actions and whether the analysis can be independently checked.
