Jump to content


Problems with the game from an experienced player(Please read wargaming)

maps gold Artilery Artillary Matchmaker

  • Please log in to reply
51 replies to this topic

Roboticus #21 Posted 04 September 2019 - 03:58 PM

    Lance-corporal

  • Player
  • 40345 battles
  • 68
  • [NOBAN] NOBAN
  • Member since:
    09-13-2011

View PostJauhesammutin, on 04 September 2019 - 02:24 PM, said:

 

This topic has been talked in numerous post over the years and the answers are still the same.

 

Short answer: no, it wouldn't be better.

Did you enjoy ranked battles? Do you know what happened to AW?

 

Does Rainbow Six have 15 v 15 players with ~80 individual "operators" you choose BEFORE the battle starts or is it just 5v5 and the MM counts only players ELO ratings? To be honest I don't even know why I bother replying to an obvious troll comment..


ranked wasn’t skill based MM. Yet another sheep that has no idea what he’s talking about. There’s one fact here. Skill based MM has not been tried, so you cannot say it wouldn’t work. With many more variables, it is besides the point, skill based MM of teams of players that have their wn8 synced and based off to establish teams with relatively similar wn8s collectively would result in far better matches than the way the MM is now. Unicums and all over 1600 wn8 on one team, the other team gets one good player and rest complete bots. Tell me that’s a good template, let me decide if you’re mad or just daft?



Jauhesammutin #22 Posted 05 September 2019 - 07:51 AM

    Second Lieutenant

  • Player
  • 23669 battles
  • 1,030
  • [FAUF] FAUF
  • Member since:
    11-05-2013

View PostRoboticus, on 04 September 2019 - 02:58 PM, said:


ranked wasn’t skill based MM. Yet another sheep that has no idea what he’s talking about. There’s one fact here. Skill based MM has not been tried, so you cannot say it wouldn’t work. With many more variables, it is besides the point, skill based MM of teams of players that have their wn8 synced and based off to establish teams with relatively similar wn8s collectively would result in far better matches than the way the MM is now. Unicums and all over 1600 wn8 on one team, the other team gets one good player and rest complete bots. Tell me that’s a good template, let me decide if you’re mad or just daft?

 

Yes I know, Ranked doesn't have SBMM. When did I say otherwise? Yet another sheep who can't read a simple forum post.

 

SBMM hasn't been tried so you can't say it would work. See, that works both ways.

 

See now you are talking about a completely different system. On your previous post you were comparing WoT to RSS in which players get matched against similar skill level players. Now you are talking about skill BALANCED matchmaking where there would be players from each skill group. Those are 2 different systems.

What's wrong with the current MM and what does SBMM fix in that?

 

Your system could be an interesting one except for couple of fundamental problems:

1. Rerolling/botting. I would instantly buy an account with 50k games and 45% winrate. I would get matched against another 45% player and guess who is going to win that one?

2. WN8. That's a 3rd party stat. Why on earth would WG balance their MM according to some random numbers some guy has created? 

3. If every 60% winrate player meets another 60% winrate player then their winrate is going to drift towards 50%. Eventually everyone is going to have a winrate of 50% and then MM doesn't have any clue about who is a good player and who isn't.

 

Just a tip for future replies: reply to the whole comment and not just a part of it (you completely ignored AW and RSS in my reply).

 

 


Edited by Jauhesammutin, 05 September 2019 - 07:54 AM.


4nt #23 Posted 05 September 2019 - 04:21 PM

    Lieutenant

  • Player
  • 29659 battles
  • 1,605
  • Member since:
    04-09-2013

View PostJauhesammutin, on 05 September 2019 - 07:51 AM, said:

 

Yes I know, Ranked doesn't have SBMM. When did I say otherwise? Yet another sheep who can't read a simple forum post.

 

SBMM hasn't been tried so you can't say it would work. See, that works both ways.

 

See now you are talking about a completely different system. On your previous post you were comparing WoT to RSS in which players get matched against similar skill level players. Now you are talking about skill BALANCED matchmaking where there would be players from each skill group. Those are 2 different systems.

What's wrong with the current MM and what does SBMM fix in that?

 

Your system could be an interesting one except for couple of fundamental problems:

1. Rerolling/botting. I would instantly buy an account with 50k games and 45% winrate. I would get matched against another 45% player and guess who is going to win that one?

2. WN8. That's a 3rd party stat. Why on earth would WG balance their MM according to some random numbers some guy has created? 

3. If every 60% winrate player meets another 60% winrate player then their winrate is going to drift towards 50%. Eventually everyone is going to have a winrate of 50% and then MM doesn't have any clue about who is a good player and who isn't.

 

Just a tip for future replies: reply to the whole comment and not just a part of it (you completely ignored AW and RSS in my reply).

 

 

Wonder why he ignored two sbmm using enviroments? Maybe because other failed epically and other isn't same format? 

 

The three points you presented are very ones Pro sbmm folk always ignore. Next up ridicule to both of us for not blindly believing him, probably a personal attacks or two, but 0 answers how to go about sbmm's cardinal problem- that it works in 15vs15 enviroment only for month or two.



shikaka9 #24 Posted 05 September 2019 - 04:40 PM

    Second Lieutenant

  • Player
  • 72185 battles
  • 1,314
  • [-SJA-] -SJA-
  • Member since:
    02-27-2013

finally from experienced player 

 

 

 

 

:trollface:



Jauhesammutin #25 Posted 06 September 2019 - 08:13 AM

    Second Lieutenant

  • Player
  • 23669 battles
  • 1,030
  • [FAUF] FAUF
  • Member since:
    11-05-2013

View Post4nt, on 05 September 2019 - 03:21 PM, said:

Wonder why he ignored two sbmm using enviroments? Maybe because other failed epically and other isn't same format? 

 

The three points you presented are very ones Pro sbmm folk always ignore. Next up ridicule to both of us for not blindly believing him, probably a personal attacks or two, but 0 answers how to go about sbmm's cardinal problem- that it works in 15vs15 enviroment only for month or two.

It might be that they think that SBMM would somehow create more close games (15-14) because the teams would be "balanced".

 

I don't think they understand the fundamental differences between WoT and competitive games like CS:GO, LoL, Dota, OW.

They don't understand that changes to SBMM would change the whole game. I'm not saying it couldn't work, but is it necessary? 



4nt #26 Posted 06 September 2019 - 09:59 AM

    Lieutenant

  • Player
  • 29659 battles
  • 1,605
  • Member since:
    04-09-2013

View PostJauhesammutin, on 06 September 2019 - 08:13 AM, said:

It might be that they think that SBMM would somehow create more close games (15-14) because the teams would be "balanced".

 

I don't think they understand the fundamental differences between WoT and competitive games like CS:GO, LoL, Dota, OW.

They don't understand that changes to SBMM would change the whole game. I'm not saying it couldn't work, but is it necessary? 

Balanced. In sbmm balance is not guaranteed, and in some cases can actually create even more imbalanced results. Two sides are equally skilled (on average or factually), both sides have equal tanks. Skill level isn't an issue since sbmm. Other side has, say, vk, 50-100 and pantera, other has maurerbrecher, somua and progetto. 

 

Now comes the question. How to balance the vk's 75% crew, no equipment on 100 and pantera player having rerolled (and failed)? Maurer player has 50% overalls like everyone, but 65% wr over 200 games on maurer? Prog player has bonded everything and 6 skill crew?

 

Nuances create pretty much disparity on sbmm as per intended by Pro sbmm posters. Not to mention that putting obj340u and something Else (as mirror obj is quite unlikely) on same level already is 2 to 5% advantage.

 

Edit: also how to introduce new players into the ladder? Movement cannot happen by only playing randoms, otherwise the system returns to current state, with differences in leagues being in name only. Essentially that feels like lots of work for getting still same dunces and uniqs within couple of months.


Edited by 4nt, 06 September 2019 - 10:03 AM.


Jauhesammutin #27 Posted 06 September 2019 - 10:06 AM

    Second Lieutenant

  • Player
  • 23669 battles
  • 1,030
  • [FAUF] FAUF
  • Member since:
    11-05-2013

View Post4nt, on 06 September 2019 - 08:59 AM, said:

Balanced. In sbmm balance is not guaranteed, and in some cases can actually create even more imbalanced results. Two sides are equally skilled (on average or factually), both sides have equal tanks. Skill level isn't an issue since sbmm. Other side has, say, vk, 50-100 and pantera, other has maurerbrecher, somua and progetto. 

 

Now comes the question. How to balance the vk's 75% crew, no equipment on 100 and pantera player having rerolled (and failed)? Maurer player has 50% overalls like everyone, but 65% wr over 200 games on maurer? Prog player has bonded everything and 6 skill crew?

 

Nuances create pretty much disparity on sbmm as per intended by Pro sbmm posters. Not to mention that putting obj340u and something Else (as mirror obj is quite unlikely) on same level already is 2 to 5% advantage.

 

Edit: also how to introduce new players into the ladder? Movement cannot happen by only playing randoms, otherwise the system returns to current state, with differences in leagues being in name only. Essentially that feels like lots of work for getting still same dunces and uniqs within couple of months.

To make it truly "balanced" it would mean that the teams are identical with tanks and there aren't any skills/equipment/bonds/crew/modules etc on tanks. Every tank would be identical. And even then we get the maps which aren't balanced. This would require at least 2 games played on with the same teams from both sides of the map..



4nt #28 Posted 06 September 2019 - 10:24 AM

    Lieutenant

  • Player
  • 29659 battles
  • 1,605
  • Member since:
    04-09-2013

View PostJauhesammutin, on 06 September 2019 - 10:06 AM, said:

To make it truly "balanced" it would mean that the teams are identical with tanks and there aren't any skills/equipment/bonds/crew/modules etc on tanks. Every tank would be identical. And even then we get the maps which aren't balanced. This would require at least 2 games played on with the same teams from both sides of the map..

And also RNG. Crucial shots not penning or lowrolling, and whoops whole flank is dead.

 

So. Kill RNG, dispersion, different tanks, all equipment, crew skills and training level, also one tier only, all maps must Be mirrored, and perferably just one map so no one needs to learn More Maps.

 

I think I prefer pong over that.



jabster #29 Posted 06 September 2019 - 10:42 AM

    Field Marshal

  • Beta Tester
  • 12771 battles
  • 26,284
  • [WSAT] WSAT
  • Member since:
    12-30-2010

View Post4nt, on 05 September 2019 - 03:21 PM, said:

Wonder why he ignored two sbmm using enviroments? Maybe because other failed epically and other isn't same format? 

 

The three points you presented are very ones Pro sbmm folk always ignore. Next up ridicule to both of us for not blindly believing him, probably a personal attacks or two, but 0 answers how to go about sbmm's cardinal problem- that it works in 15vs15 enviroment only for month or two.

 

Points 2., and 3., aren't particularly good as that just means you don't use win-rate as a balancing metric you use some other metric for evaluating 'skill'. The better point is that what does a skillbased MM achieve as for the analysis done so far there really isn't any strong link between unbalanced results and how balanced the teams were. This is compounded by teams in generally being more closely balanced than may be expected.

 

Personally I think the big flaw is the correlation between global stats and how a player performs in an individual battle is too weak.  To put it another way the average player is just too inconsistent to be able to balance teams effectively. I'd be all for some sort of skill being taken into account if I thought it would noticeable improve the overall game experience, the problem is I'm just not even slightly convinced that's true.



Jauhesammutin #30 Posted 06 September 2019 - 11:40 AM

    Second Lieutenant

  • Player
  • 23669 battles
  • 1,030
  • [FAUF] FAUF
  • Member since:
    11-05-2013

View Postjabster, on 06 September 2019 - 09:42 AM, said:

 

Points 2., and 3., aren't particularly good as that just means you don't use win-rate as a balancing metric you use some other metric for evaluating 'skill'. The better point is that what does a skillbased MM achieve as for the analysis done so far there really isn't any strong link between unbalanced results and how balanced the teams were. This is compounded by teams in generally being more closely balanced than may be expected.

 

Personally I think the big flaw is the correlation between global stats and how a player performs in an individual battle is too weak.  To put it another way the average player is just too inconsistent to be able to balance teams effectively. I'd be all for some sort of skill being taken into account if I thought it would noticeable improve the overall game experience, the problem is I'm just not even slightly convinced that's true.

 

You could use winrate as a balancing mechanic. Let's say WoT implemented an ELO system. If you have a 55% winrate in 1000-1100 ELO then you should move up to the 11000-12000 ELO bracket. Then your winrate would reset to 0%.

 

This means the winrate as a stat would be obsolete but in other games (CS:GO, OW, etc.) it already is. Players only care about rank.



jabster #31 Posted 06 September 2019 - 11:49 AM

    Field Marshal

  • Beta Tester
  • 12771 battles
  • 26,284
  • [WSAT] WSAT
  • Member since:
    12-30-2010

View PostJauhesammutin, on 06 September 2019 - 10:40 AM, said:

 

You could use winrate as a balancing mechanic. Let's say WoT implemented an ELO system. If you have a 55% winrate in 1000-1100 ELO then you should move up to the 11000-12000 ELO bracket. Then your winrate would reset to 0%.

 

This means the winrate as a stat would be obsolete but in other games (CS:GO, OW, etc.) it already is. Players only care about rank.


I was referring to overall team balance where players will tend towards 50%.



4nt #32 Posted 06 September 2019 - 11:53 AM

    Lieutenant

  • Player
  • 29659 battles
  • 1,605
  • Member since:
    04-09-2013

View Postjabster, on 06 September 2019 - 10:42 AM, said:

 

Points 2., and 3., aren't particularly good as that just means you don't use win-rate as a balancing metric you use some other metric for evaluating 'skill'. The better point is that what does a skillbased MM achieve as for the analysis done so far there really isn't any strong link between unbalanced results and how balanced the teams were. This is compounded by teams in generally being more closely balanced than may be expected.

 

Personally I think the big flaw is the correlation between global stats and how a player performs in an individual battle is too weak.  To put it another way the average player is just too inconsistent to be able to balance teams effectively. I'd be all for some sort of skill being taken into account if I thought it would noticeable improve the overall game experience, the problem is I'm just not even slightly convinced that's true.

Precisely. A ladder that makes player a part of certain position cannot rely on mere random battles' results, wether short or long term. There hasn't yet been any considerable choice: WR is diluted by Battle count, and battle count increases WGR steadily. Wn8 is 3rd party and flawed in other areas to adapt.

 

Dpg and avg exp are tier dependant, with avg exp inflated by premium acc. There isn't any certifiable skill indicator to create such ladder atm. Each in combination and average tier taken into account, with outliers in per tank performance could give rough indicator. But even so it'd place people playing, say, low tiers for fun and noobs into same position. And should sbmm note tier by tier the skill indicator? Player is on high ladder with t5 tanks, then plays tX and performance indicates he needs to be on low rung? 

 

It really isn't simple question, as some make it.



Jauhesammutin #33 Posted 06 September 2019 - 11:59 AM

    Second Lieutenant

  • Player
  • 23669 battles
  • 1,030
  • [FAUF] FAUF
  • Member since:
    11-05-2013

View Postjabster, on 06 September 2019 - 10:49 AM, said:


I was referring to overall team balance where players will tend towards 50%.

Isn't that exactly what I wrote and 4nt agreed with? Using only a winrate isn't an option since everyone will drift towards 50% (a bit less actually since draws exist).



jabster #34 Posted 06 September 2019 - 12:17 PM

    Field Marshal

  • Beta Tester
  • 12771 battles
  • 26,284
  • [WSAT] WSAT
  • Member since:
    12-30-2010

 View PostJauhesammutin, on 06 September 2019 - 10:59 AM, said:

Isn't that exactly what I wrote and 4nt agreed with? Using only a winrate isn't an option since everyone will drift towards 50% (a bit less actually since draws exist).

 

The point I was making was that win-rate will tend to 50% ism’t an argument against a skillbased MM. Maybe we are just at cross purposes here?


Edited by jabster, 06 September 2019 - 12:23 PM.


4nt #35 Posted 06 September 2019 - 12:31 PM

    Lieutenant

  • Player
  • 29659 battles
  • 1,605
  • Member since:
    04-09-2013

View Postjabster, on 06 September 2019 - 12:17 PM, said:

 

The point I was making was that win-rate will tend to 50% ism’t an argument against a skillbased MM. Maybe we are just at cross purposes here?

Most common form for sbmm porpose has been to use either wr or wn8 as basis to divvy up people. I believe we have dissected that argument for (as it is often used) sbmm in last few exchanges?



jabster #36 Posted 06 September 2019 - 12:35 PM

    Field Marshal

  • Beta Tester
  • 12771 battles
  • 26,284
  • [WSAT] WSAT
  • Member since:
    12-30-2010

View Post4nt, on 06 September 2019 - 11:31 AM, said:

Most common form for sbmm porpose has been to use either wr or wn8 as basis to divvy up people. I believe we have dissected that argument for (as it is often used) sbmm in last few exchanges?


The problem is pointing out that win-rate or WN8 aren’t metrics that can be used isn’t a good argument in itself. The only thing you show is that the poster who put it forward didn’t really think about what they’re saying not that the concept is inherently flawed.



4nt #37 Posted 06 September 2019 - 12:49 PM

    Lieutenant

  • Player
  • 29659 battles
  • 1,605
  • Member since:
    04-09-2013

View Postjabster, on 06 September 2019 - 12:35 PM, said:


The problem is pointing out that win-rate or WN8 aren’t metrics that can be used isn’t a good argument in itself. The only thing you show is that the poster who put it forward didn’t really think about what they’re saying not that the concept is inherently flawed.

But did I say it?

 

Thing is. Conceptually it could work- what I cannot see is way to implement it in a way that would achieve the goals set to it. Current measurements are not good enough, IMO. In that we may have differing view.



jabster #38 Posted 06 September 2019 - 01:00 PM

    Field Marshal

  • Beta Tester
  • 12771 battles
  • 26,284
  • [WSAT] WSAT
  • Member since:
    12-30-2010

View Post4nt, on 06 September 2019 - 11:49 AM, said:

But did I say it?

 

Thing is. Conceptually it could work- what I cannot see is way to implement it in a way that would achieve the goals set to it. Current measurements are not good enough, IMO. In that we may have differing view.


I’m not sure I ever stated you said it.

 

For the second part, I don’t see any reason why teams can’t be more balanced on performance it’s just I’m not sure what that will actually achieve beyond that.



4nt #39 Posted 06 September 2019 - 01:13 PM

    Lieutenant

  • Player
  • 29659 battles
  • 1,605
  • Member since:
    04-09-2013

View Postjabster, on 06 September 2019 - 01:00 PM, said:


I’m not sure I ever stated you said it.

 

For the second part, I don’t see any reason why teams can’t be more balanced on performance it’s just I’m not sure what that will actually achieve beyond that.

Okay. We're on same page on that then.



Jauhesammutin #40 Posted 06 September 2019 - 01:14 PM

    Second Lieutenant

  • Player
  • 23669 battles
  • 1,030
  • [FAUF] FAUF
  • Member since:
    11-05-2013

View Postjabster, on 06 September 2019 - 11:17 AM, said:

 

The point I was making was that win-rate will tend to 50% ism’t an argument against a skillbased MM. Maybe we are just at cross purposes here?

Well, yes and no. If we want to keep the current system where stats are tracked and winning gives a bonus to credits/XP then it's an argument against SBMM. But if we want to change the whole economy and stat system in the game then that argument isn't against SBMM. But if that happens is the game WoT?







Also tagged with maps, gold, Artilery, Artillary, Matchmaker

1 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users