Jump to content


Random MM is broken...


  • Please log in to reply
234 replies to this topic

da_Rock002 #221 Posted Jan 12 2020 - 12:44

    Major

  • -Players-
  • 10179 battles
  • 3,946
  • Member since:
    11-24-2016

View Postspud_tuber, on Jan 11 2020 - 09:08, said:

Many of these SBalancedMM proponents claim that ability to sway battles in your favor is a gift from MM and not the result of the player.  Hence terms such as gifted win and win rate welfare.....


You got that wrong...   again.

 

"Many" team-skill-balanced MM proponents don't say anything about the "ability to sway battles" being improved by the MM.    They usually are simply describing the fact when MM allows one sided matchups, the outcomes are strongly biased.    Heck, many of those lopsided matchups are so screwed the WRwelfare team can often underperform and still win. 

 

 As usual....  rewrite what you can't refute.



ATruk #222 Posted Jan 12 2020 - 13:22

    Sergeant

  • Players
  • 61152 battles
  • 224
  • Member since:
    03-16-2013

View PostSimplyPzB2, on Jan 12 2020 - 03:07, said:

Your 'reasons' for not implementing sbmm are nonsensical.  No good player is going to fake bad stats to get better mm.  That's saying better players will "lose more, to win more".  And no, sbmm would not drive every players win rate to 50%.  Sbmm won't create exactly balanced battles.  It will create a range of win chance, somewhere in the neigheborhood of 4-60% chance to win.  Better players will move towards 60%, bad players will move towards 40%.

 

 That depends on your definition of "good". Does "good" mean better stats (win rate or something else)? Does "good" mean ethically and morally sound?

 

 You are quite mistaken to say that "No good player is going to fake bad stats to get better mm." We call them sandbaggers. Every sport with class divisions will have sandbaggers. Every single one. Whether it's for the glory of "winning that class" or for the inevitable separate prizing structures that result from skill class divisions, there will be sandbagging (think "seal clubbing"...).

 

 While SBMM would even the playing field regarding some measure of player skill, it leaves out or ignores more important variables. Any player can destroy Unicum430 or UnicumT92, maybe with some luck if required.

 Now, do I want to battle Green430 or UnicumT92? I fear the Green430 more than the UnicumT92. You seem to think differently.

 

 Wargaming already creates pixel tanks in Tiers, "classes" if you will. MM attempts to balance these Tiers for each battle. That tank I'm in already has class limitations for battle. Player skill is up to each player.

 

 If one is concerned that each battle should be "skill balanced", then we have the problems of "how is skill determined?" and, most importantly "WHO does this balancing? and HOW are they doing this balancing?" I cannot blame Wargaming for unbalanced skill battles because they do not touch skill balancing at all. I can only shake my fist at the sky, because that is where RANDOM lives. Then I move on to the next RANDOM battle.

 

 This game is the only one that has last so many years on my computers. A major factor is that team player skill makeup is NOT manipulated. It is RANDOM. We all play on a world stage, not in little boxes of preconfigured skill levels. There is only one battle button. I would not like to wonder if the battles I play were manipulated by someone. In this way, I know the game company is not favoring me over you, or you over me.



sippo1 #223 Posted Jan 12 2020 - 14:14

    Corporal

  • Players
  • 39109 battles
  • 30
  • Member since:
    12-30-2013

So simply pz protests random MM with a platoon of skilled op tanks. Guess what, i would bot heavies with my platoon to protest SBMM if they ever tried it. I would grab every unicum, blues or green for my team. C'mon you unicums you can carry my crapola platoon, just don't  uninstall  cuz its so fair now.

 

 



Numerius_Titurius_Sophus #224 Posted Jan 12 2020 - 14:33

    Captain

  • -Players-
  • 15146 battles
  • 1,808
  • [GSH] GSH
  • Member since:
    05-07-2015

View PostSimplyPzB2, on Jan 05 2020 - 09:25, said:


But sbmm would make the gameplay better, more fun, less punshing - THUS KEEPING MORE PLAYERS PLAYING - thus make WG MORE MONEY...

 

I disagree on the money part.

 

People who spend much money.  The ones that reroll and buy OP premium tanks and spend much gold on free xp and crews.  Why do they do this?  So they can have an advantage over people who don't do this.

 

Take that away and they will likely spend less money.  They may even quit the game.  Consider a unicum who got his stats partly by doing this.  Is he going to keep playing and 'risk' his stats?  Is he going to reroll and spend much money on it if all his games are about equal skill both sides?

 

While less skilled players/people like me would love what you are suggesting, WG will never do it because it disincentives people to spend money for that advantage.

 

Some highly skilled players would love it because they could excel and, by excelling, drive other supposedly high ranked/skilled players down.  These true unicums would be in the minority though.  Most would run if your suggestion came true.


Edited by Numerius_Titurius_Sophus, Jan 12 2020 - 14:36.


NeatoMan #225 Posted Jan 12 2020 - 14:56

    Major

  • Players
  • 28785 battles
  • 21,272
  • Member since:
    06-28-2011

Block Quote

Sure, if there are only two unicums and one is in a tier 10 heave, and the other is in a tier 8 spg - it's not 'perfectly' balanced when they are on separate teams. BUT IT'S A HECK OF A LOT MORE BALANCED THAN IF THEY ARE BOTH ON THE SAME TEAM.  You missed that part...

 Is it really?  That top tier 10 is still probably still going to win more than the other.   All you done is give it the illusion of balance.  You basically just told one unicum "we're gonna take away a win from you, for no apparent reason other than it's your turn".   How is that any different then what happens in a random MM when he gets a bad team?  At least in random MM he knows if he keeps up his better than average performances he will win more.

 

Block Quote

  A BETTER way to look at very unbalanced battles is to try and find out how many unskilled teams blow out a much higher skilled team

 Why do they have to blow them out?  They just have to defeat them.

 

 

Block Quote

 

View PostNeatoMan, on Jan 10 2020 - 18:53, said:

Go ahead and set up any scenario that you think needs rebalancing, and I guarantee the good players end up with worse team mates more often than not.  It will happen every single time. 

Wrong.  Why can't you understand that it's IMPOSSIBLE FOR ANY PLAYER to have worse teammates under sbmm?  It's just so simple.  If a 'good player' ends up with 14 red tomatoes on his team - IT DOESN'T MATTER BECAUSE THE ENEMY TEAM WILL ALSO HAVE 14 RED TOMATOES.  Relative to the enemy team, the good player will NEVER have 'worse' teammates.

 They will have worse team mates than they had with random team assignment..    You can't deny it.  It is guaranteed to happen.

 

look at your scenario.  under random MM here are the possibilities:

1 purple, 14 red   vs 1 purple, 14 red  (already balanced,  happens half the time)

2 purple, 13 red   vs 15 red  (needs rebalance,  happens half the time)

       rebalance to:

       1 purple, 14 red   vs 1 purple, 14 red 

 

before balance a unicum gets to play with another unicum half the time.  A tomato gets to play with a unicum 3/4 of the time.  Under SBMM a unicum NEVER gets to play with another uncum.  Those 14 other slots will ALWAYS be filled with 14 tomatoes.  A tomato is guaranteed to have a unicum on his team.  He ALWAYS gets to play with a unicum.   So yes he did get worse team mates with SBMM than with random.  The tomato gets better team mates.

 

View PostSimplyPzB2, on Jan 12 2020 - 03:07, said:

Your 'reasons' for not implementing sbmm are nonsensical.  No good player is going to fake bad stats to get better mm.  That's saying better players will "lose more, to win more".

oh but it's perfectly acceptable when you are assigned fake bad stats through a faulty rating system.  

 

Block Quote

And no, sbmm would not drive every players win rate to 50%.  Sbmm won't create exactly balanced battles.  It will create a range of win chance, somewhere in the neigheborhood of 4-60% chance to win.  Better players will move towards 60%, bad players will move towards 40%.

 So SBMM will give good players "win rate welfare"



NeatoMan #226 Posted Jan 12 2020 - 14:59

    Major

  • Players
  • 28785 battles
  • 21,272
  • Member since:
    06-28-2011

View Postsippo1, on Jan 12 2020 - 08:14, said:

So simply pz protests random MM with a platoon of skilled op tanks. Guess what, i would bot heavies with my platoon to protest SBMM if they ever tried it. I would grab every unicum, blues or green for my team. C'mon you unicums you can carry my crapola platoon, just don't  uninstall  cuz its so fair now.

That would be the most effective way to grind up the tech trees in SBMM.  Just yolo every game as fast as you can, and you will still be guaranteed to win ~50% of the time, vs ~40% of the time now (as well as lower shared XP that the losing teams typically get).   You will then build up a low rating for all your games when you do finally decide to play your tanks for real.



NeatoMan #227 Posted Jan 12 2020 - 15:47

    Major

  • Players
  • 28785 battles
  • 21,272
  • Member since:
    06-28-2011

View Postda_Rock002, on Jan 12 2020 - 06:44, said:

  Heck, many of those lopsided matchups are so screwed the WRwelfare team can often underperform and still win.

This happens equally for everybody.   If you do nothing you win ~40% of the time in a random MM, no matter who you are.

 

What you always fail to realize is that everyone's wins/losses will be dependent on how much they contribute to their teams, and not the number of lopsided matchups. 

 

A player with good stats may get higher win chances over his next 1000 games than an average player would, but if he only contributes as much as an average player, he will only win as much as an average player.  A bad player may get lower win chances, but if they contribute as much as a good player over the same time span, they will win as much as a good player.   The win chances they got over those 1000 games are irrelevant.  It's their contributions that will determine how much each will win.

 

Nothing you can say changes that basic fact.  What you do, not the win chances, determines how much you win or lose.... period.

 

By your logic, if a complete tomato purchases a unicum account, he will get win rate welfare from that point forward.  He should expect to keep winning as much as the previous owner of the account, because he is getting higher win chances.



InfinityGamer #228 Posted Jan 14 2020 - 14:39

    Staff sergeant

  • Players
  • 35120 battles
  • 295
  • [GIVUP] GIVUP
  • Member since:
    03-29-2013

View PostNeatoMan, on Jan 13 2020 - 00:47, said:

By your logic, if a complete tomato purchases a unicum account, he will get win rate welfare from that point forward.  He should expect to keep winning as much as the previous owner of the account, because he is getting higher win chances.

 

That is a very good rebuttal to that point which I had not thought of, will be using that in the future.


Edited by InfinityGamer, Jan 14 2020 - 14:50.


InfinityGamer #229 Posted Jan 14 2020 - 15:31

    Staff sergeant

  • Players
  • 35120 battles
  • 295
  • [GIVUP] GIVUP
  • Member since:
    03-29-2013

View PostSimplyPzB2, on Jan 12 2020 - 20:07, said:

Your 'reasons' for not implementing sbmm are nonsensical.  No good player is going to fake bad stats to get better mm.  That's saying better players will "lose more, to win more".  And no, sbmm would not drive every players win rate to 50%.  Sbmm won't create exactly balanced battles.  It will create a range of win chance, somewhere in the neigheborhood of 4-60% chance to win.  Better players will move towards 60%, bad players will move towards 40%.

-

Also, gross misuse of the 'whoopsies'.  I'll explain it to you, it'll be super easy, barely an inconvenience.   There are battles that are so unbalanced that it's impossible for even the best players to win/lose.  Just look at the sample screenshots.  They show a small number of very good players who have no chance to win because the mm put so many more good players on the enemy team.  That's broken.

"No good player is going to fake bad stats to get better mm.  That's saying better players will "lose more, to win more". - That's exactly what they will do... There are already players out there who bot to make their stats look bad so they don't get xvm focused, you think more won't do it with the added incentive of fighting against easier enemies!?

 

"And no, sbmm would not drive every players win rate to 50%.  Sbmm won't create exactly balanced battles. It will create a range of win chance, somewhere in the neigheborhood of 40-60% chance to win." - Each INDIVIDUAL battle might have a win chance of 40-60%, but the thing is that spread will NEVER change. If you're a tomato the spread will be 40-60%, if you're purple the spread will be 40-60%. Having battles that aren't 'exactly balanced' doesn't matter, the point is that no matter how skilled you are you will get just as many 60% win chance battles as you will 40% win chance battles, which will equal out to 50%. SBMM will force the same odds for everyone, it is simply a fact of SBMM.

 

"They show a small number of very good players who have no chance to win because the mm put so many more good players on the enemy team. That's broken." - It's not broken, that's the way it needs to be. You can still have a good game on a losing team.

 

View PostSimplyPzB2, on Jan 12 2020 - 20:07, said:

does that mean you should just suck it up and accept the 20 bad meals you had to eat?

Yes... Those restaurants happened to be bad, oh well, onto the next meal...

But tell you what, I'll use this analogy to explain to you why SBMM forces a 50% win rate. Imagine a complete amateur cook and a pro chef with years of experience. You want them to cook meals of the same quality, but the chef is much better than the amateur, which is unfair. So the chef is given dirtier utensils, a blunter knife and more limited ingredients to work with until their meals are of comparable quality. It does NOT matter how much better the pro chef is, he will just be given worse things until his meals match the quality of the amateur's. This is SBMM with win rate, it wants the teams to be fair, so it gives the pro worse teammates and puts him up against harder enemies. If he still consistently carries his team then the SBMM will see that it's not doing a good job of balancing the teams, he will be consistently given even worse teammates and even harder enemies until the teams are balanced, a ~50% win rate is forced no matter what. I really don't understand why this is such a hard concept for you.



da_Rock002 #230 Posted Jan 14 2020 - 16:21

    Major

  • -Players-
  • 10179 battles
  • 3,946
  • Member since:
    11-24-2016

View PostNeatoMan, on Jan 12 2020 - 09:47, said:

This happens equally for everybody.    

 

 


So, you got a chart to back up that BS?   

You have no idea what battles are biased, do you.    So you're pulling the "equally for everyone" out of your.  

 

The rest of that post is just smoke.    Once again.  



da_Rock002 #231 Posted Jan 14 2020 - 16:24

    Major

  • -Players-
  • 10179 battles
  • 3,946
  • Member since:
    11-24-2016

View PostNeatoMan, on Jan 12 2020 - 09:47, said:

………..

 

What you always fail to realize is that everyone's wins/losses will be dependent on how much they contribute to their teams, and not the number of lopsided matchups. 

 

 


What you won't admit is that bit of smoke happens to apply to every single player of the 30 that wind up in each and every battle.

 

The wins come when your side of 15 players beat the other 15 players  who are also contributing as gud as they can. 



da_Rock002 #232 Posted Jan 14 2020 - 16:26

    Major

  • -Players-
  • 10179 battles
  • 3,946
  • Member since:
    11-24-2016
What I have always realized is there are 15 players on each team, 30 in the match.   And not a single one of them contributes more or less influence than they do.

jst2gr8 #233 Posted Jan 14 2020 - 16:40

    Captain

  • Players
  • 67949 battles
  • 1,423
  • [CRZY] CRZY
  • Member since:
    04-26-2011

View PostSimplyPzB2, on Jan 09 2020 - 21:26, said:

Nope.  Sbmm won't/can't/shouldn't be perfect.  It's best if it just provides a narrow competitive range.  Say 40-60% chance to win.  So better players will gravitate towards 60% and bad players will gravitate towards 40%.  Just like they do now.  The difference will be - no player will have to suffer through very unbalanced battles like the ones I posted at the start of this topic.

 

So your 40-60% will just keep players the same as they are and changing the current mm to sbmm wouldn't make a difference? Wrong or "nope" as you would say. The better players win rate would go down and the bad players would go up. And if it wouldn't change the win rates that much or at all then why change anything? SBMM is just an excuse for bad players to get a better win rate while they still don't learn the game. 

And as I stated, when you get into a platoon with 3 op E-35's and those three players are 1100 win8( can't remember your win 8's and truthfully don't care) it would not matter what the other team looks like. How do I know this? Because I have been in those same platoons. Also, your proof of unbalancing is in one game or two(again I could care less). Why aren't you posting the pics of the games when you were on the "lower skilled" team? It does go both ways. So once again, the steam rolls happen now, they happened back in the day and they would happen with SBMM. So there is no reason or good argument for SBMM except to help those with bad win rates to get better ones. 



NeatoMan #234 Posted Jan 14 2020 - 22:03

    Major

  • Players
  • 28785 battles
  • 21,272
  • Member since:
    06-28-2011

View Postda_Rock002, on Jan 14 2020 - 10:21, said:

So, you got a chart to back up that BS?   

You have no idea what battles are biased, do you.    So you're pulling the "equally for everyone" out of your. 

Don't need a chart.  it goes hand in hand with random.   It's as certain as "gravity exists, therefore the apple falls from the tree".

 

 If you say the MM is random (with respect to skill) then you must accept everything that entails.  That includes the fact that everyone gets as many teams mates/opponents stacked in their favor as everybody else.  The only difference from player to player will be due to the individual.

 

Either you have absolutely no clue what you are arguing about, or you don't believe a damn thing you say. 

 

View Postda_Rock002, on Jan 14 2020 - 10:24, said:

The wins come when your side of 15 players beat the other 15 players  who are also contributing as gud as they can. 

View Postda_Rock002, on Jan 14 2020 - 10:26, said:

What I have always realized is there are 15 players on each team, 30 in the match.   And not a single one of them contributes more or less influence than they do.

So you believe a player who consistently contributes nothing to his teams gives his teams just as much chance of winning as the guy who consistently contributes 2k to his teams?  They both play on 15 man teams, so which one of those players do you think will win more over their career, and why?

 

If what you say is true then how does one calculate win chance if individual contributions don't affect cumulative wins/losses?  Do you ever think any of this through before you hit reply?



Jaspo #235 Posted Jan 15 2020 - 02:59

    Staff sergeant

  • -Players-
  • 32093 battles
  • 495
  • [SLUGZ] SLUGZ
  • Member since:
    03-12-2015

View PostFlarvin, on Jan 06 2020 - 12:47, said:


The link below has a “Player distribution by WN8” graph, which is the only data I have seen about player skill distribution. 
 

Based on that graph, the server is not getting redder and redder. In fact, the opposite seems to be happening. Less bad and very bad players, with more average and above players. 

 

https://wot-life.com/na/serverstats/
 

 


There's some rather intriguing math going on with the wn8 distribution on that site. Apparently anything less than being in the top 37.38% of all players by global, not recent, stats is considered "below average"...so you have to be better than 62.62% of all players to be better than 50% of all players?






1 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users