Poker buddy

This repository hold the different parts of a poker assistant projects.

Installation

This project was only developed using Python and numpy.

git clone https://github.com/frgfm/poker-buddy.git
cd poker-buddy
pip install requirements.txt

Usage

If you prefer running the simulation directly without entering a shell:

python src/main.py

If you favor using objects in Python, you can use the classes and functions from the game.py file

from game import Player, Game, get_hand_value, get_winners
players = [Player() for k in range(4)]
g = Game(players)
g.deal()
g.hit()
g.hit()
g.hit()
hands = []
for k in range(4):
    print('Player %s: %s' % (k, g.get_player_cards(k)))
    hands.append(get_hand_value(g.community_cards + g.players[k].cards))
    print(hands[-1])
print('Winner:', get_winners(hands))

Either way, it should simulate a game with an output similar to:

Dealing cards to players...
Flop: ['9 Diamonds', '6 Diamonds', 'Q Diamonds']
Turn: ['9 Diamonds', '6 Diamonds', 'Q Diamonds', '3 Diamonds']
River: ['9 Diamonds', '6 Diamonds', 'Q Diamonds', '3 Diamonds', '2 Clubs']
Player 0: ['4 Spades', '10 Diamonds']
{'royal flush': None, 'straight flush': None, 'fours': [], 'full house': None, 'flush': 'Diamonds', 'straight': [], 'threes': [], 'double pairs': [], 'pairs': [], 'high': 10}
Player 1: ['5 Spades', 'A Spades']
{'royal flush': None, 'straight flush': None, 'fours': [], 'full house': None, 'flush': [], 'straight': [], 'threes': [], 'double pairs': [], 'pairs': [], 'high': 12}
Player 2: ['K Diamonds', '9 Spades']
{'royal flush': None, 'straight flush': None, 'fours': [], 'full house': None, 'flush': 'Diamonds', 'straight': [], 'threes': [], 'double pairs': [], 'pairs': [7], 'high': 11}
Player 3: ['J Spades', '10 Spades']
{'royal flush': None, 'straight flush': None, 'fours': [], 'full house': None, 'flush': [], 'straight': [], 'threes': [], 'double pairs': [], 'pairs': [], 'high': 10}
Winner: [2]

TODO

• Model a Texas Holdem Poker game
• Live computation of odds of winning depending on each player's point of view
• Investigate card recognition from pictures
• Investigate Reinforcement Learning to capture player decision-making profile