This is a WIP practice project. It does the job, but a few things are missing.
Basically it simulates a Blackjack game against a dealer with various parameters that you can change and see how it affects the outcome.
If you want to try it out: Start with the results_graphs.py where you'll set all the parameters and execute this script.
You can control:
- Number of game cycles
- Initial bet
- Initial balance
- "Stand on" value (for player)
- 'Hit' vs 'Stand' strategy (as a pandas data frame)
A few features that are available:
- Light Martingale: doubling the bet (2 times max in order to avoid total loss of balance) after the loss
- Dealer always stands on 17
WIP:
- Doubling
- Card splitting
- Insurance
When X game cycles are played through the script produces a .CSV file with these parameters for each game: Dealer's card #1 | Dealer's card #2 | Player's card #1 | Player's card #2 | Dealer's score | Player's score | bool Win/Lose | Bet placed | Balance after the game
Finally the game gives a few stats about that game cycle and shows the scatter graph (what card combos produced more frequent wins):
500 cycles | Init balance 1000 | Bet 10 | Stand on 18
Wins (number): 213, which is: 42.6% Mean score: 18.65 End balance: 800