Acting like a guardian angel of online poker, TylerRM (Andrey Streltsov) pointed out 15 more bot accounts over on GGPoker. 11 were inactive, closed by the site or by the account owners, we aren’t sure which. Four bot accounts were still active, at least when the report was published on the 4th of April, 2024.
Tyler’s posts are becoming a familiar sight on the 2+2 forum. Since the start of 2024, he detailed a huge ring of bots on WPN, once in January and then soon after in February.
This time, GGPoker is under the microscope. We’ll break down TylerRM’s recent report and talk about the implications.
The Key Points of TylerRM’s April 2024 Report
Over the past six months from around September to April 2024, Tyler was checking data from GGPoker cash games.
- 15 bot accounts: “socialJoy, Fusinica, Wolvox” and “finecarla” were the only active GGPoker accounts in early April. 11 others are inactive.
- 7.3bb/100 win rate: The bots achieved this after 825,427 cash game hands.
- $89,804.15 net won: Around $83,000 (or 92.7%) of the winnings came from NL100 and NL200 6-max tables.
- A 7-day downswing: Between January 9th to 16th, the bots went through a minor downswing. It’s not entirely clear why this would happen, but it’s uncharacteristic. The rest of the 6-month graph shows almost entirely upward progression.
The bots most and least profitable games: The 15 bots seem to generate the majority of the profit from NL100, NL200, and NL500, all in 6-max. Only 32 hands of 9-max NL100 are shown, for a small profit of $76 net won.
According to the report, almost all of the Heads-Up games were losses for the bot group. The only HU game the bots didn’t lose in was at NL50, but only saw $6.11 net won over 103 hands.
For the work he's done, TylerRM received a few huge gifts from PokerOK, which is GG Network's Russia-facing skin. They have given him a $10K package for the Mediterranean Poker Party in Cyprus, a high-roller event buy-in, plus two payments of 20,000 USDT.
He says that he may spend the donation on creating a school with an open and free course on finding bots in online poker. He'd also like to teach players to be more proficient in Hand2Note.
The move from PokerOK is slightly off-beat with their recent crackdown on poker stables, but it's nice to see Tyler getting rewarded for the valuable work he does.
How Did TylerRM Identify the GGPoker Bots?
Tyler’s analysis compares the stats of suspected bot accounts to human stats. A few inconsistencies might not be a smoking gun, but a slew of abnormal stats might be.
Identifier #1: WWSF and WaSD
Tyler notes that the WWSF and WaSD stats are “unusually high” and “challenging for humans to achieve.” The stats indicate a strong postflop ability, and while not impossible, he says it’s “rare to find.”
- WWSF stats for human regulars: ~48%
- WWSF stats for the suspected 15 bots: 54%
- WaSD stats for human regulars: ~52
- WaSD stats for the suspected 15 bots: 58%
Identifier #2: Lower than normal fold rate to river bets.
The report shows that the bots have a 58% WaSD (Won at Show Down) rate, suggesting a strong OOP checking range. When their opponent checks back on the river, their range is often going to be strong enough to win the pot.
But when they face a bet instead of a check-back, the bots are handling it more correctly than the humans. The bots have an average fold rate of 47% OOP against bets. It’s a much lower rate than OOP human players, who fold approximately 62% in the same spot.
Identifier #3: C-bet range construction.
The bots’ in-position flop c-bet range implies a weaker range with large bet sizings. This lends itself to higher fold rates versus check raises after 45-90% c-bet sizes.
Human regs would fold to these flop check-raises around 38%, but the suspected bots folded at much higher rates. Sometimes, they’d be folding up to 66% of the time in the described situation, or as low as 48%. Fusinica, an active account in early April, was the exception to this rule with a stat of 30%.
Identifier #4: SRP c-bet sizing.
In Single-Raised Pots (SRPs) the bots didn’t have c-bet sizes below 40% when at least one opponent was in position against them – but Tyler says humans would choose this sizing 21% of the time. This sizing was completely absent from the suspected bot accounts. However, against two OOP (Out-Of-Position) players, the bots started to include some sizings between 25% and 40%.
Identifier #5: Losses between January 9th and 16th.
The bots went through a hiccup in mid-January and had significant losses. Tyler also notes that they “played many random hands,” which doesn’t match the kind of experienced online poker player who would be showing such a profitable graph.
Key Takeaway of TylerRM’s Report
TylerRM’s report may seem like a handful of assumptions to casual poker players or those unfamiliar with statistics like WWSF, WaSD, and all of the percentages.
One identifier on its own is not a confirmation. Perhaps two, or even three, wouldn’t be a high standard of proof either. However, the smoking gun in this case are the trends we see across 15 separate accounts.
The unusually high rates of winning after the flop and winning at showdown. The unusually low fold rates out-of-position to river bets. The complete absence of a bet sizing that a fifth of human regs have. The losses the accounts had in mid-January.
Keeping these trends in mind, what is more likely? Are these 15 accounts just human players with abnormal strategies, with the same trends, at the same time? Or, is it more likely that the 15 accounts are using similar configurations with automated parameters? The answer seems quite clear.
Tyler exposed a $10,000,000 void in the WPN player economy in January 2024, but these aren’t his first rodeos. While he has a history of poker training, Tyler’s been one of the main sources of information on combating bots and bot farms recently. While making an effort to support fair online poker and he’s motivated others to join the crusade (like Moosegills).
When we interviewed TylerRM a few years ago, he told us how to find poker bots at tables. The next time he posts a report or a guide, you can be sure we’ll cover it.
With all of this talk about bots, it isn’t clear where players should go for poker. We’ve identified some of the safest ways to play poker and there are definitely options. If you want less risk or financial exposure, this 5-minute read will help you.
