Bot vs Human Winrates
How much better is a bot than a human?
We know a bot cannot lose, and so the best result a human player could achieve, in purely theoretical terms, would be to break even (in reality, unachievable).
But how would a bot's winrate change against a spectrum of human strategies?
Considering HU NLHE & HU PLO only, and assuming 100k+ hand samples:
- At what WR would it beat the top HS professionals?
- At what WR would it beat the midstakes professionals?
- At what WR would it beat the average rec?
And why?
4 Replies
I don´t have the capacity to answer your question, but if you have a RIO Elite subscription, you can look for a video by Kevin Rabichow where he compares two HU bots, and the winrate gain by the best bot was mindblowing, double digit. I believe the 2 bots were RuseAI and Slumbot.
Given any human will have to resort to simplifications and heuristics, and will have imperfections and innacuracies all over the place, I´m pretty sure even the best HU HS reg in the world would suffer a double digit lossrate massacre.
Guess I was probably the last one here to learn that GTOWizard bought Ruse, and that’s the reason why they came up with the AI feature? 😃
