For today's blog I'll be creating a game clustering that uses as input only the information that we might reasonably know pre-game - for example, the pre-game team MARS Ratings, Bookmaker prices (or some metric derived from them), and information about the game venue.