General DiscussionBringing back the old MMR method (diamond, platinum, gold, etc)

pros are at 5-6k so if u got 4-5k ur "decent" pubber

bamboo wants attention for his high mmr....

Hi,

I've only been playing for the past 2 years, but instead of the old method which kind of sounds similar to SC2 ratings which I was used to.

I figure wouldn't it be better just to place a numerical value associated with your rank instead of a bracket like that? Like "50th place in the world out of 2.5 million people" it seems the numerical MMR points is kind of similar to a bracket rating of Diamond/platinum/gold etc.

Back years ago during first year of SC2 I was placed in Diamond league, then they expanded out to "master league" and "grand master League" so Grand masters would be top 200 in the world, and masters would be the top 1%. I was in master league once, but wasn't aware out of how many people I was the top 1% in. It would put you in a bracket in the masters league out of 100 people, but there were several brackets of master leagues which was not in any particular order...

Sorry long story short I figure a numerical value of where you stand out of x amount of players would be better. Sorry for the long post.

its impossible to say what the exact percentiles are but u can make a rough estimate.

assume that everyone above 5k exclusively play on the top 10 pages of live games.
assume that everyone above 5k play an average of 2 hours a day.
assume that there are 3 miliion active players in dota 2

there are 60 players on each page of live games so the total amount of players above 5k is (60*10)*(24/2)= 6000 players

above 5k would then be the top 0.2 percentile (99.8-100).

this would imo correlate well with the numbers for unranked matchmaking where >4100 is the top 1 percentile (99-100).

dude. i just want to know if im still better than 99% ppl like i used to be. because i notice the mmr system changes and the pressure i get from game wasnt the same anymore.

and 4100 isnt = top 1%( i can confirm 100%)
the statistic is wrong

u are definately better than 99% of ppl at this game if u are at 4.7k. i think thats safe to say.

if u were closer to 4k i would be more unsure.

additionally: my calculations were only estimates, feel free to come up with better ones urself

anything below 5k is shit, i dont even give a fuck if you got 2k, 3k or 4k, ur garbage at dota and should give up

theres some major stupid shit going on with 5k+ rated players and if it was up to me, i would put those players in 2k-3k where they belong but sadly volvo will never do such a thing

mmr is very inflated, with a fresh smurf account, you could easily get 5.5k to 6k if you manage to win 95% of your games until you reach level 13, i know this guy started with 5.9k but now he has 5k or something http://dotabuff.com/players/161098039, i dont need to point out his main acc has 50% wr either #russiastronk

they should increase it to amount of matches played such as 500 instead of level 13, that would be wonderful

says kami

what is this wave guy trying to say

He is trying to say that he cannot accept losing at any cost and if he can't win nearly 100% of games by whatever outside the game tactics it's the games fault... AND if you don't also tryhard in the super extreme way he does.., then you suck. But all his near 100% win accounts keep getting reported so now he just makes no-games accounts to troll the forum in hopes that by constant repetition of his supposed greatness people will believe he can easily get to the very highest levels MMR in the world and win nearly all games.

But he is not very good at expressing himself. So I thought I would help.

thats some good bait, especially where you called me a troll, have you tried playing dota and winning a match instead of analyzing shit out of your ass and assuming things wrong 100%?

for analyzing, you're not even doing a decent job at it and the truth is, you're still stuck at 30000 with so "much" awareness

didn't you quit dota right after you found out your ranked? link me at least 3 of my supposedly "100% win accounts"

oh hey i remember spectating one of my friends matches and it was yours, i even took a screenshot of it last month

http://steamcommunity.com/sharedfiles/filedetails/?id=201873943

fucking please dont play dota, you embarrass your own rating

Some people really like masturbating to numbers.

@wave Ok bro, you pro.Someone will give a fuck about it soon don't worry.

wave u mad?

relentless is spamming friend requests on me, what a kid lol

^Haha too funny lol

he left a huge post on my steam profile so now its back to private, i just dont know why these kids have to lie about me on the internet LoL

Why would somebody lie on the internet?

with modern technology scientists have created a way for other users to experience how relentless felt like when his rating turned out to be 3000

[IMG]http://puu.sh/6wYXu.gif[/IMG]

Because of dbuff reputation (too important)

incoming damage control by relentless in 10 9 8 7 6 5 4 3 2 1 if he havent killed himself yet ..

just realized i said 30000 and not 3000, i bet relentless felt really happy there huh wont even bother to edit i t

@Zano

That banana looks tasty, very want!

Funny banana dog. I don't think abusing people on the thread really answers the question for bamboo.

Good point Cpt. Caveman. So to answer the question...

One way to look at it in hard numbers, since top 1% of all dota players is not very precise, is to look at tMMR scores. Only about 1.4% of dota players have a current tMMR score but its mostly the more serious players (whether high or low skill) who bother to put together a team and actually play team ranked games.

It is still about 110,000 players. Of those teams only 0.56% are above 4750. As a fraction of all dota players... 0.0016% have tMMR over 4750.

To put it into words 4750 MMR is an extremely high score share by tiny, tiny fraction of the dota world... and even high enough to look pretty good among the honest and some exaggerated self reporting of scores on dotabuff. There are pro teams rated in the top 50 teams in the world whose players have average scores near 4750. Although teams in the top 20 would be more like 5k to 5.5k average if they made a team for tMMR. The highest anyone has ever obtained was 5.8k tMMR.

why is it only 1.4% have total mmr

Team MMR. Most people simply don't use it. You have to actually make a team and then play 3 games to get one. Most just stack.

http://dotabuff.com/teams
You can view the ladder here.

The main problem with team matching is you cannot do challenge matches. So people nearly always play teams with ratings far above them or below them since so few are in the que. This is why the current #1 team named themselves "Volo fix TMM pls".

If you go through the games for the first page of teams with ratings over 5k you will see that they almost never play each other. All the games for nearly all the top rated teams were played against teams in the 4.5k to 4.8k range because there are almost never two 5k + teams in the que at the same time. And when the 5.2k team easily beats the 4.5k team they gain zero points... which makes the whole team MMR pointless.

But when you start a team the initial tMMR is the average of your personal MMR scores.

I like how relentless doesn't give a crap about wave... This is a good lesson. Please ignore attention whores who mindlessly flame.