0 0 157 Views

Computer scientist and poker player Ben Kushigian talks to Carlos and Andrew about the limitations of poker solvers, what we don’t know we don’t know about them, how they could be improved, the potential for open source solutions, and more. Plus, he’s got an interesting spot from a satellite bubble for the strategy segment.

Open Source Solver

The github repo is https://github.com/b-inary/postflop_solver. This is the backing solver (so no user interface). There are two user interface versions:
+ https://github.com/b-inary/wasm-postflop

+ https://github.com/b-inary/desktop-postflop

And here is a link to comparisons between postflop solver and different solver techs: https://github.com/b-inary/wasm-postflop#comparison

This is the paper that it’s based on: https://arxiv.org/abs/1809.04040

Strategy

5 players remain, 4 get paid $55+5 tickets, 1 gets $8 (so we’re on the ticket bubble). This spot seemed like a crucial moment where I had the opportunity to eat some variance but with high payoff.

Hero (BB): 5.61 BB
UTG (UTG): 12.86 BB
CO (CO): 2.22 BB
BTN (BTN): 4.5 BB
SB (SB): 12.8 BB

5 players post ante of 0.1 BB, SB posts SB 0.5 BB, Hero posts BB 1 BB

Dealt to Hero: Kc3h
Dealt to CO: ???

fold, CO raises to 2.12 BB and is all-in, fold, fold, Hero ???