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

Hurk #1741 Posted Mar 13 2018 - 17:51

    Major

  • Players
  • 55852 battles
  • 17,382
  • [KGR] KGR
  • Member since:
    09-30-2012

View PostSoTrue, on Mar 12 2018 - 07:16, said:

 

No such thing as a fixed game you say....

Spoiler

When the 'random' mm puts all the skill onto one team THAT IS A FIXED GAME.  There is no other way to look at it.  Burying your head in the sand, chalking it up to 's**t happens', making yourself feel better because sometimes you get the good team....   Are all absolutely weak excuses for the crap that is random mm.  We can clearly see these mismatches happen.  It's very easy to fix.  That WG hasn't yet shouldn't stop, we the consumers, from logging our complaints about it and making them change it.

-

No one is saying we are 100% in control of the match.  What we are saying is MM is 100% control of the match.  Battles like the one above happen every time I've ever played.  (Not every battle, but certainly several times a session).  This is crap.  It needs to stop.  We are tired of it.

 

Playing in a 'fixed' outcome game is about 'gameplay'.  Sure, I can set myself up to farm damage and experience - but that's not really what I'd call 'good gameplay'.  Give me a level playing field every time...

one anecdote is not proof of anything. 

there is nothing fixed about the team you are on. the next battle you are no longer on that team, thus its random, its anything but fixed. 

MM doesnt control anything except a queue of people waiting to play.

learn to win 5% chance games. for those that dont pay attention, thats a 1 in 20 chance. you can win these matches, if your team simply doesnt play "average" and the enemy team doesnt play "average". etc. 

 

the next match, when you are on the other team, the results are different. the only constant is YOU. YOU determine your win rate. not the MM. 



NeatoMan #1742 Posted Mar 13 2018 - 18:20

    Major

  • Players
  • 28198 battles
  • 20,721
  • Member since:
    06-28-2011

View Posteteam, on Mar 13 2018 - 11:42, said:

First of all I for one don't feel the least bit cornered.

The behavior of skilled players vs unskilled players in game is different and recognizable.  If it wasn't you wouldn't be worried about having more reds on your team with SBMM.

XVM is just the tool that shows there are many matches that are not balanced and no different than you using data and spread sheets to justify your position.

So based on your turn off XVM logic we could go to some form of SBMM and you could not track skill compression and everything would be fine.  Right?

I wouldn't be able to tell the difference between SBMM from what we got now.   I wouldn't notice any improvement in game play whatsoever, and I'd still see bad players doing everything possible to lose games.   What I would notice is that it will be harder to win, since the MM would actively be trying to balance me all the time.

 

On a separate note, SBMM may actually decrease damage farming, because farming would needlessly push up your stats, only making it harder to keep winning against the SBMM.  You might be better off letting the tomatoes fight over the scraps of a victory.   hmmm...



SoTrue #1743 Posted Mar 13 2018 - 18:59

    Major

  • Players
  • 33655 battles
  • 3,302
  • [ATTOP] ATTOP
  • Member since:
    04-01-2011

View PostHurk, on Mar 13 2018 - 08:51, said:

one anecdote is not proof of anything.

there is nothing fixed about the team you are on. the next battle you are no longer on that team, thus its random, its anything but fixed.

MM doesnt control anything except a queue of people waiting to play.

learn to win 5% chance games. for those that dont pay attention, thats a 1 in 20 chance. you can win these matches, if your team simply doesnt play "average" and the enemy team doesnt play "average". etc.

 

the next match, when you are on the other team, the results are different. the only constant is YOU. YOU determine your win rate. not the MM.

 

ha ha ha ha ha ha ha.   "learn to win the 5% chance games"   ha ha ha ha ha...   Oh good, one, really.

SoTrue #1744 Posted Mar 13 2018 - 19:02

    Major

  • Players
  • 33655 battles
  • 3,302
  • [ATTOP] ATTOP
  • Member since:
    04-01-2011

View PostKliphie, on Mar 13 2018 - 06:39, said:

 

As it's never been used in the game, what is the source of this "real" data?

 

View PostNunya_000, on Mar 13 2018 - 06:41, said:

 

He does not understand the difference between hypothesis and proof.

 

The real data is the data collected by members of the community.  The data shows the distribution of the types of battles in terms of skill imbalance.  The data has shown about 20% of the battles are fixed in your favor, about 20% are fixed against you, and about 60% are between roughly equal teams.   Here's what you need to know, coming to a discussion without any data to back up your position, when debating those who have brought data - makes you look like fool.  Do us all a favor, install xvm, record your results and come back when you have something to support you position.

SoTrue #1745 Posted Mar 13 2018 - 19:04

    Major

  • Players
  • 33655 battles
  • 3,302
  • [ATTOP] ATTOP
  • Member since:
    04-01-2011

View PostNeatoMan, on Mar 13 2018 - 07:53, said:

It has been shown to happen in only one tank for you, and only one tank for me.  That is not really proof.  At best you can say it can be done in certain specific instances.

 

"it has been shown to happen", thank you for admitting defeat and that I'm right.  If you, and the other random mm fan boys, were right - it would be impossible for anyone to have better than 50% win rate under sbmm.  Yet we've both managed to do it.

Kliphie #1746 Posted Mar 13 2018 - 19:42

    Major

  • Players
  • 32996 battles
  • 5,578
  • [GFLC] GFLC
  • Member since:
    07-20-2012

View PostSoTrue, on Mar 13 2018 - 13:02, said:

 

 

The real data is the data collected by members of the community.  The data shows the distribution of the types of battles in terms of skill imbalance.  The data has shown about 20% of the battles are fixed in your favor, about 20% are fixed against you, and about 60% are between roughly equal teams.   Here's what you need to know, coming to a discussion without any data to back up your position, when debating those who have brought data - makes you look like fool.  Do us all a favor, install xvm, record your results and come back when you have something to support you position.

 

So if it happens naturally 60% of the time and the other 40% cancel each other out, what would SBMM actually do other than give actual credence to the tinfoiler's belief that WG is "rigging" the outcome?  

eteam #1747 Posted Mar 13 2018 - 20:03

    Sergeant

  • Players
  • 74210 battles
  • 244
  • Member since:
    05-06-2013

View PostNeatoMan, on Mar 13 2018 - 17:20, said:

I wouldn't be able to tell the difference between SBMM from what we got now.   I wouldn't notice any improvement in game play whatsoever, and I'd still see bad players doing everything possible to lose games.   What I would notice is that it will be harder to win, since the MM would actively be trying to balance me all the time.

 

On a separate note, SBMM may actually decrease damage farming, because farming would needlessly push up your stats, only making it harder to keep winning against the SBMM.  You might be better off letting the tomatoes fight over the scraps of a victory.   hmmm...

 

​I believe that the probability of a good game experience is dramatically increased when teams are balanced.  (I guess good game play experience is subjective and we weight things differently depending on perspective)

Yes some underdog teams win and yes some mismatched team games can provide a good game experience. But over the long term imo skill balanced games have a higher probability of providing a better game experience just because those very unbalanced games are unlikely to provide it for me. When you put the majority of reds on one team the mistakes are compounded further detracting from the game experience.  I also believe that over the longer term you would notice a change in game play experience while acknowledging that it would be harder for you to win.



StiffWind #1748 Posted Mar 13 2018 - 20:43

    Major

  • -Players-
  • 18479 battles
  • 2,076
  • [MOV] MOV
  • Member since:
    03-15-2017

View PostMiseryMachine, on Mar 13 2018 - 06:32, said:

I wonder what masochistic behavior keeps people like this Stiffwind here. He is bad at the game and sore at better players, criticize the game but won't leave! lol. Such a loser.

 

I stick around just to irritate sychophantic slags just like you.  Didn't you know that already?

Now...let's see just what kind of clueless wonder you really are.  Define "bad at the game".



StiffWind #1749 Posted Mar 13 2018 - 20:46

    Major

  • -Players-
  • 18479 battles
  • 2,076
  • [MOV] MOV
  • Member since:
    03-15-2017

View PostStaz211, on Mar 13 2018 - 06:33, said:

 

You literally just proved yourself wrong.

 

Rigged denotes intent. You admit the matches are random. With random matches, there will be diminishing values as you approach either extreme end of the bell curve. That is the inherent nature of a random, unmanipulated data. Matches at either end aren't rigged/fixed, they are the product of randomness. Matches at either end aren't unfair, because they will be balanced out with a relatively equal number of matches at the other end of the curve. With that in mind, as you yourself admit, the only thing that moves a player along the curve is individual skill.

 

You just proved yourself dumb.  Rigged DOES imply intent...in this case, to make battles end more quickly.  However, the rigging of this nature also produces other consequences.  If you actually took the time trying to understand the post BEFORE ignorantly slagging it, you might not stick your foot in your mouth so often.  Just some advice.

 



StiffWind #1750 Posted Mar 13 2018 - 20:46

    Major

  • -Players-
  • 18479 battles
  • 2,076
  • [MOV] MOV
  • Member since:
    03-15-2017

View PostJer1413, on Mar 13 2018 - 12:02, said:

 

Of course it's not true. Better players will always have better odds of winning, just like everything else in life.

 

You said it yourself "all things being equal". Well guess what, skill is not equal between players.

 

What part of "all things being equal" do you not understand?

 



NeatoMan #1751 Posted Mar 13 2018 - 21:17

    Major

  • Players
  • 28198 battles
  • 20,721
  • Member since:
    06-28-2011

View Posteteam, on Mar 13 2018 - 14:03, said:

​I believe that the probability of a good game experience is dramatically increased when teams are balanced.  (I guess good game play experience is subjective and we weight things differently depending on perspective)

 

Yes some underdog teams win and yes some mismatched team games can provide a good game experience. But over the long term imo skill balanced games have a higher probability of providing a better game experience just because those very unbalanced games are unlikely to provide it for me. When you put the majority of reds on one team the mistakes are compounded further detracting from the game experience.  I also believe that over the longer term you would notice a change in game play experience while acknowledging that it would be harder for you to win.

but I haven't seen anything that shows those heavily unbalanced games are common enough that getting rid of them affects overall game play very much.  The only way I've seen anyone identify a large number "crap" games is by XVM.  I consider games crap that play like crap, not games that some stupid number is tells me is crap.  There is a HUGE difference.  

 

Your crap and budha's crap is not the same as what I consider crap.   Why should I get all excited about SBMM when it does absolutely nothing to improve my games, makes the game harder for me, and I will still see just as much crap? 
 

View PostSoTrue, on Mar 13 2018 - 13:04, said:

"it has been shown to happen", thank you for admitting defeat and that I'm right.  If you, and the other random mm fan boys, were right - it would be impossible for anyone to have better than 50% win rate under sbmm.  Yet we've both managed to do it.

A broken clock is also correct twice a day.  If that's how you judge success, then no wonder why you keep coming up with the conclusions you do.



Hurk #1752 Posted Mar 13 2018 - 21:37

    Major

  • Players
  • 55852 battles
  • 17,382
  • [KGR] KGR
  • Member since:
    09-30-2012

View PostSoTrue, on Mar 13 2018 - 11:04, said:

"it has been shown to happen", thank you for admitting defeat and that I'm right.  If you, and the other random mm fan boys, were right - it would be impossible for anyone to have better than 50% win rate under sbmm.  Yet we've both managed to do it.

 

Hearhtstone is a game with skill based matching that dynamically matches players in an immediate fashion at the start of each match, you are placed against players of equal or higher skill, and only against lower skill if queue times drag on. as its a 1v1 matching, it suffers much less than WoT would in building a "proper" match. 

 

now look at the decks. 

https://hsreplay.net...#sortBy=winrate

 

oh dear... decks with a 60% win rate due to balance issues. 

WoT is no different. 

 

oh... and, you are wrong. the only thing SBMM will do is shrink the width of the bell curve. instead of players averaging 40-65%, they will instead average 46-55%; at that point will you then argue that SBMM is "rigged" and "fixing matches" ? 



eteam #1753 Posted Mar 13 2018 - 22:27

    Sergeant

  • Players
  • 74210 battles
  • 244
  • Member since:
    05-06-2013

View PostNeatoMan, on Mar 13 2018 - 20:17, said:

but I haven't seen anything that shows those heavily unbalanced games are common enough that getting rid of them affects overall game play very much.  The only way I've seen anyone identify a large number "crap" games is by XVM.  I consider games crap that play like crap, not games that some stupid number is tells me is crap.  There is a HUGE difference.  

 

Your crap and budha's crap is not the same as what I consider crap.   Why should I get all excited about SBMM when it does absolutely nothing to improve my games, makes the game harder for me, and I will still see just as much crap? 
 

A broken clock is also correct twice a day.  If that's how you judge success, then no wonder why you keep coming up with the conclusions you do.

 

​In one of the last posts you suggest that you see bad players doing everything to lose games.  So does it not make sense that if you put all those so called bad players on one team there will be a compounding effect and so the probability of it playing out like crap is a lot more likely?  I am not suggesting you get excited about or accept SBMM as a solution but just acknowledge the fact that huge mismatched games are not likely to produce a good game experience for anyone despite our different perceptions of crap games.



Jer1413 #1754 Posted Mar 13 2018 - 22:49

    Captain

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

View PostStiffWind, on Mar 13 2018 - 19:46, said:

 

What part of "all things being equal" do you not understand?

 

 

The part where one player is better than another, and subsequently wins more.

Staz211 #1755 Posted Mar 13 2018 - 23:28

    Major

  • Players
  • 27045 battles
  • 3,906
  • Member since:
    06-11-2012

View PostEmperorJuliusCaesar, on Mar 13 2018 - 07:42, said:

 

I, and many other with green recents, that are clearly carrying out weight and more, that are sick of getting stuck on the red team, or being put on the other side of very imbalanced games.  We don't want auto-wins or auto-losses that are 40% of our games.  That's not a challenge, that's slot-machine based MM.

 

Those of us with Purple recents/overalls get stuck on the red team too. The difference between these two groups is that one group can turn more red team games into wins than the other group. There is no auto win or auto loss for 40% of your games. There is no slot machine. You play better, you win more games; you play worse, you lose more games. Nothing in that system is rigged. You, and others arguing your position, refuse to take accountability of your own play. If you feel like matches are out of your control to win or positively influence, then you aren't that great a player. It's that simple.

 

 

View PostSoTrue, on Mar 13 2018 - 09:34, said:

 

Yes, there are 40% of all battles are fixed by the mm.  The current mm does not balance teams based on skill.  About 40% of all battles occur between significantly mismatched teams.  The data has been presented, the data does not lie.  How does it determine the outcome before hand?  Did you really just ask that question?  If I put all pros on one team, and all noobs on the other team - I just determined the outcome before the start. Fact.

 

It has been proven with real data that sbmm will NOT push everyone to 50%.  Get with the times man...

 

First statement: You can repeat this all you want, but it doesn't make it true. The data not only does not support your position, but has proven it to be incorrect by people in your own camp. You just made the same exact arguments myself and other refuted only a few posts ago. Saying the same disproven statement over and over isnt going to change reality. 

 

Second statement: Blatantly false. In the system you are proposing there can be no outliers. Everyone will be forced towards 50%. 

 

View PostStiffWind, on Mar 13 2018 - 14:46, said:

 

You just proved yourself dumb.  Rigged DOES imply intent...in this case, to make battles end more quickly.  However, the rigging of this nature also produces other consequences.  If you actually took the time trying to understand the post BEFORE ignorantly slagging it, you might not stick your foot in your mouth so often.  Just some advice.

 

 

I take responses that contain insults and no substantive argument as admission of defeat. Thank you for your concession. 



NeatoMan #1756 Posted Mar 13 2018 - 23:47

    Major

  • Players
  • 28198 battles
  • 20,721
  • Member since:
    06-28-2011

View Posteteam, on Mar 13 2018 - 16:27, said:

 

​In one of the last posts you suggest that you see bad players doing everything to lose games.  So does it not make sense that if you put all those so called bad players on one team there will be a compounding effect and so the probability of it playing out like crap is a lot more likely?  I am not suggesting you get excited about or accept SBMM as a solution but just acknowledge the fact that huge mismatched games are not likely to produce a good game experience for anyone despite our different perceptions of crap games.

From what I've seen it's more a matter of incompatible play styles.  Yoloers cause one kind of issue, timid players and campers cause another.   Most teams can survive a few of each.   Too many of one kind and it's a problem, no matter how balanced you make the stats on both teams.  

 

The huge skill mismatches do not come close to 40% of games.  I see the huge mismatches begin to affect games beyond 2:1 odds, and even then not every one of those end up being bad games. That leaves a large number of games that are perfectly fine as far as skill goes.  Budha claims it's not about who wins and loses, but that's exactly what SBMM is doing a large majority of the time.

 

You want to get rid of blowouts with SBMM, yet are addressing only a small part of what causes them, and in the process are sweeping up a large number of perfectly competitive games, and redistributing the wins and losses.



Staz211 #1757 Posted Mar 13 2018 - 23:49

    Major

  • Players
  • 27045 battles
  • 3,906
  • Member since:
    06-11-2012

View PostSoTrue, on Mar 13 2018 - 12:59, said:

 

ha ha ha ha ha ha ha.   "learn to win the 5% chance games"   ha ha ha ha ha...   Oh good, one, really.

 

Out of 100 games, you will win 5 of them. Hence 5%. 

 

It's this attitude that separates good players from bad players. This is why bad players feel like the MM is rigged; because they're not good enough to actually see the impact they can have on a match. People win those 5% chances.



Staz211 #1758 Posted Mar 14 2018 - 00:04

    Major

  • Players
  • 27045 battles
  • 3,906
  • Member since:
    06-11-2012

This thread is a good resource for helping people to understand why some blow outs happen, since people seem to be hung up on blow out matches. Read it, understand it, apply it. If you do, the matches you play will begin to make a heck of a lot more sense, as opposed to blaming losses on rigging. 

 

http://forum-console...ss-math-of-wot/



AMSRocker #1759 Posted Mar 14 2018 - 01:17

    Major

  • Players
  • 39263 battles
  • 6,087
  • [SPRAY] SPRAY
  • Member since:
    11-07-2011
Page 88 and counting... :popcorn:

EmperorJuliusCaesar #1760 Posted Mar 14 2018 - 04:18

    Major

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

View PostNeatoMan, on Mar 13 2018 - 05:23, said:

View PostEmperorJuliusCaesar, on Mar 13 2018 - 08:15, said:

I've not had battles be over in less than 3 mins when they weren't very unbalanced. 3 mins is a super fast game.  If you supposedly don't have XVM, then you wouldn't know, the difference is, WE DO KNOW AND WE KNOW WHY IT HAPPENS.

Out of the 1200+ battles since they introduced the new MM I have only had one battle that was over in less than 3 minutes, and only 50 that were over in less than 4 minutes, and 5 of those were not even close to being blowouts (they were fast cap wins when no one had a clear advantage in numbers).   40% of all battles?   Without XVM less than 5% of my battles fit your criteria. 

 

You're not helping your case when it comes to you stating MM is truly random.  If that's the MM you get, I'm quite jealous.  I get on both sides of sub and barely over 3 min games nearly daily.  Or could it be that you don't actually play much anymore and spend most of your time in the forums defending this crap slot machine MM we have now, the one that drives players away from the game.






3 user(s) are reading this topic

0 members, 3 guests, 0 anonymous users