This is a simple evaluator for Prisoner's Dilemma strategies. It takes in python code that defines a strategy and pits it against other strategies in a round-robin tournament. The strategies are evaluated based on the total score they achieve in the tournament.
Demo: https://prisoners-dilemma-evaluator.vercel.app/
To evaluate a strategy, you need to define a python script that outputs the first move of the strategy and takes in the subsequent moves of the opponent. The outputs of the code should be cooperate or defect. After set number of rounds, the program will be terminated by sending a stop to its stdin.
Here is an example of a simple strategy:
import random
print("cooperate", end="\n")
inp = input()
while inp != "stop":
print(random.choice(["cooperate", "defect"]), end="\n")
inp = input()Example stdin:
cooperate
defect
cooperate
defect
stop
As the input has 5 lines, the strategy will play 5 rounds of the game, one blind move and 4 moves against the opponent. There must be a total of 5 lines in the output.
To submit a strategy, send a POST request to the server with the following parameters:
POST /strategies
{
"email": "[email protected]",
"code": "python_strategy_code"
}To evaluate a strategy, send a POST request to the server with the following parameters:
POST /trigger-evaluation
{
"strategyId": "strategy_id"
}The server will respond with the evaluation results.
Feel free to contribute to this project by creating a pull request.
This project is licensed under the MIT License - see the LICENSE file for details.