Poker-AI.org

When I read this thread I asked myself, "Can the equilibrium bot beat the rake?"

That depends whether it's playing against an equilibrium bot.

There are many strategies that will tie against an equilibrium strategy and some are very exploitable.

Surely any strategy that ties against an equilibrium strategy is an equilibrium strategy? Any variation from equilibrium will become a losing bot.

Consider rock, paper, scissors. You play (1/3, 1/3, 1/3). I play (1, 0, 0). We tie.

Ooh, good example!

You play anything and we tie! However, RPS is a example of a game where the equilibrium strategy can't possibly be a winning strategy. I'm hoping Poker isn't the same.

Edit: Maybe my first post should say - "It depends on the opponent"

I just got that quote from Schnizlein's paper on State Translation.

I had in mind the equilibrium boat against the average player. Can a bot beat the rake without selecting an opponents.

I think, it should be able to beat the rake. Of course there are situations, where you will tie, but when I look at SNG push or fold game for instance, a GTO strategy still make a profit, especially against players who are too tight. Of course this doesnt guarantee, that it will be the same in other games, but I find it likelly.

The problem is that rake is usually huge.
In SNGs rake on mid-stake is roughly 10% which means you already have to have a 10% edge against other players to make money. I'm not sure this is achievable anymore with just an equilibrium strategy. The average player is pretty good these days unfortunately.

_________________
Cheers.