Jump to content


The Matchmaker – How It Works


  • Please log in to reply
399 replies to this topic

Community #1 Posted 16 June 2014 - 04:01 PM

    Sergeant

  • Content Team
  • 0 battles
  • 25,696
  • Member since:
    11-09-2011
Curious about the system that assembles teams for random battles? Check out our new infographic to find out.
 

WPT_Vuteek #2 Posted 16 June 2014 - 04:28 PM

    Private

  • Player
  • 5906 battles
  • 10
  • Member since:
    12-28-2013

This is okay, but when will you start matching players according to their win rate %.

So many matches where you get like 30% chance to win is incredible stupid.

I know that queue time would be higher if you consider dividing players according to their win%, but everybody would like to see interesting close win matches.



Pater_Leppich #3 Posted 16 June 2014 - 04:33 PM

    Warrant Officer

  • Player
  • 53928 battles
  • 930
  • [_LGS_] _LGS_
  • Member since:
    02-22-2011

*
POPULAR

Xvm's winchance formula is not made by wargaming and most surely they won't balance the mm around some 3rd party application. Just disable the damn thing.

 

EDIT: Negrep for saying my opinion, whoever you are have fun with your 5% chance your brain gives you for discussion.


Edited by Psarandis, 16 June 2014 - 05:21 PM.


Knowno #4 Posted 16 June 2014 - 04:43 PM

    Private

  • Player
  • 14292 battles
  • 9
  • [EGYPT] EGYPT
  • Member since:
    01-31-2014

WG is a car MM is the engine.

 

So we must accept some malfunctioning and imbalanced battles ,but the players are afraid of a conspiracy theory that the lost games are sometimes intentional :)



Alecsandru010 #5 Posted 16 June 2014 - 04:48 PM

    Warrant Officer

  • Beta Tester
  • 34423 battles
  • 719
  • Member since:
    09-15-2010
It doesn't. :bush:

Smokun #6 Posted 16 June 2014 - 04:48 PM

    Private

  • Beta Tester
  • 12598 battles
  • 26
  • Member since:
    08-20-2010
so 'random' -.-

Monster95alpha19 #7 Posted 16 June 2014 - 04:52 PM

    Brigadier

  • Player
  • 35389 battles
  • 4,000
  • [TKEA] TKEA
  • Member since:
    10-28-2011

WG, what i don't understand is why aren't you using your own ranking system ( Personal Rating ) in making balanced teams.For example: The difference between the top PR player and worst PR player to be around 2000-3000 points.Or, to make it by WR.

 

Oh, you think its balanced? :

Spoiler

 


Edited by Titaniu, 16 June 2014 - 04:54 PM.


Lord_Demon #8 Posted 16 June 2014 - 04:52 PM

    Colonel

  • Player
  • 32658 battles
  • 3,643
  • Member since:
    06-01-2011

View PostWPT_Vuteek, on 16 June 2014 - 05:28 PM, said:

This is okay, but when will you start matching players according to their win rate %.

So many matches where you get like 30% chance to win is incredible stupid.

I know that queue time would be higher if you consider dividing players according to their win%, but everybody would like to see interesting close win matches.

How exactly would this work in the long run? Grouping the worst players that now have WR below 40% in same battle would mean half of them win, and their WR would constantlymove towards 50%.

 

Same for the really good players.

 

It might increase the extremes a bit, but would practically make everyone from the average (say 46-54% WR) just move toward the 50% WR.

 

 



malasOmerta #9 Posted 16 June 2014 - 04:54 PM

    Private

  • Player
  • 24738 battles
  • 8
  • [LTUTT] LTUTT
  • Member since:
    05-19-2012

just separate the players into 2 groups:

1. All players who have win rate lower than 48%. These are usually very bad players or bots.

2. All the rest players with win rate higher than 48%.

 

And make MM for these 2 groups separate, so that no shamelessly bad players get into the experienced player battles.

This would remove the biggest pains and negative effects of the bot-like players.

 

Nothing fancy or prolonging the battle queues.



Zedrick #10 Posted 16 June 2014 - 04:55 PM

    Staff Sergeant

  • Player
  • 15260 battles
  • 350
  • Member since:
    11-03-2012

View PostPsarandis, on 16 June 2014 - 04:33 PM, said:

Xvm's winchance formula is not made by wargaming and most surely they won't balance the mm around some 3rd party application. Just disable the damn thing.

 

I think WPT_Vuteek was talking about the winrate and not XVM's winchance-guess. which can be off. Taking the winrate into account would be useful for more balanced games, I'm guessing they don't because it would lead to longer queues.



Spanguolis #11 Posted 16 June 2014 - 04:56 PM

    Lance-corporal

  • Player
  • 21158 battles
  • 99
  • Member since:
    09-29-2013
I have simply no comments...:facepalm: 

Attached Thumbnails



Pater_Leppich #12 Posted 16 June 2014 - 04:57 PM

    Warrant Officer

  • Player
  • 53928 battles
  • 930
  • [_LGS_] _LGS_
  • Member since:
    02-22-2011

View PostZedrick, on 16 June 2014 - 04:55 PM, said:

 

I think WPT_Vuteek was talking about the winrate and not XVM's winchance-guess. which can be off. Taking the winrate into account would be useful for more balanced games, I'm guessing they don't because it would lead to longer queues.


Talking about "30% chance to win", i doubt he means something else


Edited by Psarandis, 16 June 2014 - 04:57 PM.


WPT_Vuteek #13 Posted 16 June 2014 - 05:02 PM

    Private

  • Player
  • 5906 battles
  • 10
  • Member since:
    12-28-2013

View PostLord_Demon, on 16 June 2014 - 05:52 PM, said:

How exactly would this work in the long run? Grouping the worst players that now have WR below 40% in same battle would mean half of them win, and their WR would constantlymove towards 50%.

 

Same for the really good players.

 

It might increase the extremes a bit, but would practically make everyone from the average (say 46-54% WR) just move toward the 50% WR.

 

 

Valve did exactly this in their game Dota 2.
YOu get 25 pts when you win and you lose 25pts when you lose, You got rating number that describe you were you belong and after a while when you reach your actual rating number you play games with ppl who are skilled just like you are.

 

Spoiler

 


Edited by WPT_Vuteek, 16 June 2014 - 05:05 PM.


StealthAce #14 Posted 16 June 2014 - 05:05 PM

    Private

  • Player
  • 16202 battles
  • 15
  • [-FU-] -FU-
  • Member since:
    01-27-2014

I think I see there extraordinary situations more often than one should see...

 

One question regarding this. Why doesn't it take more time to organise the games? Most of the times it just pushes me in any game I fit in princible but it has not been one or two times when I am the only on in lowest tier and rest are one or two tiers higher. I am not talking about scout tanks now but just ordinary tier 4-6 medium tanks or TDs... 

 

When it takes about 20-30 sec to arrange the game, it always feels like mm works better. WG I am not in that kind of hurry so take your time with mm and make decent matches.


Edited by StealthAce, 16 June 2014 - 05:09 PM.


WPT_Vuteek #15 Posted 16 June 2014 - 05:10 PM

    Private

  • Player
  • 5906 battles
  • 10
  • Member since:
    12-28-2013

View PostZedrick, on 16 June 2014 - 05:55 PM, said:

 

I think WPT_Vuteek was talking about the winrate and not XVM's winchance-guess. which can be off. Taking the winrate into account would be useful for more balanced games, I'm guessing they don't because it would lead to longer queues.

I was talking about winrate of a player to be considered into matchmaking, not a XVM winchance.

I don't agree about that weight balance ratio they are mentioning in explanation, because most of high skill players think maus is just a sack of hp. it doesn't mean maus should be considered as more usefull than WT auf for example.



Raffy5 #16 Posted 16 June 2014 - 05:11 PM

    Lieutenant

  • Player
  • 26711 battles
  • 1,977
  • Member since:
    07-09-2012

The funny thing is what the guy that did the video explaining how MM works and when he said something like:

"MM makes team so the result is unpredictable". If you balance the teams just by tank weight what the guy said there is a BIG FAT LIE, or maybe the russian to english translation was not done properly.


Edited by pussinbootsraffy, 16 June 2014 - 05:13 PM.


jagdcommander #17 Posted 16 June 2014 - 05:14 PM

    Warrant Officer

  • Player
  • 39623 battles
  • 860
  • Member since:
    03-18-2011

View PostPsarandis, on 16 June 2014 - 05:33 PM, said:

Xvm's winchance formula is not made by wargaming and most surely they won't balance the mm around some 3rd party application. Just disable the damn thing.

 

Why do some players use XVM? Because they believe it gives them (and quite possibly does) an advantage over players who stick with the standard game. It is cheating! Personally, I'd ban all external mods if it were possible to make the game the same for everyone.

 

I can't understand why players get upset over MM. Over the long term, even though with differing vehicles it's impossible to match perfectly, you will be on the better side 50% of the time anyway.

 

Damn, just reintroduce Karl.



Knowno #18 Posted 16 June 2014 - 05:16 PM

    Private

  • Player
  • 14292 battles
  • 9
  • [EGYPT] EGYPT
  • Member since:
    01-31-2014
why no one is appreciating the processing power working on this .. so whining tankers :izmena:

Trollhammeren #19 Posted 16 June 2014 - 05:16 PM

    Private

  • Player
  • 23331 battles
  • 45
  • Member since:
    07-29-2012
Yes, that's why it is so lovely when you t-34-76 platoon of two gets into a match only with you two being the only mediums and t5 in the game .... you can just feel the balance ... the wide open balance of it :D Also when grinding a stock tank it seems that you always get the worst tiering 9/10 times. Yes its not rigged :P Feels like you are in a Casino , house always wins. 

Master77 #20 Posted 16 June 2014 - 05:16 PM

    Private

  • Player
  • 42190 battles
  • 33
  • Member since:
    07-23-2011
Skill based MM is smth that should be considered URGENTLY.  I would rather want WG to allocate more resources for finding a solution to this problem than to ...errr lets say ... improving the game graphics, or sounds ...etc.




1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users