Page 2 of 6 FirstFirst 1 2 3 4 ... LastLast
Results 11 to 20 of 59

Thread: MMR very low after calibration, even after winning 9 out of 10 games

  1. #11
    Basic Member Shmendrich's Avatar
    Join Date
    Oct 2012
    Posts
    2,579
    Quote Originally Posted by Lajkonik View Post
    http://blog.dota2.com/2013/12/matchmaking/

    The median is 2250 because previously new players were asked about their experience (beginner and intermediate players started at 2000 and advanced players started at 3000). Nowadays new accounts start at 3000 IIRC.

    New accounts are tested by smurf detection in their first few [unranked] games to prevent smurfs from ruining games for new players. This is why you can have high MMR with a low winrate - you can boost the account to your [more or less; +/- 300 points] actual rating in 5-10 games. In case you overshoot, stomp first 5 games, get a rating 500 higher than your actual skill level - say 5.5k - and then proceed to lose the next 10 games, you will have 5.25k rating with 33% winrate.
    Er....I read that post a while ago, and it doesnt says anything like that. I just checked it again and no.

    1 Doesnt says players start with 3000
    2 It doesnt says that your existing rank is used when you start playing ranked to balance you.

    Thats the 2 facts he gave and neither is contained in that post...I dont see why you even linked it.

    And as for the rest of your post again it sounds like an educated guess, but I dont see anything backed up by any facts. (particularly the numbers you are using)

    We know about smurf detection, uncertainty and such, but most of the numbers I just dont see where you are getting them from except for the median rank.
    Make a NON automated report system.
    I would rather wait 10 minutes for a balanced, fun and exciting match that will last 1+ hours. Than wait 10 seconds for a poorly balanced crapmatch that will end in 30- minutes
    Valve, How I'm I supposed to respect certain boundaries to prevent a mute if I dont know them because they are different for each and every person on each and every game I play?
    The only type of player that dislikes having his stats public are the players that suck and cheaters

  2. #12
    Basic Member
    Join Date
    Sep 2013
    Posts
    88
    Pick someone that can deal insane amount of hero damage and stomp hard... boom, 4k at least. This work even if you lose more than win.

  3. #13
    Basic Member Shmendrich's Avatar
    Join Date
    Oct 2012
    Posts
    2,579
    Quote Originally Posted by sirolatiato View Post
    Pick someone that can deal insane amount of hero damage and stomp hard... boom, 4k at least. This work even if you lose more than win.
    It would be pretty stupid from Valve to take hero damage in to consideration. A good support is 10 times better player than a good carry. Its way more difficult to play support properly.
    Make a NON automated report system.
    I would rather wait 10 minutes for a balanced, fun and exciting match that will last 1+ hours. Than wait 10 seconds for a poorly balanced crapmatch that will end in 30- minutes
    Valve, How I'm I supposed to respect certain boundaries to prevent a mute if I dont know them because they are different for each and every person on each and every game I play?
    The only type of player that dislikes having his stats public are the players that suck and cheaters

  4. #14
    Basic Member
    Join Date
    Mar 2012
    Location
    Perth, Western Australia
    Posts
    241
    State the average rating of your opponents in all your calibration games. This is all that matters.

  5. #15
    Basic Member
    Join Date
    Mar 2014
    Posts
    16
    Quote Originally Posted by TcheQ View Post
    State the average rating of your opponents in all your calibration games. This is all that matters.
    How should I know that?

    And more importantly, so if I lose against 10 randomly selected pro gamer 5 stacks, I get 5k+ mmr, but if I beat 10 randomly selected noob teams, I only get 2k? I don't have high opinion about valve and it's developers (especially how I got practically lied to and misled by them, wasting a shitload of time and some money) but this would be way stupider than I can imagine them to be .

  6. #16
    Basic Member hoveringmover's Avatar
    Join Date
    Feb 2013
    Posts
    2,509
    Quote Originally Posted by Shmendrich View Post
    It would be pretty stupid from Valve to take hero damage in to consideration. A good support is 10 times better player than a good carry. Its way more difficult to play support properly.
    It has already been established that they are doing this.
    I will mercilessly add to my ignore list anyone that makes an incredibly annoying signature.

  7. #17
    Basic Member Paranoia14's Avatar
    Join Date
    Jul 2013
    Posts
    161
    Quote Originally Posted by totemizer View Post
    How should I know that?

    And more importantly, so if I lose against 10 randomly selected pro gamer 5 stacks, I get 5k+ mmr, but if I beat 10 randomly selected noob teams, I only get 2k? I don't have high opinion about valve and it's developers (especially how I got practically lied to and misled by them, wasting a shitload of time and some money) but this would be way stupider than I can imagine them to be .
    I couldn't agree with you more. It wasn't too difficult to implement a decent system. When they brought in 'Ranked' I actually had hopes in the game again and I resumed playing it only to find out ranked was the new normal matchmaking since normal had gone down to absolute crap. All in all, it's all still bugged as hell but they don't need to care since they've made 80+ million out of it in a couple of months only in compendium sales even with a bugged system. Life's good.

  8. #18
    Basic Member McMillanMe's Avatar
    Join Date
    Apr 2013
    Location
    Russia, Ivanovo
    Posts
    191
    6+9\5 = 10.20? lmao.
    Counting mmr takes not only winrate, but your impact in game.

  9. #19
    Basic Member
    Join Date
    Nov 2012
    Posts
    161
    Its probably bad idea to use heroes like Omni, Dazzle and Treant in calibration matches. Their hero damage and tower damage will look very low even then compared to other supports, and Valve probably cant be assed to create separate impact calculation formula for every hero.

  10. #20
    Basic Member Crowfeather's Avatar
    Join Date
    Jan 2013
    Posts
    3,319
    Its very easy to smurf a high MMr, that's why a lot of people do it who are fed up with being in the "trenches".

    Its also why so many 4k+ get really pissy about their teammates who have low game count.
    Quote Originally Posted by Typhox View Post
    Nope. Windows Explorer is not a database
    Take a look into the mirror! You think you are completely unfailable. With your limited knowledge, you think that you know more than someone who is very deep in the matter/an expert.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •