This is a tiny project that aims to settle a debate about how to best opening move in 6 card golf.
This repo implements a simple simulation of 6 card golf with three different starting strategies
- flip cards in the same column
- flip cards in different columns
- flip cards in the same column unless the first one is a two then flip different columns
python main.py
Running simulation for strategy: Same column
Average score after 500000 games: 9.63
Running simulation for strategy: Different columns
Average score after 500000 games: 9.26
Running simulation for strategy: Flip if first not two
Average score after 500000 games: 9.64
Strategy Comparison:
Different columns vs same column:
Difference: 0.38 points
p-value: 0.0000
Statistically significant
Different columns vs flip if first not two:
Difference: 0.38 points
p-value: 0.0000
Statistically significant
The best strategy is: Different columns
tbh, i wasn't expecting this - and thought the 3rd strategy would be best since you're acting on more information.
warning; this simulation misses many of the complexities of real world card play. This simulation is of a single player who always plays the drawn card. Additionally the simulation plays greedily and always choose the move that has the best impact on their score at the current moment. However in real play, the greedy move may not be the best move since there are other players and the best move long term may not be the best move at the moment. anyway this helped think though the game mechanics and going forward i will be flipping in two different columns