We use machine learning to uncover regularities in the initial play of matrix games. We first train a prediction algorithm on data from past experiments. Examining the games where our algorithm predicts correctly, but existing economic models don't, leads us to add a parameter to the best...

Algorithms are increasingly used to guide decisions in high-stakes contexts, such as who should receive bail, be approved for a loan, or be hired for a job. Motivated by a growing body of empirical evidence, regulators are concerned about the possibility that the error rates of these algorithms...