Valve never said they where changing their algorithm though. So why exactly are people surprised that it is similar to the system they had before? People just seem to think they are much better than they actually are.
Examples of this are players who play heroes like QoP, WR, BH etc who have many many ways to escape so when they lose their score may seem better than their teammates BUT they had little to no positive impact on the game.
Mute am not the stats man but when I get the link from another forum poster(I have seen a couple of them posted in the multiple MM QQ threads) will you concede that you are wrong and let it go?
you dont seem to understand that it's not actually considered as your MMR as long as you dont have 10 games
it's like TMM based on your MMR from unranked games
and i dont know why people think it's okay to play vs 4-5 stacks when everyone on your team's solo
that's basically saying: there are 5 shit players who are supposed to play together against on average higher skilled individual players having shit teamwork
i don't think a system based on that concept is considered good
me going and winning every match I play against 8 year olds in basketball does not mean I should be in the NBA
you do not understand it, you should stop there before you make a fool out of yourself and start complaining about something you do not understand. Now if you came here asking "how does MM work I am a little confused on it" you might not have come off as a tool
@atum blabla
Everything counts in MMR, 1 guy queued along with a party, a 5-man party, they won 10 matches in a row, still, however, everyone had his own party MMR, and actually the best player along them got the highest MMR, so yes, KDA counts for sure.
I just played a game with legit russians with 4.3k tmmr. I don't know why there is so much hate against russians, you just gave to adjust your playstyle to be a lot more aggresive and don't hesitate before you go in. Btw, I've met american players that've behaved a lot worse than most russians
@Tinkering`Around
well there was a game where i did poorly (>10 deaths) but still won and in the next one my KDA was over 10 and both increased my mmr for 'bout the same
It's not broken, it's just seems like we're all starting new accounts all over and either have to be good enough to drag the terribles to victory or slowly gain rating to where we aren't queued with them as much. I get matched with retards in regular matches but this is much worse.
Also what happened to the no parties when solo option?
@atum 金槍魚
In the first game where u fed, u still won, so u should gain MMR and not lose them (although u can lose sometimes).
In the second game u did much better, and won, so u gain MMR.
However, I am pretty sure u gained much more points in the second game where u did well than in the first game.
Also I should add that, u can somehow lose MMR even when u win, when does this happen?
When u r playing against players with similar-lower MMRs, play bad, and still win.
Though this might be really difficult to achieve.
yea people are not even doing the 10 matches before they complain about how broken the system is.................even then you need more matches to find where you fit
Ok guys here we go,wanna share with you some information. All those I m checking with console with commands as already many of you know(also 1 more game and 10 already). 4445 mmr after 7 games,after 8 loss 4400 mmr. 9 game also lose but with score 22-7-10 or smth and was best in my team but my team was rly bad and I lost but now I m 4500 mmr. So you can get some mmr also when loss if you played best in your team seems like.
Also consider that the ranked system tries to queue with players who have similar experience at dota.
How does the system gather this information?
By surveying all the players. Just kidding. It does so by counting the number of games u played, and queuing u against ones with close ones, this would affect the gaming experience u are having.
@BogiDotO
That explains wut I mentioned earlier, there are cases where u lose MMR even when u win, and vice versa.
Kinda stupid but I am pretty sure the rating banks it mostly off the win and how much of a stomp it was....this is from what I have seen/heard from others.
e.g. GOLD/EXP difference plays a part - W/L seems to play the biggest part - KDA probably has a small handle.
So, as playing as a support with low KDA, Gold and EXP along with the loss means you will have a shit end game rating...not impressed.
havoc u dont seem to understand how this works, I played support 6/10 games, like solo support. Yet I managed to get 5000+
Just accept that you are not good as you think you are.
@wacoo mega lols dude I don't see you playing supports ever, basically greedy heroes and still that win rate tsk tsk for shame....
oh and I am as aware as any as to how this works..if you would care to read I stated from what I have heard from others and seen from people and my own rating.
I consider myself an average player but a pretty decent support. From the structure of your post "I played support 6/10 games, like solo support. Yet I managed to get 5000+" its seems you would like to cast illusions of grandeur on people, however the lack of evidence supporting your claim I find equally humorous.
Edit: It seems to set from the first game then fluctuate by about 30-50 points for each additional game. Anyone actually finished the 10 and has the start and finish score so we can see the spread and if it actually "calibrates"
I think it starts from your normal MMR with ~800 uncertainty and at the end of your games you'll end up somewhere between (MMR - uncertainty) and (MMR + uncertainty), where uncertainty gets lower after each game.
I heard about this dota account command long time ago. I read it somewhere on playdota or like that...
I just won test game and mmr went from 3551 to 3533.. so you can actually get - rank if you won but played not so good ( had 2-5 + some 15 assists with ds)
Really disappointing that you just start at your hidden mmr (which most people should have an idea of already if they played TMM at all) and there's virtually no change up or down during the 10 games. Would have been nice to get a clean slate.
I agree cause lets take my account for instance, in the early transitions of dota2 I played like trash on a stick and found it hard to adjust but then got friends and trained a little and managed to move out of the Dota1 mindset and it rapidly started to improve.
The old stuff will forever haunt me and look pretty bad all round but meh, c'est la vie e.g my invoker :P
a clean slate for old as balls accounts does look appealing
Valve never said they where changing their algorithm though. So why exactly are people surprised that it is similar to the system they had before? People just seem to think they are much better than they actually are.
Examples of this are players who play heroes like QoP, WR, BH etc who have many many ways to escape so when they lose their score may seem better than their teammates BUT they had little to no positive impact on the game.
Also winrate is not what determines skill
there are plenty of players on page 1 games with under 50% win rates
It only matters on who you play against
show me one page 1 player with wr below 50%
Me 4445 just saw at console,must say this is really accurate for me. Good job Volvo = )
Mute am not the stats man but when I get the link from another forum poster(I have seen a couple of them posted in the multiple MM QQ threads) will you concede that you are wrong and let it go?
yes but it wont happen becaue they are full of shit theres no such a thing as top page player with winrate below 50%
you dont seem to understand that it's not actually considered as your MMR as long as you dont have 10 games
it's like TMM based on your MMR from unranked games
and i dont know why people think it's okay to play vs 4-5 stacks when everyone on your team's solo
that's basically saying: there are 5 shit players who are supposed to play together against on average higher skilled individual players having shit teamwork
i don't think a system based on that concept is considered good
So Mute can you say this exact phrase
"I will never post about MM ever again if you can link a player who has played on page 1 with under 50% win/loss"
I want you to write it out
solo_competitive_rank: 3678
solo_competitive_rank_uncertainty: 643
solo_calibration_games_remaining: 6
Sampson never go full retard and ask people for such shit.
dunno how that uncertainty thing works
Idk, I got 5043 solo. Most games were with or against stacks... doesn't really matter tbh.
I will never post about MM ever again if you can link a player who has played on page 1 with under 50% win/loss
competitive_rank: 4520
competitive_rank_uncertainty: 737
calibration_games_remaining: 8
expecting slightly higher, but not by much, perhaps 4600... so gj valve!
http://dotabuff.com/players/150855327
http://dotabuff.com/matches/410767581
never post again please
damn i went full retard i forgot about all those failed smurfs with 0 wins playing top pages xO you got me there bastard
http://dotabuff.com/players/124296606
http://dotabuff.com/matches/408933733
you dont understand how little win/ loss means.....
only thing i dont understand is how poorly designed mm can be
me going and winning every match I play against 8 year olds in basketball does not mean I should be in the NBA
you do not understand it, you should stop there before you make a fool out of yourself and start complaining about something you do not understand. Now if you came here asking "how does MM work I am a little confused on it" you might not have come off as a tool
4700 after 1 win & 2 losses, ohwell
The system is broken, second ranked game. My team were retarded russians and the other team were a 3 man party.....
won 5 out of 5 solo ranked games so far. my rating seems to increase by about ~100 after each win. KDAs doesnt seem to matter.
Because when you where doing the calibration matches you fought 3 friends in the same party therefore the system is broken?
@atum blabla
Everything counts in MMR, 1 guy queued along with a party, a 5-man party, they won 10 matches in a row, still, however, everyone had his own party MMR, and actually the best player along them got the highest MMR, so yes, KDA counts for sure.
I just played a game with legit russians with 4.3k tmmr. I don't know why there is so much hate against russians, you just gave to adjust your playstyle to be a lot more aggresive and don't hesitate before you go in. Btw, I've met american players that've behaved a lot worse than most russians
@Tinkering`Around
well there was a game where i did poorly (>10 deaths) but still won and in the next one my KDA was over 10 and both increased my mmr for 'bout the same
It's not broken, it's just seems like we're all starting new accounts all over and either have to be good enough to drag the terribles to victory or slowly gain rating to where we aren't queued with them as much. I get matched with retards in regular matches but this is much worse.
Also what happened to the no parties when solo option?
^ that option got removed, check the patch notes
@atum 金槍魚
In the first game where u fed, u still won, so u should gain MMR and not lose them (although u can lose sometimes).
In the second game u did much better, and won, so u gain MMR.
However, I am pretty sure u gained much more points in the second game where u did well than in the first game.
Also I should add that, u can somehow lose MMR even when u win, when does this happen?
When u r playing against players with similar-lower MMRs, play bad, and still win.
Though this might be really difficult to achieve.
yea people are not even doing the 10 matches before they complain about how broken the system is.................even then you need more matches to find where you fit
Ok guys here we go,wanna share with you some information. All those I m checking with console with commands as already many of you know(also 1 more game and 10 already). 4445 mmr after 7 games,after 8 loss 4400 mmr. 9 game also lose but with score 22-7-10 or smth and was best in my team but my team was rly bad and I lost but now I m 4500 mmr. So you can get some mmr also when loss if you played best in your team seems like.
do not think much about the number you get after the first match, alot of people are getting like 700+ uncertainty......
Also consider that the ranked system tries to queue with players who have similar experience at dota.
How does the system gather this information?
By surveying all the players. Just kidding. It does so by counting the number of games u played, and queuing u against ones with close ones, this would affect the gaming experience u are having.
@BogiDotO
That explains wut I mentioned earlier, there are cases where u lose MMR even when u win, and vice versa.
after 2 wins
competitive_rank: 4526
competitive_rank_uncertainty: 743
calibration_games_remaining: 8
solo_competitive_rank: 3947
solo_competitive_rank_uncertainty: 864
solo_calibration_games_remaining: 7
1-2 w/l
Kinda stupid but I am pretty sure the rating banks it mostly off the win and how much of a stomp it was....this is from what I have seen/heard from others.
e.g. GOLD/EXP difference plays a part - W/L seems to play the biggest part - KDA probably has a small handle.
So, as playing as a support with low KDA, Gold and EXP along with the loss means you will have a shit end game rating...not impressed.
competitive_rank: 4197
competitive_rank_uncertainty: 901
calibration_games_remaining: 9
doto unsure about me
After a single loss :(
havoc u dont seem to understand how this works, I played support 6/10 games, like solo support. Yet I managed to get 5000+
Just accept that you are not good as you think you are.
After 1 loss solo and 1 win and 1 loss with stack of 4.
competitive_rank: 3517
competitive_rank_uncertainty: 658
calibration_games_remaining: 8
solo_competitive_rank: 3482
solo_competitive_rank_uncertainty: 719
solo_calibration_games_remaining: 9
Edit: After one more loss solo, rank actually went up!
solo_competitive_rank: 3495
solo_competitive_rank_uncertainty: 689
solo_calibration_games_remaining: 8
LE: Won solo game, rank went down! Yeah, I played like shit but had good team! :D
solo_competitive_rank: 3439
solo_competitive_rank_uncertainty: 683
solo_calibration_games_remaining: 7
Another win
solo_competitive_rank: 3494
solo_competitive_rank_uncertainty: 647
solo_calibration_games_remaining: 6
Another win
solo_competitive_rank: 3508
solo_competitive_rank_uncertainty: 610
solo_calibration_games_remaining: 5
After 1 loss in a party then 2 solo queue wins
competitive_rank: 3023
competitive_rank_uncertainty: 783
calibration_games_remaining: 9
solo_competitive_rank: 3090
solo_competitive_rank_uncertainty: 733
solo_calibration_games_remaining: 8
OA*_Havoc Badger
It not depends on win/loss , just the way you play. I lost 7 games 3 wins in this 10 ranked. And got 4500
solo_competitive_rank: 4112
solo_competitive_rank_uncertainty: 607
solo_calibration_games_remaining: 6
2 wins / 2 losses.
@wacoo mega lols dude I don't see you playing supports ever, basically greedy heroes and still that win rate tsk tsk for shame....
oh and I am as aware as any as to how this works..if you would care to read I stated from what I have heard from others and seen from people and my own rating.
I consider myself an average player but a pretty decent support. From the structure of your post "I played support 6/10 games, like solo support. Yet I managed to get 5000+" its seems you would like to cast illusions of grandeur on people, however the lack of evidence supporting your claim I find equally humorous.
Edit: It seems to set from the first game then fluctuate by about 30-50 points for each additional game. Anyone actually finished the 10 and has the start and finish score so we can see the spread and if it actually "calibrates"
What is considered good?
I'm 1-2 after 3 games and my rating is over 5k.
is 4.2K good? I think if i play some more games and win, it'll be higher
I think it starts from your normal MMR with ~800 uncertainty and at the end of your games you'll end up somewhere between (MMR - uncertainty) and (MMR + uncertainty), where uncertainty gets lower after each game.
6_din_49 Not really you are really wrong. You start at your mmr which you already (invisible) have and always had. that 10 games are just casual.
^And what did I say? "Normal MMR" is the invisible one.
I heard about this dota account command long time ago. I read it somewhere on playdota or like that...
I just won test game and mmr went from 3551 to 3533.. so you can actually get - rank if you won but played not so good ( had 2-5 + some 15 assists with ds)
Really disappointing that you just start at your hidden mmr (which most people should have an idea of already if they played TMM at all) and there's virtually no change up or down during the 10 games. Would have been nice to get a clean slate.
I agree cause lets take my account for instance, in the early transitions of dota2 I played like trash on a stick and found it hard to adjust but then got friends and trained a little and managed to move out of the Dota1 mindset and it rapidly started to improve.
The old stuff will forever haunt me and look pretty bad all round but meh, c'est la vie e.g my invoker :P
a clean slate for old as balls accounts does look appealing