These bots are regularly going down 20%+ on high cards duels
Show HN: Play poker with LLMs, or watch them play against each other
61–70 of 101 posts
Re: Show HN: Play poker with LLMs, or watch them play against each other
#62Placing full GPT 5.2 versus fast/flash models of main competitors is unfair, would love to see more balanced table.
Re: Show HN: Play poker with LLMs, or watch them play against each other
#63This is fun!
Given online is now bot-riddled, I half-finished something similar a while back, where the game was adopting and 'coaching' (a <500 character prompt was allowed every time the dealer chip passed, outside of play) an LLM player, as a kind of gambling-on-how-good-at-prompting-you-are game. Feature request! The rake could pay for the tokens, at least.
Re: Show HN: Play poker with LLMs, or watch them play against each other
#64Earlier quoted context omitted.
To expand on this - an LLM will try to play (and reason) like a person would, while a solver simply crunches the possibility space for the mathematically optimal move. It’s similar to how an LLM can sometimes play chess on a reasonably high (but not world-class) level, while Stockfish (the chess solver) can easily crush even the best human player in the world.
How does a poker solver select bet size? Doesn't this depend on posteriors on the opponent's 'policy' + hand estimation?
Nash equilibrium. Optimal strategy for online poker has been known for like literally 20 years right now
Re: Show HN: Play poker with LLMs, or watch them play against each other
#65Earlier quoted context omitted.
To expand on this - an LLM will try to play (and reason) like a person would, while a solver simply crunches the possibility space for the mathematically optimal move. It’s similar to how an LLM can sometimes play chess on a reasonably high (but not world-class) level, while Stockfish (the chess solver) can easily crush even the best human player in the world.
How would an LLM play like a human would? I kind of doubt that there is enough recounting of poker hands or transcription of filmed poker games in the training data to imbue a human-like decision pattern.
Also, if you set the bar for human players low enough, pretty much any set of actions is human-like. :p
Re: Show HN: Play poker with LLMs, or watch them play against each other
#66So strange that people are into this, but were not into the much stronger non-LLM poker agents.
Re: Show HN: Play poker with LLMs, or watch them play against each other
#67this could make for an interesting new benchmark
Re: Show HN: Play poker with LLMs, or watch them play against each other
#68I used to play professionally, and I still play in the casinos. These LLMs are playing better than most human players I encounter (low limits). They're kinda bad, but not as criminally bad as the humans.
lol what? I just watched Grok fold pocket jacks preflop, no raise/limps ahead.
That's, in fact, a much smaller mistake than I see humans make every hour.
Re: Show HN: Play poker with LLMs, or watch them play against each other
#69Earlier quoted context omitted.
The solvers don't typically work in real time, I don't think. They take a while to crunch a hand.
"Solvers" normally means algorithms which aim to produce some mathematically optimal (given certain assumptions) behaviour. There are other poker playing programs [0] - what we called AI before large language models were a thing - which achieve superhuman performance in real time in this format. They would crush the LLMs here. I don't know what's publicly available though. [0] e.g. https://en.wikipedia.org/wiki/Pluri…
Solvers, in a poker context, are a category of programs. They run a simulation after you enter the known information.
Like piosolver, as an example.
The best poker-playing AI is not beatable by anyone, so yes, it would crush the LLMs.
Re: Show HN: Play poker with LLMs, or watch them play against each other
#70This is very cool, one piece of feedback: watching the table as the AI plays while seeing the reasoning is difficult as they're on other sides of the screen. It could be nice to have the reasoning show up next to the players as they make their moves.
Yep, exactly. It's very difficult currently.