The other answers here are fairly uninformed. AI has been reliably beating humans in NHLE since 2019 with just deep learning [1], modern solvers can play within a small epsilon of a nash equilibrium (a perfect strategy) and are effectively unbeatable.

Referencing a solver while playing (what people call real-time assistance or RTA) is definitely a problem in online poker and is always prohibited. Solvers however play in a fairly predictable way and poker sites can detect if people are using them, though I'd imagine imperfectly. Saying that 99% of online poker is vs AIs is a hilarious overstatement.

[1] https://www.science.org/doi/10.1126/science.aay2400

Makes me think the AI should know and learn to try to play more like humans, so it doesn't get detected. AI needs to make itself look more stupid than it really is, to win in this game. And perhaps that has been part of Poker always, not "revealing your cards".

But it feels to me a bit awkward that I can't use all the tools that could help me play a better game. But of course rules are rules. In Formula-1 you can only drive a specific type of car. But the car-machine still helps you move faster. The car has a cockpit and I assume it is much computer controlled.

The paper referenced here is well beyond “just deep learning”.

AI is a problem, but unlike Chess and Go, Poker involves each player having information that is unknown to the other players, and actively misrepresenting it.

I see. That's a notable distinction. You got know when to hold them, and when to fold them :-)

I've had a very long standing debate with friends about the nature of poker. I'm neither a statistician, game theorist, nor mathematician so I'm very open to being corrected. My intuition having played recreationally is that the absolute optimal move in poker is relatively trivial to calculate compared to chess/go. That is to say, most experienced players would reach this threshold without too much training. Obviously, what makes poker fun/interesting is that you are trying to guess the strategies of other players, based on your interpretation of their behaviour, and react accordingly. If all that is true, then I would submit, that at any sufficiently high (i.e. non beginner) competitive level it probably comes down to luck, since I don't really believe that the second skill described is really learnable. So when we say "John is the world poker champion for 10 years running and is obviously far better than any mid tier player" are we not saying, essentially, that John is a human polygraph machine. Why, then, is that absolutely astounding skill not being harnessed for something else?

> are we not saying, essentially, that John is a human polygraph machine

No we are not. Poker skills (even at a high level) go beyond live tells.

But if we accept the mathematics of poker are not that complicated, then are those other skills not highly generalizable to any job requiring extraordinary insight into human behaviour?

playing good theoretical poker is very complicated.

People spend thousands of hours trying to grasp it.

In a 100BB 6-max game in a single raised pot BTN vs loose player on Bb, who called your raise: is the J83r a good flop for a small bet? Maybe a bigger bet is the proper size? This is a simple question that has often a correct answer and the question is different for each flop, for each preflop spot and potentially for different opponents.

Now once you have figured out betsizes on the flop - go figure out which actual hands you are supposed to bet. Hmm, you should play most hands with mixed strategy of check and bet, but with different probability distribution so that your distribution of good hands / bad hands on the future streets after called is right on a variety of turn cards.

You can spend all life mastering that part alone and we are talking only about the first decision on the flop in single raised pot.

Tldr; poker theory is not something easy and quick to master.

Definitely. And this isn't even bringing in what can be perceived as the "fuzzy math" of who has the range advantage, combinatorics, and consideration on how to proceed down the game tree on different streets (when the board changes texture).

There's certainly a part of me that wants to go thru this curriculum just to say that there's no way it could help someone to be better than an 'average player'.

Perhaps if the population on which the average is based is...the world?

Yes, but my comment wasn't directed at that point. It was just that we can't reliably say that being a (theoretical) top 3 live game poker player allows us to isolate the most important (differentiable) skill at that level being live tells.

Further, I disagree with your point that the mathematics of poker (implicitly top tier highest level poker) are not complicated.

(edit) to your top line point: > My intuition having played recreationally is that the absolute optimal move in poker is relatively trivial to calculate compared to chess/go.

While the branching factor of Go is between 10^250 to 10^361 (and chess is ~35), poker is 10^17 to 10^165 depending on the game variant: ""Thus, the game tree for no-limit has a much larger branching factor and is significantly larger; there are 10^165 nodes in the game tree for no-limit, while there are around 10^17 nodes for limit (Johanson 2013)" - Sam Ganzfried, Reflections on the First Man Versus Machine No-Limit Texas Hold ’em Competition

...though it's the hidden information component which ratchets up the complexity and leads to a game tree size that's many orders of magnitude higher. If you're curious about that, there's some great info here on other aspects which add complexity: stack depth, multiplayer scenarios, etc: https://poker.cs.ualberta.ca/publications/billings.phd.pdf

That said, I do agree with a relaxed version of the point you're getting at: some subset of high level poker skills can generalizable quite well to other jobs.

100% possible. I'd ~bet~ that 99% of online poker cash games are adversarially against various forms of AI.

The future of online play will be catered more toward teaching, solving, and tracking hands, rather than playing competitively for profit. I recently started a side project in Rust to do just that, mainly for love of the game and desire to learn the language.

Hoping to release later this year and perhaps productize if it gains traction.

I would sell that 99% number for a six figure sum if you'd like to define terms, an arbiter, and setup escrow.

AI has been able to crush heads up (2 player) No Limit Hold Em for a while now and 6 player for a few years. In that time frame though, a lot of other games have gained popularity, and playing a mix of games in the same sessions. AI/solvers are also normally optimized at certain stack sizes (measured in poker as how many big blinds each player has) and do not do as well in very deep stacked games. The trend of playing more games, newer games, and playing deeper will probably continue to stay ahead of AI, and people that study it. I would not play low to mid stakes NLHE online at all on any side at this point personally, just too likely to run into someone using assistance that will crush you easily.

There's been a surge in discussion of online cheating in poker. some online poker players certainly use real time assistance. There's also bots grinding online too.

The prevalence is an open question, though. I imagine it's rather high since there's a huge incentive to cheat.

There have been effectively unbeatable 1v1 limit hold’em bots online for at least 15 years - this has been a problem for a long time. Poker popularity ebbs and flows, we seem to be in a bit of a “boom” period right now.

Yes, yes, and: yes.

Yes AI will solve this. Right now because it's a hidden information game it's not as solved. Alberta Uni has done some great work here.

I'm trying to work on this myself in my spare time.