Quote:
Originally Posted by Mugsy
How did you make those changes?
|
I have a Google spreadsheet that takes data from the exported csv of the fighters, and a formula takes each fighter's performance points and modifies it based on any losses from their last 5 bouts. The more recent the loss, the higher the penalty. Then I have another formula to determine each organization's "bias" for a fighter by using the mod function, the fighter's ID number and a multiplier to magnify those biases between fighters.
The major orgs mostly agree on the top 10 fighters, with some minor variations, but the minor orgs were setup to have wildly different top 10s in comparison.
Previously, I was "simulating" the organization's "bias" by using random numbers, but it was causing wild swings in ranking changes. Now that I've changed it, orgs biases are consistent for each fighter, so ranking changes should make more sense. If someone climbs the rankings, it is because they either won their last fight, or were below someone who lost their ranking due to losing/retiring.
The spreadsheet takes these results and sorts the fighters for each ranking in two sheets (regional and world) that displays them in the format I've been posting