Jump to content


Skill Based Balancing (Not SBMM) Example


  • Please log in to reply
23 replies to this topic

Jhuderis #1 Posted Jul 12 2019 - 22:03

    Sergeant

  • -Players-
  • 5642 battles
  • 161
  • [WCTNT] WCTNT
  • Member since:
    05-15-2018

These teams are already chosen by the existing matchmaking system. Swapping team members of the same/similar tank type from one side to the other would improve balance of matches imho. 

 

 

Averaged Win Rate (added all W/R and divided by 15)

Left Team - 49.73 W/R

Right Team - 54.6 W/R

 

There is no argument I can come up with that doesn't say the "Right Team" has a huge statistical advantage to win this game. We all know the difference in player skill to be a 54% player vs. a 49% player, let alone the 2 super unicums and 3 unicums on the "Right Team". They could probably all individually carry games, let alone 5 of them on one team.

 


With the crudely "balanced" version I did visually in 5 mins, the rates stack up as follows:

Left Team - 52.46 W/R

Right Team - 51.06 W/R

I did this my "feel" just looking and swapping, one time. I was able to balance it within 1.5% vs. the original balance of 5%+. A computer algorithm would kick my butt at making it way closer, way faster. 

Anywhoo, feel free to shred me. I just wanted to actually lay out an example of skill balancing, since most of the discussion/rejection of a new system is around SBMM not balancing after the matchmaking. 

Edit: The 200IQ guy you can't see has a 59% W/R. 


Edited by Jhuderis, Jul 12 2019 - 22:05.


Gothraul #2 Posted Jul 12 2019 - 22:22

    Major

  • -Players-
  • 2426 battles
  • 5,720
  • Member since:
    11-17-2014
Last thing any team needs is 45% taters especially when they are your top tiers in a 3-5-7 hell.

Boghie #3 Posted Jul 12 2019 - 22:30

    Captain

  • -Players-
  • 8482 battles
  • 1,142
  • [WCTNT] WCTNT
  • Member since:
    04-10-2016

You have a platoon issue.

 

The only player you could move would be the IS-3A.



JakeTheMystic #4 Posted Jul 12 2019 - 22:46

    Major

  • Players
  • 24843 battles
  • 2,658
  • [OPIC] OPIC
  • Member since:
    12-30-2011

firstly, showing their names could result in you getting a strike for name and shame.

 

Secondly, WG doesnt accept wn8 as a measurement of player skill, at most they would use PR. The problem with PR is that its solely based on number of battles + average damage, you could essentially play a crap ton of tier 10 games and get a fairly decent PR, but it wouldnt represent anything conclusive of how the player actually performs. 

 

Just remove platoons from the game all together. Tomatoes that platoon with other tomatoes are the ones throwing games, not because of the MM itself. 



Ndtm #5 Posted Jul 12 2019 - 23:02

    Major

  • Players
  • 17694 battles
  • 4,598
  • Member since:
    05-01-2012

View PostJakeTheMystic, on Jul 12 2019 - 22:46, said:

firstly, showing their names could result in you getting a strike for name and shame.

Eh.. in this case i don't think that applies, i mean, it's in the name "name and shame", the op is talking in a general type way and not really talking down on a certain group or individuals either



Jhuderis #6 Posted Jul 12 2019 - 23:21

    Sergeant

  • -Players-
  • 5642 battles
  • 161
  • [WCTNT] WCTNT
  • Member since:
    05-15-2018

Oh I didn't know there was a problem with showing names and certainly wasn't intending anything negative about any particular player. Hope the post can stand for further discussion. If not, I can remove it. 

 

Also, they could just rules lawyer platoons and say "well, you're still in the same game together, just on opposite sides for balance sake" haha. ;)


Edited by Jhuderis, Jul 12 2019 - 23:23.


Nixeldon #7 Posted Jul 12 2019 - 23:23

    Major

  • Players
  • 60879 battles
  • 2,207
  • Member since:
    10-30-2011

View PostJakeTheMystic, on Jul 12 2019 - 16:46, said:

firstly, showing their names could result in you getting a strike for name and shame.

 

Secondly, WG doesnt accept wn8 as a measurement of player skill, at most they would use PR. The problem with PR is that its solely based on number of battles + average damage, you could essentially play a crap ton of tier 10 games and get a fairly decent PR, but it wouldnt represent anything conclusive of how the player actually performs. 

 

Just remove platoons from the game all together. Tomatoes that platoon with other tomatoes are the ones throwing games, not because of the MM itself. 

That is not how PR is calculated.



Boghie #8 Posted Jul 13 2019 - 00:39

    Captain

  • -Players-
  • 8482 battles
  • 1,142
  • [WCTNT] WCTNT
  • Member since:
    04-10-2016

On the edge of 'name and shame', but to make an important point:  Look at the record of the 'Super Unicum' in the IS-3A.  He might be - or can become - a Super Unicum, but is he one right now?

The other is likely a re-roll, but he is platooned with the Skorpian G and they should be moved as a unit.

The SU-130PM player has played a total of 18 games in TDs.

The Skorpian is a re-roll.

And the GW Tiger player has to be moved as a unit with the SU-130PM.

 

Are the re-roll stats as good as they appear?  I would say yes, but it is not a certain thing.  Is a guy with 18 games in an entire tank class as good as his stats in another class?  Is a low tier seal clubber 'Super Unicum' the equivalent of a high tier Super Unicum (by the way, he/she is quite good at high tier as well so maybe weight the MM by game count)?  Is is smart or fair to break up platoons?  Lastly, what is one Unicum worth?  Maybe moving two/three 'Green' players could work as well.

 

Oh, but now it gets a bit more complicated.

 

All in all, all this worry about individual games being kinda lopsided is kinda silly.  Most games are fine.



cavalry11 #9 Posted Jul 13 2019 - 00:52

    First lieutenant

  • Players
  • 49593 battles
  • 626
  • [11BAT] 11BAT
  • Member since:
    06-24-2013
No they arent

Omega_Weapon #10 Posted Jul 13 2019 - 01:38

    Major

  • Players
  • 53120 battles
  • 2,648
  • [GRIEF] GRIEF
  • Member since:
    11-15-2011

View PostJhuderis, on Jul 12 2019 - 16:03, said:

Anywhoo, feel free to shred me. I just wanted to actually lay out an example of skill balancing, since most of the discussion/rejection of a new system is around SBMM not balancing after the matchmaking.


Your tinkering may have evened those 2 teams out skill wise, but it also takes the random element out of random match making. Skill based match making leads to all individuals winning about 50% of the time because more successful players are immediately punished by having poor players added to their team to "balance" them. I really don't see what is fair about that.



RingoMcHarrison #11 Posted Jul 13 2019 - 16:14

    Major

  • Beta Testers
  • 36226 battles
  • 5,272
  • Member since:
    11-22-2010

Block Quote

Skill Based Balancing (Not SBMM) Example

 

 

So you're balancing skill (by swapping players) but it's not skill balanced?

 

Nice mental gymnastics.



dunniteowl #12 Posted Jul 13 2019 - 19:37

    Major

  • -Players-
  • 31291 battles
  • 8,047
  • Member since:
    09-01-2014

If there were money or standings (actual, honest to God, standings) that made a difference in one's outcomes of games, then Skill Balanced, Skill Based or, hell, even Skill Enforced match making might make sense.

 

No-one's life is going to be negatively affected by ANY of the outcomes in Random Public Matches based on inherent skill imbalances in the Random Match Maker we currently have.

 

I will state, ONE MORE TIME (though probably not the last time) that Skill Balancing, Skill Based, Skill ANYTHING type of match making DOES NOT BELONG IN A NON-COMPETITIVE ENVIRONMENT.

 

Yes, we, as players compete to win.  In the same way we compete to win in a sandlot baseball game at the corner lot or joining a Saturday Party game of football where more beer is spilled then blood.  We play to win, but there's nothing but the spirit of competition and the desire to win that drives us and our outcomes are tempered by the fact that we had fun, it wasn't important for anything other than that and we all had a good time, win or lose.

 

You cannot remotely compare this game's 'competiing' players as 'competitive' in the sense that the reason they compete is because it is their job or profession as members of their sport/league/division.  Skill Balancing, Skill Based, Skill Enforced Match Making are all intended to separate TEAMS of players by their level of overall ability in order to fairly compete in a game that has something more than just a trophy riding on the outcome.  It is for PROFESSIONAL LEVEL COMPETITION or, failing that, for League Level Athletics/Sports/Events that require folks to train together as a team and play against other teams that do the same thing.

 

Clan Wars?  Yeah, that's a good candidate for Skill Balancing.  League Play?  Yeah, that's a good candidate for Skill Matching.  Random Pubs?  Nope.  Sorry.  Cant help ya.  The team you are now on and the team you are playing, will never exist again.  Well, maybe not Never, though the odds are super highly against it.  As in Universally Astronomically Unlikely.

 

 

Everyone that promotes this idea thinks that, by doing this, the games will be more 'fair' and more 'fun' and there simply isn't any truth to that if you understand what the PURPOSE actually is for Skill matching.  It's designed for professional or truly competitive league level playing.  Bowling teams. Baseball teams.  Soccer teams.  Football teams.  Hockey and Basketball, any professional team level sport that also has scaled leagues for playing as amateurs -- all these are candidates for Skill Based or Skill Balanced Match making, because they require them to be properly competitive where the outcomes have more than 'just for fun' implied to them.

 

Yes, it sucks getting your backside handed you repeatedly.  Hell, I've had a few 28% and 34% days, playing many matches and just going, "Hmm.  Well, I'm not doing that well today.  Is it just me?"  You know, if it hurts to lose, then get used to losing.  This game will certainly get you used to that.  Even a 65% player still loses 35% of their games, which is pretty close to 1 loss every 3 games.

 

Get used to losing.  Learn to enjoy the challenge and accept that many times you lose or get exited, because someone else was:

 

     A) Better Than You

     B) Luckier Than You.

     C) Faster Than You.

     D) More Powerful Than You.

               or

     E) More Clever Than You.

 

And how can you be upset at someone else for YOU losing out?  You lost the exchange.  Maybe you also lost the game.  It sucks to lose.

 

 

There's another game 'right around the corner' from you.  Simply pick a tank, take a breath and click "Battle!"

 

 

Losing sucks and I KNOW, WITHOUT A DOUBT, that the majority of these threads that keep asking for all this extra work to be placed into the match maker are losing more than they like and think, that by proposing this, they aren't going to lose as much.  Unless you suddenly get much better at playing?  Not going to change with a new MM.  And, please remember, Skill Based/Balancing/Enforced MM is not for you and it won't solve the issue that makes you propound its use.  It's for folks who Have To Compete as part of their obligation of playing.

 

People, I am just at 50% as a player.  I lose EVERY OTHER GAME!  Folks who are lower than that lose more than that and that sucks, too.  Sadly, this game is going to provide you plenty of practice at taking it with some manner of dignity at whatever level you can manage.

 

I know it sounds trite, but you know what makes it okay for me when I'm having a 'bad day' and losing way more than I like?  I take the time to tell myself this one thing:

 

It's just a game.  it's just for fun.  It doesn't mean anything.  

 

 

 

GL, HF & HSYBF!
OvO

 

 

 

     


Edited by dunniteowl, Jul 13 2019 - 19:40.


Heavylemon5454 #13 Posted Jul 13 2019 - 20:27

    Corporal

  • Players
  • 11731 battles
  • 63
  • [P-W] P-W
  • Member since:
    09-18-2012

View Postdunniteowl, on Jul 13 2019 - 10:37, said:

 

I will state, ONE MORE TIME (though probably not the last time) that Skill Balancing, Skill Based, Skill ANYTHING type of match making DOES NOT BELONG IN A NON-COMPETITIVE ENVIRONMENT.

 

Boom, you are exactly on point here. You hit the nail through the board here.

It's a random battle. It's not competitive. It's similar to jumping into a non-competitive match in CS:GO. It's similar to playing a casual match in TF2. It's like being divided into teams for a casual game of dodgeball. 

 

If you want a more balanced game with more balanced teams, play the ranked matches. Dabble in the clan wars. Join a professional dodgeball team. Do note that balance is not entirely able to prevent you or the enemy team from being roflstomped by the other respective team.



Roggg2 #14 Posted Jul 15 2019 - 19:17

    Major

  • -Players-
  • 21963 battles
  • 2,143
  • [D-DAY] D-DAY
  • Member since:
    05-27-2015
The question is not if it can be done (easily, quickly, or otherwise).  The question is should it be done.  And I think no.  Lucky for me, WG agrees with that so far.  I want random so I have a chance to influence my win rate.

Jhuderis #15 Posted Jul 16 2019 - 19:58

    Sergeant

  • -Players-
  • 5642 battles
  • 161
  • [WCTNT] WCTNT
  • Member since:
    05-15-2018

View PostRoggg2, on Jul 15 2019 - 10:17, said:

The question is not if it can be done (easily, quickly, or otherwise).  The question is should it be done.  And I think no.  Lucky for me, WG agrees with that so far.  I want random so I have a chance to influence my win rate.

 

Curious why you feel you'd influence the game less if the teams were better balanced. Wouldn't your input to the situation be the same regardless? Honest question. 



Jhuderis #16 Posted Jul 16 2019 - 20:02

    Sergeant

  • -Players-
  • 5642 battles
  • 161
  • [WCTNT] WCTNT
  • Member since:
    05-15-2018

View Postdunniteowl, on Jul 13 2019 - 10:37, said:

Losing sucks and I KNOW, WITHOUT A DOUBT, that the majority of these threads that keep asking for all this extra work to be placed into the match maker are losing more than they like and think, that by proposing this, they aren't going to lose as much.  Unless you suddenly get much better at playing?  Not going to change with a new MM.  And, please remember, Skill Based/Balancing/Enforced MM is not for you and it won't solve the issue that makes you propound its use.  It's for folks who Have To Compete as part of their obligation of playing.   

 

For me, it's about making games more fun by reducing the chances for lopsided blowouts. Those aren't fun in either direction imho. A close loss is better to me than a rofl stomp victory, even if I'm on the winning side. 

 

I realize that I, as a 49.8% player am losing slightly more than I win. I can change that by getting better. While I get better, I want the games to be more balanced because they become nail-biters, clutch situations etc.

 

Again, I'm not advocating one way or the other, just curious about the in-match balance vs. skill based matchmaking and whether that might make for improved gameplay overall. 



ethics_gradient #17 Posted Jul 16 2019 - 20:27

    Major

  • Players
  • 41374 battles
  • 2,354
  • [DHO-X] DHO-X
  • Member since:
    04-22-2011

The lopsided battles are caused by the auto-loaders and to a lesser extent, the ultra high alpha guns.  Now a mistake is punished by death, and the loss of a tank can start a cascade effect.

 

More importantly, the conceptual problem with what you propose is that it reduces the incentive for bad players to become good.  After all, they can always rely on a steady stream of good players to support them as they know that MM is no longer random.  Even better, a player can have his bot "play" for 9 battles, and then play the 10 th battle himself, and enjoy having a bunch of good players carrying him.  No thanks.

The solution is not changing MM.  The solution is to have a better player base and fewer players who are absolute death for their teams.  I mean how does a player manage to "achieve" a 40% win rate?


Edited by ethics_gradient, Jul 16 2019 - 20:28.


Omega_Weapon #18 Posted Jul 17 2019 - 01:32

    Major

  • Players
  • 53120 battles
  • 2,648
  • [GRIEF] GRIEF
  • Member since:
    11-15-2011

View PostJhuderis, on Jul 16 2019 - 13:58, said:

Curious why you feel you'd influence the game less if the teams were better balanced. Wouldn't your input to the situation be the same regardless? Honest question. 

 

If you are a good player and can kill say 4 enemies in a given battle, that normally helps your team's chance to win significantly. That is why win rates are higher for some players. Under SBMM however, the player who kills 4 enemy tanks has their extra contribution neutralized by an opposing player with matching skill who will also likely kill 4 tanks. Skill based match making removes skill as a deciding factor right from the get go so matches will come down to RNG and dumb luck almost every time. That results in everyone (good and bad players) ending up with a 50% win rate over time.



ethics_gradient #19 Posted Jul 17 2019 - 03:08

    Major

  • Players
  • 41374 battles
  • 2,354
  • [DHO-X] DHO-X
  • Member since:
    04-22-2011
For example, I just got out of a battle.  One guy on my team had 49k battles, so he's very experienced. Problem, he's a [edited]on purpose.  He was playing a leopard (tier V), which he took to a corner of the map and just parked.  No scouting, no movement, nothing.  He has a 44% win rate and a 234 WN8.  He literally is a leech.  Now tell me, how does a guy like that deserve to be teamed-up with good players to make the teams even?  If it were up to me, he'd only be teamed with other jerks and bots, because they all deserve each other.  Under the OP's system, he would be fed a steady diet of unicums to make for his lack of care or effort.  49k battles of wasted time, and he wasted the time of everyone on his team.  yeah, I'm pissed.

Trakks #20 Posted Jul 17 2019 - 08:15

    First lieutenant

  • Players
  • 3801 battles
  • 760
  • Member since:
    11-24-2011

View PostNdtm, on Jul 12 2019 - 22:02, said:

Eh.. in this case i don't think that applies, i mean, it's in the name "name and shame", the op is talking in a general type way and not really talking down on a certain group or individuals either

He is highlighting the poor performance of these players and leaving their names intact. It is name and shame regardless of whether or not he is deliberately trying to be personally insulting to those guys. 






1 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users