ELO ratings were invented by physics professor Arpad Elo to rate chess players, but the formula is flexible enough to handle any type of competition. Teams gain points for winning games and lose points after losing. The number of points the winner wins is the same as how much the loser loses, so it is a zero-sum system. More points are transferred after upset wins, games won by wider margins, and playoff games.
Because ELO ratings show relative strength of teams, we can generate win probabilities for games based on the ELO rating difference between the teams involved. National Statistical calculates these probabilities before each game, presents them on the website, then highlights in green for correct and red for wrong.
