Show HN: Searching for optimal strategies in a puzzle game with chance (Drop7)

A puzzle game called Drop7 has been used to test the limits of artificial intelligence and human strategy, with a researcher seeking to find an autonomous strategy that averages over one million points per game. Despite automated strategies averaging 300,000-400,000 points, an experienced human player can achieve over a million points. The researcher invites others to contribute to the project by competing on the global leaderboard and submitting their human strategies. The project also serves as a testbed for using agents as research collaborators and evidence collectors. The researcher seeks feedback from experts in game theory, stochastic planning, and learned value functions.