>"What that means: if you pick a player at random, you’ll get the winner right about 10% of the time (this varies with how many people are left, but 10% is an average). My model’s #1 pick is right just over 20% of the time. That 2x improvement holds for most of the game, basically after two episodes have elapsed (once the features have enough data)."

Ok... but no one picks a winner at random. How does this do against anyone who has watched the show and picks someone based on their knowledge of the game? Seems like a really complicated way of just barely doing better than a dartboard.