- “WHR” (Whole History Rating) is now available within the current ranking pages.
- WHR is yet another mathematical system that is based on Elo, but is drastically different. While the standard, Elo, and Glicko rating systems are executed once to produce ratings output, where ratings move incrementally after each bout, the WHR system depends on numerous iterations to build the current ratings. In doing this, the system is able to use knowledge derived from previous iterations to build a supposedly more accurate picture as it can also use what fighters and their chain of opponents did before AND after their respective bouts. You can read more about that system here -> https://www.remi-coulom.fr/WHR/. There is also more info in the FAQ page. Note: The actual WHR ratings have been increased by 1750 points as a display-level modification, so the rating point values match up better with Elo and Glicko for comparison purposes. WHR ratings as posted are likely to change in the coming days as we dig in to try and improve prediction rate.
- Glicko-1 ratings reduced by 500 points. This is a display-level modification so the ratings match better with Elo (which start at 1000 instead of 1500) and WHR systems.
- For the new “Underrated” and “Overrated” designations, we’re moving over to an Olympic-type model by dropping the highest and lowest ranks of the 4 mathematical ranking systems, before making the calculation.