Jean_Dessaut, on Aug 07 2012 - 06:35, said:
Nah, I think they just took raw numerical efficiency as their point to call "average". In fact their quantiles are just chosen at regular intervals of size 300 efficiency rating. By their categorisation, 90% of the player base are "average" or worse. I've even seen a post in a thread on the EU server where a guy with 1700 rating said he considers everyone below 1200 to be noobs. He's entitled to think that, but calling 9 out of every 10 people he plays with in pubs noobs is not a good way to go about winning.
For that matter, someone at the median (somewhere around 830 or so) should on average contribute about half as much (in terms of scouting, capping, kills, dmg) as someone in the top 2%. This is a pretty good argument for not completely dismissing the lower quantiles.
You hit the nail on the head. According to the Efficiency Rating definitions, 10% of players are Good or better, 90% of players are Average or worse . . . .
As to the usefulness of the measure . . . I know I am an average player so my score is not as important to me as the information about the other players (total games played, games played and victories for the tank they are using in the current battle, etc.).
A couple suggestions to improve the rating formula:
1) Get rid of the Average Destroyed. The formula already values the Average Damage per game, and effectively values the same dimension twice.
2) Include a value for Average Experience adjusted by Average Tier. The WoT designers already have an algorithm that evaluates performance across multiple dimensions. That value should be included as better players score more experience than poor players, even in games that end in Defeat.