An implementation of the simulation described in this video and the strategies mentioned in More Effective Choice in the Prisoner's Dilemma. Here's an overview of the experiment.
The video generally notes that strategies that were "nice"(did not defect without cause), "forgiving"(went back to cooperating), and "not a pushover"(retaliated) were the most successful.
My hope is this code can allow more people to try this out for their own and write their own strategies.