Jump to content


how come you don't take skill/win rate into account in matchmaking


  • Please log in to reply
1967 replies to this topic

TerrorJoe #1461 Posted Feb 28 2018 - 17:52

    Corporal

  • Players
  • 20114 battles
  • 38
  • Member since:
    04-23-2011

View PostNeatoMan, on Feb 28 2018 - 15:15, said:

If the teams are so obviously stacked then it should be no problem at all to tell them apart without the need for stats.  Yet when challenged to do so, budha did no better at picking out the stacked teams than flipping a coin.   The only way he could claim he could tell games apart is when he narrowed his criteria for balance beyond what even he claims is acceptable, and then guessing everything else was unbalanced (i.e. he stacked the criteria in his favor).  And this was only looking at blowouts, which he claimed should be the easiest to tell apart.   He never even attempted to look at the remaining 75%-80% of games that aren't blowouts to see if he could tell them apart.

 

In addition, these games that are so stacked that they lead to certain victory are in no way as common as he pretends.  You and him both present the extreme example as if it is the norm.  The examples you present are outliers even among the games classified as unacceptable by him.

 

​... Again.... You CAN tell them apart. The difference is, XVM gives you visual indication ****PRIOR**** to match starting. The visual indicator that exists, is shown to be quite accurate, therefore it is just as reasonable to measure the outcomes for a large sample size of battles and the statistics for the players in each match to come to the same conclusions.

​This seems like you two are competing for who can commit the most logical fallacies. Just because a tool exists that simplifies the data gathering, or even automates it, does not mean without it... the claim in which it proves is any less substantiated.
Both of you are essentially claiming "You are obtaining your evidence too efficiently, you must obtain it the more difficult way otherwise it proves nothing".  That alone is 3 logical fallacies. You might as well admit to trolling now.

​Now.... can XVM indicators have a psychological effect on some users that ultimately ends up influencing the match? Sure... anything is possible. Impossible to realistically prove however.

​What has been proven in this very thread is the exact due course of topic. The current MM is not Random. The current MM does stack teams a majority of the time (Higher than 10% diff in comp quality). SB/SWAP MM would solve the issues of current MM without increasing Que times.

Kliphie #1462 Posted Feb 28 2018 - 18:02

    Major

  • Players
  • 32996 battles
  • 5,562
  • [GFLC] GFLC
  • Member since:
    07-20-2012

View Postkruppw, on Feb 28 2018 - 10:52, said:

What has been proven in this very thread is the exact due course of topic. The current MM is not Random. The current MM does stack teams a majority of the time (Higher than 10% diff in comp quality). 

 

  I've seen this claimed often but the proof is never presented.  Perhaps it was somewhere in the previous 73 pages.  I've seen plenty of evidence showing imbalanced teams but nothing to indicate that they were formed by design.  

Nixeldon #1463 Posted Feb 28 2018 - 18:02

    Major

  • Players
  • 60881 battles
  • 2,287
  • Member since:
    10-30-2011

View Postkruppw, on Feb 28 2018 - 11:52, said:

Both of you are essentially claiming "You are obtaining your evidence too efficiently, you must obtain it the more difficult way otherwise it proves nothing".  That alone is 3 logical fallacies. You might as well admit to trolling now.

 

I am saying you are just as clueless as budha. It has nothing to do with efficiency. You don't know how a balanced match should play.

 

Why balance a match if it does nothing more than satisfy OCD? The only fallacy being committed is you, like budha, can't answer a simple question and resort to circular reasoning. 

 

 

 



HalfPastZulu #1464 Posted Feb 28 2018 - 18:45

    First lieutenant

  • Players
  • 24183 battles
  • 825
  • [ITE] ITE
  • Member since:
    10-02-2013

View PostHisNameWasSethRich, on Jan 23 2018 - 16:31, said:

Skill based MM does not work. This has been argued already.

 

Skill based MM does work they just don't want to do it for business reasons.

 

I see many negative WT references but since i play both I can tell you that the match ups were better in WT when they used skill based and thats why they flocked to WT from WoT in droves. They now use something more akin to WoT which is driving popularity down again.



TerrorJoe #1465 Posted Feb 28 2018 - 19:18

    Corporal

  • Players
  • 20114 battles
  • 38
  • Member since:
    04-23-2011

View PostNixeldon, on Feb 28 2018 - 17:02, said:

 

I am saying you are just as clueless as budha. It has nothing to do with efficiency. You don't know how a balanced match should play.

 

Why balance a match if it does nothing more than satisfy OCD? The only fallacy being committed is you, like budha, can't answer a simple question and resort to circular reasoning. 

 

 

 

 

​So more of the same... Red herring... Ad hominem... appeal to the stone....

"Why balance a match if it does nothing more than to satisfy OCD"?   Sorry but that gets you a 0/10 and a troll report.

​Still waiting on this "Simple question"

NeatoMan #1466 Posted Feb 28 2018 - 21:42

    Major

  • Players
  • 28198 battles
  • 20,718
  • Member since:
    06-28-2011

View Postkruppw, on Feb 28 2018 - 11:52, said:

​... Again.... You CAN tell them apart. The difference is, XVM gives you visual indication ****PRIOR**** to match starting. The visual indicator that exists, is shown to be quite accurate, therefore it is just as reasonable to measure the outcomes for a large sample size of battles and the statistics for the players in each match to come to the same conclusions.

what do you tell non-XVM users when nothing has changed from their perspective?  Just as many blowouts, games aren't noticeably closer.     "you just gotta trust us, this is SOOOO much better" :sceptic:

 

given budha's constant butchering of math and statistics I wouldn't trust him on any of this.



Chester380 #1467 Posted Mar 01 2018 - 02:33

    Captain

  • Players
  • 33803 battles
  • 1,191
  • Member since:
    03-09-2012

View PostNeatoMan, on Feb 28 2018 - 15:42, said:

what do you tell non-XVM users when nothing has changed from their perspective?  Just as many blowouts, games aren't noticeably closer.     "you just gotta trust us, this is SOOOO much better" :sceptic:

 

given budha's constant butchering of math and statistics I wouldn't trust him on any of this.

 

People don't need XVM to know a game sucks...are you dense?

EmperorJuliusCaesar #1468 Posted Mar 01 2018 - 03:01

    Major

  • Players
  • 37052 battles
  • 5,744
  • [EOR] EOR
  • Member since:
    03-16-2014

View PostNixeldon, on Feb 28 2018 - 02:39, said:

View PostEmperorJuliusCaesar, on Feb 28 2018 - 05:26, said:

 

You're of course free to lie to yourself all you want, but no one is buying it.  We know better, it's far too easy to see.

I have no reason to lie about anything.

 

How do you know a mismatch?

 

 

When it's a sub 4 min steamroll, it's unbalanced.  When you see a Type 5 platoon that waited in queue long enough for MM to "loosen" the MM rules, it's unbalanced.  Some prefer to keep their head in the sand, I understand.



EmperorJuliusCaesar #1469 Posted Mar 01 2018 - 03:03

    Major

  • Players
  • 37052 battles
  • 5,744
  • [EOR] EOR
  • Member since:
    03-16-2014

View PostJer1413, on Feb 28 2018 - 04:14, said:

View PostSoTrue, on Feb 28 2018 - 02:55, said:

 

 

Sure, better players skew their own bell curve.  BUT No matter what your skill level, MM PICKS THE TEAMS.   So you can be a unicum and get 15 tomatoes, while the enemy gets 3 uniucms, 10 greens, and 2 yellows.  That is 100% mm rigging that battle.

 

 

 

You're arguing based on an extreme example that almost never happens. And if it does, I deal with the battle as it plays out and move on to the next one. I don't curl up in a little ball, murmuring "no fair, no fair".

 

 

 

 

It happens often enough that thread after thread get posted about the 3-4 min blowouts that are all too common now.  It happens often enough that people, especially newer players see MM for the slot machine win/loss decider before the battle starts, it happens often enough that it's causing WG to lose players and openly state they know MM is an issue and will keep working on it until they get it right.  Many of us don't think it will be right until they take skill into account.  Those that know how to game the current system of course, don't want that change.



EmperorJuliusCaesar #1470 Posted Mar 01 2018 - 03:06

    Major

  • Players
  • 37052 battles
  • 5,744
  • [EOR] EOR
  • Member since:
    03-16-2014

View PostJer1413, on Feb 28 2018 - 04:14, said:

View PostSoTrue, on Feb 28 2018 - 02:55, said:

 

 

Sure, better players skew their own bell curve.  BUT No matter what your skill level, MM PICKS THE TEAMS.   So you can be a unicum and get 15 tomatoes, while the enemy gets 3 uniucms, 10 greens, and 2 yellows.  That is 100% mm rigging that battle.

 

 

 

You're arguing based on an extreme example that almost never happens. And if it does, I deal with the battle as it plays out and move on to the next one. I don't curl up in a little ball, murmuring "no fair, no fair".

 

 

 

 

It happens often enough that thread after thread get posted about the 3-4 min blowouts that are all too common now.  It happens often enough that people, especially newer players see MM for the slot machine win/loss decider before the battle starts, it happens often enough that it's causing WG to lose players and openly state they know MM is an issue and will keep working on it until they get it right.  Many of us don't think it will be right until they take skill into account.



EmperorJuliusCaesar #1471 Posted Mar 01 2018 - 03:10

    Major

  • Players
  • 37052 battles
  • 5,744
  • [EOR] EOR
  • Member since:
    03-16-2014

View PostHalfPastZulu, on Feb 28 2018 - 09:45, said:

View PostHisNameWasSethRich, on Jan 23 2018 - 16:31, said:

Skill based MM does not work. This has been argued already.

 

Skill based MM does work they just don't want to do it for business reasons.

 

I see many negative WT references but since i play both I can tell you that the match ups were better in WT when they used skill based and thats why they flocked to WT from WoT in droves. They now use something more akin to WoT which is driving popularity down again.

 

And that's why you always see posts from the "good" players claiming people left WT because of SBMM.  They, and the "good" players they know left.....because they didn't want a fair fight. 



StiffWind #1472 Posted Mar 01 2018 - 03:14

    Major

  • -Players-
  • 18473 battles
  • 2,076
  • [MOV] MOV
  • Member since:
    03-15-2017

View Post_Gungrave_, on Feb 28 2018 - 15:43, said:

 

Unfortunately the Skill MM trio(SoTrue, StiffWind, and Emperor) is never going to shut up so its highly unlikely that this thread will ever die.

 

Long Live The Thread!

 



Nixeldon #1473 Posted Mar 01 2018 - 04:21

    Major

  • Players
  • 60881 battles
  • 2,287
  • Member since:
    10-30-2011

View PostEmperorJuliusCaesar, on Feb 28 2018 - 21:01, said:

 

When it's a sub 4 min steamroll, it's unbalanced.  When you see a Type 5 platoon that waited in queue long enough for MM to "loosen" the MM rules, it's unbalanced.  Some prefer to keep their head in the sand, I understand.

 

Why couldn't you simply state that blowouts were the issue? Is it because you know that color-coded matches won't reduce blowouts enough to notice a difference?

 

Even budha knows that is the case. That's why he can't define a balanced match except by what XVM tells him.



NeatoMan #1474 Posted Mar 01 2018 - 04:34

    Major

  • Players
  • 28198 battles
  • 20,718
  • Member since:
    06-28-2011

View PostChester380, on Feb 28 2018 - 20:33, said:

People don't need XVM to know a game sucks...are you dense?

Yes there are sucky games (usually described as quick blowouts), the point is you can balance games all you want and there will still be just as many sucky games.   So what have you accomplished other than redistributing wins and losses?

 

you guys keep going back and forth between

"We don't need XVM.. mmm durrr"

and

"Again.... You CAN tell them apart. The difference is, XVM gives you visual indication ****PRIOR**** to match starting... mmm durrr."

 

Which is it?  make up your minds.  What exactly are you trying to address?  skill unbalance or blowout games?  They are not the same thing, and treating one doesn't fix the other.  You are conflating two separate issues.

TerrorJoe #1475 Posted Mar 01 2018 - 13:50

    Corporal

  • Players
  • 20114 battles
  • 38
  • Member since:
    04-23-2011

View PostNeatoMan, on Mar 01 2018 - 03:34, said:

Yes there are sucky games (usually described as quick blowouts), the point is you can balance games all you want and there will still be just as many sucky games.   So what have you accomplished other than redistributing wins and losses?

 

you guys keep going back and forth between

"We don't need XVM.. mmm durrr"

and

"Again.... You CAN tell them apart. The difference is, XVM gives you visual indication ****PRIOR**** to match starting... mmm durrr."

 

Which is it?  make up your minds.  What exactly are you trying to address?  skill unbalance or blowout games?  They are not the same thing, and treating one doesn't fix the other.  You are conflating two separate issues.

 

Oh man..... more fallacies...

​1.) As has been clearly stated it is the Blowouts that have been defined as the "sucky games". But... you know.... keep moving the goal posts.... red herring... false attribution... cherry picking... etc

​2.) Going back and forth? Oh look let us add Straw Man to that list as well.  You should try quoting in context there Genius. Both quotes are stating that XVM is not *needed*.

​No one is asking for perfect 50%v50% games. Which is clearly what you are attempting to also Straw Man here.


​Even with the math that has been presented in this thread where it has been proven that SB/SWAP MM will not increase que times, it was clearly shown under the pretense ruleset that was NOT strict on requiring 50%v50% plucking.

​46% v 54% matchups are NOT the problem.....,.  The 40% v 60% Matchups ARE the problem. The 30% v 70% Matchups ARE the problem.

​For the relative stats of each person combined for each respective team, no matter how you acquire them, if the difference is only something like 8% can easily be overcome by a single death on the team holding the statistical advantage. In those same regards with a difference of something like 15%, now you are talking about the equivalence of one team statistically having an entire +1 player advantage over the other team (Essentially 13v12). Which increases the higher the gap.

​How about the argument from repetition that you both seem to like about "redistributing wins and losses"? Sure it will do that, but ONLY for average players and players whos stats *currently* are completely detached from the realistic skill level (like a unicum who simply platoons all the time or manipulates MM or seal clubs in a lower tier). THOSE players will see their stats drop. That is because skill should have more weight than MM, desires should be met through what is deserved not through luck. Simply for the fact the WR average does NOT hover around 50% currently shows without any doubt the current MM is broken and further proves the math that the current average statistical gap between teams is actually equaling the difference of an entire player (and more in many cases).

​So what *actual* defense against SB/SWAP MM is there other than these so called "Good" players, afraid that just maybe... just maybe their stats are a complete misrepresentation of their actual skill levels in a more fair environment?



 

Edited by kruppw, Mar 01 2018 - 13:53.


Jer1413 #1476 Posted Mar 01 2018 - 13:55

    Captain

  • Players
  • 48111 battles
  • 1,546
  • [RR13] RR13
  • Member since:
    02-24-2013

View Postkruppw, on Mar 01 2018 - 12:50, said:

 



​1.) As has been clearly stated it is the Blowouts that have been defined as the "sucky games". But... you know.... keep moving the goal posts.... red herring... false attribution... cherry picking... etc




 

 

 

But without XVM, how do you know the Blowouts were caused by an un-balanced match-up? Balanced matches end in a blow-out all the time.

 

 



NeatoMan #1477 Posted Mar 01 2018 - 14:01

    Major

  • Players
  • 28198 battles
  • 20,718
  • Member since:
    06-28-2011

View Postkruppw, on Mar 01 2018 - 07:50, said:

​1.) As has been clearly stated it is the Blowouts that have been defined as the "sucky games". But... you know.... keep moving the goal posts.... red herring... false attribution... cherry picking... etc


​46% v 54% matchups are NOT the problem.....,.  The 40% v 60% Matchups ARE the problem. The 30% v 70% Matchups ARE the problem.

Other than changing which team wins, how will that affect blowouts?   Nothing in the data suggests that getting rid of those unacceptable XVM matchups will significantly affect the number of blowouts, or game duration.   It's just balance for balance (XVM's) sake.



_Gungrave_ #1478 Posted Mar 01 2018 - 14:23

    Major

  • Players
  • 45503 battles
  • 16,299
  • [-SRP-] -SRP-
  • Member since:
    12-07-2011

View PostStiffWind, on Mar 01 2018 - 03:14, said:

Long Live The Thread!

 

I honestly think the only reason you three keep this nonsense thread going is because you guys enjoy the circle jerk of giving each other forum rep.

da_Rock002 #1479 Posted Mar 01 2018 - 14:24

    Major

  • -Players-
  • 9704 battles
  • 3,770
  • Member since:
    11-24-2016

View PostJer1413, on Mar 01 2018 - 07:55, said:

 

 

But without XVM, how do you know the Blowouts were caused by an un-balanced match-up? Balanced matches end in a blow-out all the time.

 

 

 

 

Don't know how he knows, but after a battle, the Teams report has the info.   All you have to do is click on the players to see each side's stats.


 

Balanced matches do end in blowouts.   However, it's far from being all the time.


 

The real problem is that around 40% of the matchups WoT creates are so unbalanced the greatly stronger side naturally kills off the newbie side in a heartbeat.    The belief that some of those actually result in upsets is fantasy.    The matchups that create the 40% roflstomp statistic are what is wrong with this situation.    The top players side almost never loses.    Like almost never.   There is a "top players side" in those about 100% of the time.    So when do the top players who've wound up in those battles experience "random" outcomes?    Almost never.    So they benefit almost every time.   So who else is in those battles and what do they experience?    The newbies are usually about half the losing sides.   Some newbies do wind up on the winning side....   about 1/3 of those roflstomps they see.    About 1/3 the time isn't random wins.   About 2/3 the time isn't random losses.  They're losses all the same.    

 

So those battles naturally occur to help the top players and naturally screw the newbies.    THAT'S WHAT'S WRONG.


Edited by da_Rock002, Mar 01 2018 - 14:27.


NeatoMan #1480 Posted Mar 01 2018 - 14:43

    Major

  • Players
  • 28198 battles
  • 20,718
  • Member since:
    06-28-2011

View Postda_Rock002, on Mar 01 2018 - 08:24, said:

Don't know how he knows, but after a battle, the Teams report has the info.   All you have to do is click on the players to see each side's stats.

Which is doing the exact the same thing XVM is doing. 

 

Block Quote

Balanced matches do end in blowouts.   However, it's far from being all the time.

 Nor do unbalanced matches always end in blowouts either, especially when classifying such a broad range of games as unbalanced, as budha does.

 

Block Quote

The real problem is that around 40% of the matchups WoT creates are so unbalanced the greatly stronger side naturally kills off the newbie side in a heartbeat.    The belief that some of those actually result in upsets is fantasy.    The matchups that create the 40% roflstomp statistic are what is wrong with this situation. 

 It does not happen that often.

 

Block Quote

   The top players side almost never loses.    Like almost never.   There is a "top players side" in those about 100% of the time.    So when do the top players who've wound up in those battles experience "random" outcomes?    Almost never.    So they benefit almost every time.   So who else is in those battles and what do they experience?    The newbies are usually about half the losing sides.   Some newbies do wind up on the winning side....   about 1/3 of those roflstomps they see.    About 1/3 the time isn't random wins.   About 2/3 the time isn't random losses.      So those battles naturally occur to help the top players and naturally screw the newbies.    THAT'S WHAT'S WRONG.

 Yes they win, but do not win in blowouts as often as you claim.  Remove those unbalanced matchups and you'll still get almost the same number of blowouts.

 

 

One out of 10 games are so unbalanced that they are noticeably faster.  Of those only 1 out of 3 lead to blowouts.   That's 3.1% of all games played that have fast blowouts caused by large skill unbalance.   The rest of the games play out like the balanced games, so removing those 10% grossly unbalanced games doesn't change overall blowouts or game duration significantly.






1 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users