jacobmake

I’ve finally come up with an algorithm I think works well. Without getting too much into the programmatic details, here is a brief explanation of how this will work. Each game falls into one of three categories;

• Normal – Players are pretty evenly matched. Default values are given to the winner and subtracted from the loser. A modifier of how close the game was is also factored in.
• Upset – When the player who won the game is ranked much lower than the favorite the product of the game is an upset. The underdog (winner) absorbs a percentage of the favorite’s (loser’s) points. The loser’s rank also decreases by an amount relates to the difference in rank. A modifier of how close the game was is also factored in.
• Predicted – This is the opposite outcome of ‘upset’. Very little points (if any) change hands here because the winner was heavily favored. This prevents good players from “beating up” the weaker players. A modifier of how close the game was is also factored in.

This should keep things pretty balanced and competitive. I should be able to post some screenshots soon!