Now let’s talk about bots and cheating software in poker. No one will argue that they are harmful to poker and the poker ecosystem, but bots pose a much smaller risk than cheating software. At first glance, this might seem counterintuitive. After all, cheating software requires a human to press the buttons, while a bot can be run on multiple accounts simultaneously without requiring significant computing power. This is partly true, but poker sites find it much easier to track bots due to their similar behavior and statistics.
It’s not just about the naturalness of mouse movements, but also the timings that accompany specific game actions of these players.
Economics plays an even bigger role here. A team of developers doesn’t spend so much time creating a bot just to run it on a single account in a cash game at medium limits. Usually, cheaters use manipulation to create entire networks of bots, which are sooner or later exposed, and there are several reasons for this!
(Bot networks) consist of many players with almost identical stats that don’t change over time.
Each bot needs a fake identity, documents, bank cards, address, utility payments, etc. Most often, bots are detected precisely because of discrepancies in the “player’s” payment information.
Bots in poker in Regulated and Unregulated Poker Rooms:
If you compare regulated and unregulated poker rooms, which do you think catches more bots and why? That’s right, regulated rooms usually uncover far more bot networks. But most people think that this is due to the regulatory bodies that oversee them, forcing them to have a security service and investigate suspicious cases. This isn’t necessarily the case. Often, the reason websites operate without regulation is that it’s a deliberate business decision. Unregulated sites can accept deposits without requiring players to verify their identity. You can create multiple accounts there immediately and use Bitcoin, other cryptocurrencies, or a dozen other random deposit methods, of which there are now many.
At the same time, to play in a regulated poker room, you must provide a copy of your ID, a utility bill, and have a bank account or bank card in your real name. Consequently, bot operators face much greater difficulties in regulated poker rooms: they not only have to invent identities for all their accounts, but also prove the legal origin of their bankrolls. That’s how things are.
Fraudulent Software
The operators of poker bots sometimes face the same problems. If you create fraudulent software of this type and want to sell it to a large number of people, they will also need real identities, and they will all play very similarly to each other.
The reason why a bot is harder to detect is not so much due to natural mouse movements or the ability to chat. The point is that real people can deviate from the script. For example, the bot advises a raise on the turn, but the person thinks: “I know this player, and he almost always has a good hand in this spot, so I won’t raise the turn.” And if he does this often enough, his stats deviate so much from the optimal ones that he becomes not so easy to detect.
This leads us to one of the main ways to detect bots and bots using assistance software β analyzing game statistics. In the past, such analysis was carried out manually or semi-automatically: the poker room’s security service found players with the most similar stats and ran their hands through a solver to see how often they played optimally.
Is it true that artificial intelligence is killing poker?
Now, this semi-automatic verification is much more automated: the software itself flags players with certain criteria, after which specialists begin a detailed study of these accounts. The criteria can be different: for example, some accounts always play at the same table; in this case, collusion or exchange of information about pocket cards may be taking place, they may be calling each other separately via messengers and discussing cards to eliminate players from the table and appropriate the winnings, and then divide them. Whether they are bots or real people. Or, in specific spots, all the flagged accounts demonstrate statistics that the average player would find very difficult to achieve without using fraudulent software. But now, to detect artificial intelligence, poker rooms themselves are using their own artificial intelligence, and this is a long-term trend that will only grow and develop.
There is another non-obvious fact that helps catch cheaters: most poker players do not win in the long run. Depending on the popularity of the poker room and the proportion of amateurs at the tables, 75-90% of accounts are losing money. With a large sample size, there is no need to check losing accounts: even if they are cheating, they are not doing it well, and they don’t significantly harm the poker ecosystem; the room also earns rake from them, but for the players, of course, this is more painful. This somewhat narrows the search: you only look at winning accounts, compare them with others, etc.
This is a very brief overview of how bots and cheating programs are detected. There are other methods: for example, the poker clients of some rooms take screenshots of your screen during sessions to see if you are using cheating programs. Rooms can also check the processes running on your computer.
Comments are closed