The Game Attribute Rankings page is your ticket to reliving the best and wildest games of a particular season. The categories available for exploration are…
Excitement: A measure of how much the win probability changed over the course of a game (using a win probability algorithm that assumes the teams are of equal strength).
Tension: A measure of how close the win probability is to 50% throughout the game (using a win probability algorithm that assumes the teams are of equal strength).
Dominance: A measure of the average expected margin of victory for the winning team throughout the game (using an expected margin algorithm that assumes the teams are of equal strength).
Minimum win probability: The minimum win probability that the winning team experienced during the game.
FanMatch: The game that had the highest pre-game Thrill ScoreTM. Note that these games may have sucked but they were the ones we would have expected to be the best match-ups based on the information we had at the time.
Upsets: Games are ranked based on the winner’s pre-game chance of winning.
Busts: The model makes predictions on margin of victory. This category ranks those games by how badly the model missed.
Max Deficit: A list of the biggest comebacks based on the biggest deficit a team faced.
Window: Take the largest lead and deficit for a team in each game. Now rank the games by the smallest difference between those two numbers. That is what you get in this list.