General Discussion

General DiscussionMy opinion how to calculate SINGLE MM PLAYER SKILL , I wanna know wha...

My opinion how to calculate SINGLE MM PLAYER SKILL , I wanna know what you guys think about this? in General Discussion
BogiDotA

    This is how i calculate skill mm lvl for players , for example my friends or me in MM.

    Anyone of us can watch live games,and choose a hero at live games. For example if you play slark there will be like 200-400 slarks pages at same time..it`s matter how much ppl already online at that time with playing slark. As you can read when you choose hero at live games it`s written ( LIVE GAMES ARE SORTED BY AVERAGE SKILL,HIGHEST TO LOWEST). YES ! Its totaly true. Games with most op visitors and page 1,2,3 are always pro players and very usually you can see for example some players who played int3 etc. They are at page 1 cos they are best with skill. Same for you,if you playing so good mm it`s doesnt matter you are 60% win rate 1000 games or 55% or 70% if you at first pages your MM is high and you can can say that you are rly good. Someone who is 50% win rate he cant be top pages never. For example when i played slark 4-5 hours ago i was 7/280 pages with slarks. That mean there was like 6 page(30 games) 30 players which for that time are better skilled than me in (MM CALCULATE) Same you can do for heroes you play,work same for any hero. You can see how you improve for example if you 15/300 page you : that and its 20 keep that in mind next time you play hero and if its like 12/350 you can see for sure that your MM grow and you are than in higher MM. I can pretty easy check my live games and see where i m cos i m always first 15 page with any hero depens on how much players online etc. so i can say my mm is pretty high but not even close too most high mm for pros. I know is maybe hard to check it if you are like 30-40+ page but..ask some of your freinds maybe when you are in game to check it for you. Also i must say (THIS ONLY WORKS IF YOU PLAY ALONE MATCHMAKING) if you play with freinds is totaly different results cos you and your friends are not at same mm skill lvl,maybe with some friends not big difference but is always a difference. For example If i play with friend who is better than me,I will be like 2-3 page with heroes not liek 7-15 where i m usually when playing alone.Same if you play with someone who is worse you get 20-30-40 page etc.. or more.. I hope someone read this cos i think is some of rare ways where you can see your real mm skill. Cheers guys and gl in mm

    Kanye Best

      Im sending this to valve right now, gr8 suggestion m8

      HeLL_RAISeR

        this is not a way to calculate a player MMR. this is the way to show you your MMR(quite complicated). easier would be to just show you your MMR.
        Spoiler: this won't happen because dota players should trust mm system 100% and be sure that their rating is always correct( read terms and conditions).

        Woof Woof

          didnt read but i will tell you one thing valve dont give a fuck about any suggestions unless they are posted on reddit and upvoted by tons of casuals

          BogiDotA

            I hope there will be soon mmr or dbr like it existed already but nothing new with that..

            Woof Woof

              also those turds there are mainly responsible for destroying dbr

              Woof Woof

                my last hopes about visible rating ever vanished when valve fucked dotabuff it was more than obvious that they care only about casuals and ladder or visible rating wont ever be a part of dota 2 client itself or any valves site

                Relentless

                  You are right Bogidota. Its not convenient, and if you are nowhere near the top you just can't do it because it would require looking through way to many pages...but this is one good way to try to figure out your hidden MMR level. Normally it might be impossible to find yourself with many thousands of pages of games...but if you sort by hero you can dramatically cut down the number of games on the live pages and see %wise how close you are to the top.

                  If you are in the top 10% and there are only 100 pages of games for a certain hero you can find yourself in the first 10 pages. That is much easier than searching 600 pages to fnd the top 10% of all 6,000 pages.

                  Keep in mind that being on page 5 does not necessarily mean you are rated higher than being on page 6 because the order depends on the average MMR of the teams. If you que with players rated above you...then you will be on a team with a higher average MMR. Contrapositively if you que with players weaker than you would will end up on a team with a lower average MMR and be farther back on the live pages.

                  Bu yorum düzenlendi
                  Chris.

                    The measurment you chose is quite inaqurate. You don't always get matched up with evenly skilled players. There is often a huge gap, sometimes I play on P1 sometimes on P20. It always depends on who you get matched up with.

                    Relentless

                      That's true but if you qued with 4 other 4k+ rated players you would be very likely to be first page. Again...its clumsy but you could use this to figure you who in your party moved the game up or down.

                      If you sort on a hero that is played a lot there could still be a ton of games...but with more games its more accurate. Pudge currently has 1650 pages of games going, while Chen has 174. A top 1% player should be in the first 16-17 pages of pudge...that's hard to find still. A top 1% player should be on the first 1-2 pages of Chen...that's easy to find.

                      You do not have to be playing the hero you search. It just has to be in your game. If you know which heroes are more rare you can use them to check more easily. Again keep in mind one check doesn't give you a very precise idea of your score. You should do maybe 5 with the same party and average it to be more certain.

                      Bu yorum düzenlendi
                      Chris.

                        Why filter it for heros? It makes the whole thing more inaccurate. Just check the first 20 pages and if you are in those you are top .5% or some similar percentage.

                        Relentless

                          I just suggest filtering for heroes to expand access to more people. It does make it a lot less accurate if you are trying to decide whether you are top 0.5% or top 1%...but if you are only looking to determine top 10% or top 20% its fine.

                          Bu yorum düzenlendi
                          Vaikiss`742.

                            u can't messure players skill in team games unless this game is played by robots who makes no mistakes has no emotions and shit like that

                            otherwise every system is flawled by things mentioned in this post

                            6_din_49

                              @Relentless
                              I think the pudge distribution is not very uniform on each bracket. My bet is that ~97% of pudge games are in normal bracket. Can someone with plus give us some numbers?

                              Relentless

                                86.5% of Pudge games this month were in Gold+Silver+Bronze....that roughly corresponds to Normal. Normal is supposed to be 85% of players so Pudge is only slightly over-represented. Pudge is popular in every skill bracket. He is only edged out for most popular in the Diamond games and only just barely there by NP.

                                Vaikiss I agree that measuring individual skill in a team game is very difficult. But this page checking thing is just to give people a rough idea of where they stand.

                                Bu yorum düzenlendi
                                Gerry

                                  Pudge Breakdown:

                                  Bronze 61,191
                                  Silver 386,279
                                  Gold 1,886,374
                                  Diamond 121,792
                                  Platinum 645,133

                                  Bu yorum düzenlendi
                                  Gerry

                                    Class Pudge Matches Total Matches Percent Used (Pudge / 1/10 of Total)
                                    1 Bronze 61,191 1,573,871 38.9%
                                    2 Silver 386,279 10,957,100 35.3%
                                    3 Gold 1,886,374 57,068,913 33.1%
                                    4 Platinum 645,133 19,481,385 33.1%
                                    5 Diamond 121,792 3,838,825 31.7%

                                    Sorry... don't know how to format this better.

                                    Bu yorum düzenlendi
                                    BogiDotA

                                      Relentless you are right,as i said this can be well calculate only if you play solo mm(without party). :)

                                      SmallKek

                                        its not actually that accurate to calculate like that. i can conclude that ur currently better than 97% of people in rating. however u have to check ur opponent profile. to be diamond at 1000 games u need at least 53% although there are some exception people with 51-52% made it. at 500 games at least 55%. it can be very easily determined if u have reach the standard. to be appearing in top few pages usually do not 100% determined u as a diamond rating. even i play in 1st page of live game in this low game account and i dont think i am diamond. u have to be able to play in first few pages consistently. like 9 out of 10 times

                                        Bu yorum düzenlendi
                                        SayRay

                                          Diamond doesn't even exist anymore. DBR is gone, why do people obsess over Diamond and all that still?