baruuum / ipdgame Goto Github PK

This is a Axelrod-type computer tournament where players play the iterated prisoner's dilemma against each other.

• Spring, 2024: GOB (submitted by Gracelynn Osei-Bosompem)

New strategies can be added as .R files into the ./strategies directory.

1. The strategies should functions with the same name as the file (e.g., if the file name is RANDOM.R, the function's name should be RANDOM)
2. The general signature of the function is f(ego_past, alter_past, ...) where
   1. ego_past is the history of ego and
   2. alter_past is the history of alter with
   3. cooperation being coded as 1 and defection as 0
3. Currently, the only additional arguments passed via the ... in the roundrobin.R tournament-style are (although other arguments could be added as well)
   1. n_rounds: total moves that are played between the players
   2. current_round: which move the current one is

Other tournament structures can be added into the ./tournaments directory (e.g., playing on a grid or network).

1. the_game function defines prisoner's dilemma
2. roundrobin

Running the run.R file will run the tournament. The current rules are

1. each player plays against each other as well as a copy of themselves in a n_rounds = 50 move iterated prisoners dilemma
2. the whole tournament is run n_iter = 5 times to reduce stochastic variability
3. The payoff structure is:
   • mutual cooperation: 3 points
   • mutual defection: 1 point,
   • temptation to defect: 5 points,
   • sucker's payoff: 0

• data.table (>= 1.14.9)