General Discussion

General DiscussionFactors that determine individual rating

Factors that determine individual rating in General Discussion
Standin.n00b

    Hi,

    I just realized that rating system for dotabuff has been disabled (was away for sometimes). Since KDA ratio are not included in the rating determination (as mentioned by the dotabuff), I kinda wonder what factors are used to determine individual skill level.

    I think knowing the factors can help most players to get better and create a better gaming experience for us.

    So care to share your view?

    Epokhe

      Dotabuff rating used just wins, losses, and skill rating determined and changed with them. How much rating you get from a match was determined by looking at your teammates skill level, then enemies skill level, and then did you lose the match or win it. For example, if you play with the most skilled players(rating-wise) and your enemies are like, Merlini, TC, universe, etc., you don't gain so much rating if you win, because you have nice allies. But if you are with players with bad rating, and you play against the same team, if you win, you will get tons of rating. It works same way you just lose less rating when you lose if you play against a good team. Roughly, this was how dbr working

      Vaikiss`742.

        wins win%

        Stock(TME)

          I am sure kda is included into real mmr, but depending on the hero im sure, or well pure dmg done to certain hero as in way of contribution. Im sure you get much more points if you are 20 0 5, when everyone in team is sucking and you win, then when you are 0 10 3 and your team wins game for you.

          They made way they calculate "skill" "contribution" whatever , with each hero and that + depending of win or lose influences it a lot.

          Also why should you not be rewarded if you are one who actually won game even if 4 guys in your team were higher rated, but they sucked that game.
          Same as why should you be punished if they were not in mood to play.

          Dotabuff didnt include kda since they were reverse engineering the rating. Dotabuff just wanted to tell you somewhat your mmr, based on games you were playing and their outcome. So you should not think about it. It was not accurate ofc, but really close.

          cupcakebruh

            ^no, I am quite sure they do not, as they can be completely irrelevant to your actual presence in the game, and many heroes naturally deal significantly more damage than others. I don't mean that by way of defensive/offensive skills such as heroes like Omniknight or Dazzle who may deal more damage by allowing allies to deal more, but heroes deal damage in large AoE or are able to target a lot of enemies at once like Zeus, Spectre, Leshrac are going to be dealing a lot more damage than single target heroes.

            Anyway I am quite sure that what Epokhe said is what is currently being used. You gain points for winning or losing, and the amount you gain or lose depends on the rating of other people in the game, which is also effected by stacks.

            The issue is that there are so many differences in each game that while it may be appropriate, and better in some cases to allow for KDA to be involved in your rating, there are many games in which KDA means absolutely nothing. You are playing Alchemist? You drop acid spray and if your team wins that fight, you get 5 assists. Same with zeus, spectre, furion, etc....you sit in your base after level 6 and just KS with zeus? Your KDA is good, but you didn't do shit.

            Этот комментарий был изменён