TylerRM exposed a nearly $10,000,000 void in the Winning Poker Networks player economy. Through his report, the use of bots of major poker sites is under more scrutiny than ever.
This isn't our first time chatting with TylerRM, whose real name is Andrey Streltsov. We caught up with him in 2020, for a discussion with an eerily similar theme as the one you're about to read. The interview is in Russian, but most browsers can translate it quickly.
Our Interview with Andrey "TylerRM" Streltsov
GipsyTeam profile and Telegram link: TylerRM // t.me/GTO or GTFO
– Is it possible to understand during the game that you are playing against a bot?
– [TylerRM] Of course, but provided that you know what characteristics a particular bot farm meets. If you study well-known public cases, then it will be quite possible to notice similar patterns during the game. Well, if a new farm appears that has not been studied before, then it will be difficult to understand what is happening directly at the table.
– What indicators in statistics should you pay attention to if you notice a suspicious player at the table? And in general, how can you understand that a player is suspicious if you play in a room without a HUD?
– We have made the latest statistics of bots publicly available just so that regulars can study the features of their game and answer this question on their own.
It would not be very good for me to give a direct answer publicly since bots can read this and get rid of vulnerabilities. And so let them at least rack their brains a little about where they are pierced.
– Do I understand correctly that the bot’s statistics will now differ from GTO more than the reg’s statistics?
– As a rule, yes. Bot farms are designed to exploit the field in order to get the maximum advantage; playing GTO is not the most profitable strategy for them. However, there are also GTO-oriented tips, for example the Uranus project: I saw Jardani and Xorybur bases (officially blocked by WPN with a $35k refund), they were GTO-oriented, used many different sizings, andtried to be balanced. It’s not necessarily one at a time; you need to study many farms and manufacturers of tips to understand their features.
– Where are there more bots – in the cash games or in MTT, at what limits are they more common?
– If we take the example of a botnet on WPN, then it’s about the same. There are also spins and Omaha. They definitely play spins successfully, and, judging by the latest data, Omaha too.
Early insights into the botting issue in the PLO Cash Player Pool at $200 on @ACR_POKER
— JNandez (@JNandezPoker) January 5, 2024
I've spotted 11 suspected bot accounts spending about 2 hours on this. (There are a lot more bots)
Noteworthy findings
– 14 key stats show uncanny similarities, this is very unusual and… https://t.co/aB51Y513pA pic.twitter.com/R62wUpL8fm
At the beginning of the year, Fernando Habegger wrote that he found 11 supposed bots on ACR with a large number of identical stats and very low WTSD for regs. Accounts play PLO200 with high win rates, but only play 1-2k hands per month, which is also not typical for winning regs.
In the past, bots tended to sit at mid-stakes, skipping games that were too cheap and too expensive. Too cheap ones, like NL10 are simply not interesting to them, and in an expensive game from NL5K or regular tournaments for $630, they are afraid of losing to the regs who have studied their features. Therefore, the bots felt best at medium limits. I don’t know what will happen in the future, but I assume that they will make attempts to climb to high stakes.
– What should a player do if he discovers a bot?
– Firstly, you need to discuss the case with at least one of the players you know. I see a lot of cases of misconceptions about bots among regulars with a not very high level of competence, when they consider a bot to be a player who simply ran over them, won an all-in, or called a bluff with a weak hand. Such accusations are often simply an emotional reaction to a loss. Well, if the case raises suspicions among other players with whom you are discussing the game, feel free to write reports to the security service of the poker room.
– Are the current methods of combating players against bots effective? Is there anything that can be improved/added?
– Eddie Harari recently wrote that it would be nice if there was an independent committee to assess the fairness of the game, which would receive hands from rooms encrypted by nicknames and analyze them.
There are many problems with the integrity of the game – these are not only bot farms, but also chip dumping, and multi-accounting, and hacking of rooms by hackers, as was the case with MoneyTaker69 , and hacking by some part of one’s own team, as was the case with POTRIPPER (a superuser on Ultimate Bet), and more complex problems – RTA , hosting, softplay, exchange of information on dead cards. In an ideal world, there would even be several such committees, and they would compete with each other.
At the moment, rooms are reluctant to share problems with players and the press due to the fact that their discussion damages the image and affects the number of deposits attracted. For many years, there has been a tendency to limit the amount of data provided to players – trackers, Sharkscope, and mining have been banned. I believe that to ensure trust, as much data as possible should be publicly available, just like in any sport. If a public game is played, data about its results and progress must be stored. Without transparency, there will be less and less trust.
– Is there now a central discussion thread for rooms with suspicious nicknames, so that players have the opportunity to see everything in one place?
– There's no such thing. There are separate topics on GipsyTeam and 2+2, separate posts in bot farm discussion threads with suspicious nicknames, some useful posts on Twitter, but there is no one place where you can see everything. In an ideal world, it shouldn’t exist, because if the security service were working normally, the problem itself wouldn’t exist. If the rooms work as poorly as WPN now, or even collude with bot farms, then one can imagine that such sites will appear.
– In which room does the security service work most adequately – banning suspicious players and responding to reports?
– Many actually work well. But the case with WPN is unclear to me – I don’t understand why their team refuses to check anything and ignores the gaming farm for many years. Most well-known rooms react sensibly to feedback from players and investigate any cases of fraud.
GGPoker has responded to my reports more than once, as have 888poker and Chico. They say that PokerStars has a good security team. Judging by the fact that I have never encountered suspicious cases at the tables there, this is true. I don’t have contact with all rooms, but in most cases their response to reports of table fraud is adequate.
– There is an opinion that bots can be beneficial for rooms. At a minimum, there should always be a game. Do you believe in this?
– Yes, some of them were immediately announced by the creators as assistants for the poker room. The developers believed that they could make bots with a changing level of play that would play poorly with weak players and well with strong ones. In this way, the bot would win a little against everyone and solve the problem of weak players losing too quickly.
If the poker community were atomized, this could even happen. But now the community is quite lively: there are forums where you can openly discuss problems; there are streamers who communicate with the audience; there are many funds within which players openly discuss anything. In general, you need to understand how poker marketing works – new players come to poker when they come across the successes of real living people, see their stories, and want to surpass them. Therefore, if any of the room managers think that they can replace regulars with bots, they are, of course, mistaken.
Vlad BootieInUFace
I play with bots on the WPN network almost every day. In recent weeks there have been 2-3 nicknames at the table at limits NL1-2k.
You can understand that you are playing against a bot if you have a lot of experience and understanding of GTO. In this case, those who play too mechanically or, on the contrary, too exploitatively, will be visible. Statistically, this can only be understood if there is a large sample of hands. Then you can conduct a detailed analysis of all lines and see strategies that are atypical for the field.
Without statistics, all that remains is to watch showdowns and look for plays that do not fit into standard human logic. If there are many such hands, this is a reason to take a closer look at your opponent.
An important caveat is that in both cases you need to have a certain competence in order to assess how much what you see fits into normal human behavior.
At the moment, I have not seen bots playing close to optimal (unlike players with a hint). Those that are now widespread, and this is, in fact, one network from Russia – both in cash games and in MTT – play according to a static, although often updated, exploit-oriented strategy, so you can beat them, if you understand them well, with a high win rate.
Usually, I inform my friends about suspicious players, and if the suspicion is confirmed, we begin to disassemble the game. I write reports only on those sites where the support at least somehow reacts to them (of the large sites this is actually only PokerStars).
Anton Nko1 // Ride and come what may
I encountered bots on WPN in blitz (fast poker) and at regular tables.
Their distinctive features:
– High cold calls
– VPIP 27+
– WWSF 50+
– Limps
– High RFI with SB
– High 3-bet from BB vs SB
– High aggression against checks
– Large bets against checks, large overbets with X’s
– Teamplay exhibitions
– Cold calls 3-bet
– High 4-bet/fold
– High 5-bet/all-in from the blinds, shove with any
I reported the accounts I found, but it was useless on WPN. In August, the bots in fast poker disappeared and now blitz50 is dead, and the game doesn’t always play at blitz200 either.
The adjustments for the bots are constantly changing, as they are constantly patching up holes. The strategy has to be changed almost every week.
If there is no HUD in the room, then I’m afraid it will be extremely difficult to detect a bot or RTA; you need to play a lot and closely monitor your opponents. And with no hand history at all, pitchforks.
Looks for operators in rooms and conducts investigations on the blog, here are the latest (in Russian).
My story is more about operators, not bots. I played with them for almost six months at NL100-200 on GGPoker – both with those who took first places in the leaderboards, and with those who, with a high degree of probability, are copies of WPN bots. I didn’t save strange hands, since the first ones disguised themselves very well as simple rake bots and didn’t do anything supernatural, and the second ones surprised me with outlandish aggression, but I concentrated on my game against the fish.
It is difficult to detect bots. For example, in the first batch of banned accounts, there was initially nothing suspicious, except that from the first day of their appearance, they began to grind for first places in the leaderboard. The patterns in the second game are simply difficult to see without having statistics at the table.
If there is no HUD, then there is no way to detect bots or bot operators. Except in cases where you know in advance about some exclusive patterns in the sweepstakes.
It is possible to adapt to the game of bots and this is what the regulars at WPN did, by analyzing their databases. I didn’t try to adjust – as soon as I identified the likelihood of the operators, I immediately reported the accounts. However, the nuance is that, in a vacuum, it is difficult to suspect and blame one account. You need a pack of similar accounts, with the same stats and the same graphs; in my case, these accounts even changed the game in the same way at the same time.
