Matchmaker Discussion Thread & MM Balance

[GarrusBrutus]

4 minutes ago, Greek_Patriot said:

People with no balls to play solo dont get to critise other who do.

So do you want me and my boyfriend to carry you to 70% winrate? If you pay me more than wargaming I might even consider your offer.

[Greek_Patriot]

Just now, GarrusBrutus said:

So do you want me and my boyfriend to carry you to 70% winrate? If you pay me more than wargaming I might even consider your offer.

No thanks. I prefer trying alone.

[Darkeid]

My humble apologies, but i think that it cant be possible answering like you represent WG. What we are seeing is ordinary players (and 4 non WG employees? Forum fanatics?) being online almost every work day in a single forum location/thread stating what can not & what can be improved. P.S. even the moderator has denied having any official role in the MM system or being a WG employee. About 30-40 pages ago a suggestion was offered to separate players by account winrate or PR , which after a partly offensive, partly internet bully behavior style - dialogue came to a conclusion that it cant happen because it would be unfair to the higher WR% players playing more difficult games. 

After all this negativity, ideally i would like for all of us to overcome the impolite & spiteful remarks and move on with the actual purpose of this thread. One suggestion could be to test this idea of adding an account WR parameter in the MM system for 1 weekend trial run and run one of the surveys after to see what players think. Increase rewards per battle for the Higher % of ppl and finally separate the player base by account WR and divide it into tiers (that the paid WG game developers will find fair) .

Another approach that possibly could improve this imbalance is to keep everything as is for people above 40% ? 45% ? . Keep people below 40% / 45% (potential griefers) only being matched up with new players. Raise the newbie limit to 500-800-1000? account battles total and at night times use bots. As you have already seen from the MM monitor pics , players below 200 battles are often put in regular randoms partly causing the imbalance discussed.

The real question is will any WG official actually read this? Or is this just a place that has to exist due to the size of the community. Peace. 

[ColonelPete]

9 minutes ago, Darkeid said:

About 30-40 pages ago a suggestion was offered to separate players by account winrate or PR , which after a partly offensive, partly internet bully behavior style - dialogue came to a conclusion that it cant happen because it would be unfair to the higher WR% players playing more difficult games.

No. It would just not work.

WR and PR would equalize over time. The 40% WR player would get closer to 50% and the 60% player would get closer to 50% until both will meet each other in battle, making the system useless and giving you the same battles as now.

12 minutes ago, Darkeid said:

Another approach that possibly could improve this imbalance is to keep everything as is for people above 40% ? 45% ? . Keep people below 40% / 45% (potential griefers) only being matched up with new players. Raise the newbie limit to 500-800-1000? account battles total and at night times use bots. As you have already seen from the MM monitor pics , players below 200 battles are often put in regular randoms partly causing the imbalance discussed.

Even now WG has to put bots in battles Tier I to V to fill ou tthe missing team slots. Increasing the newbie protection even further would make the situation worse.

 

The best option would be to prevent that people with 200 battles get into high Tier and make sure their end up in Tier V, Tier VI max, that they do not face too many experienced players and have time to learn before they reach high Tier.

[Greek_Patriot]

1 minute ago, ColonelPete said:

No. It would just not work.

WR and PR would equalize over time. The 40% WR player would get closer to 50% and the 60% player would get closer to 50% until both will meet each other in battle, making the system useless and giving you the same battles as now.

Even now WG has to put bots in battles Tier I to V to fill ou tthe missing team slots. Increasing the newbie protection even further would make the situation worse.

 

The best option would be to prevent that people with 200 battles get into high Tier and make sure their end up in Tier V, Tier VI max, that they do not face too many experienced players and have time to learn before they reach high Tier.

Bro it would be better to play with less people in each team rather playing with bots or people that are worse than bots.

[Triplexmazi]

Are you not going to matchmaking to the average rating of the player ?

The newcomer and the player with 5 years of experience enter the same match, this is not fair

 

I think these should be taken into consideration

-Battles

-Win rate

-average damage

-Experience

-Kills / deaths

 

this way there will be much better balanced and quality battles

 

(if not possible, there may be a separate game mode in which these will be active) 

 

It should also be not matched two-level difference

[SkipperCH]

Removed several posts and edited several.

 

This threads topic is a matchmaker discussion thread. There is no need or room for personal attacks against each other. So behave.

[ColonelPete]

36 minutes ago, GargencUA said:

Are you not going to matchmaking to the average rating of the player ?

The newcomer and the player with 5 years of experience enter the same match, this is not fair

That is why new players have newbie protection and veterans cannot do most missions at Tier I to IV.

[Darkeid]

17 hours ago, 159Hunter said:

That's random MM for you. You'll be matched up with good and bad players. 

And yes, it is an attitude problem. All we see in this topic is people complaining about how they lose games due to MM.

Never due to their own mistakes, or the mistake of another player. Or the great play by an enemy player. No, it's always the problem of MM.

That is an attitude problem.

You are only half correct, if I remember @Excavatus' numbers correctly: 4/10 are almost a guaranteed loss, 3/10 almost a guaranteed win.

So you have 3/10 games to influence yourself.

And if is like that; why is that a problemt? We can't all have 50% WR, that's just not possible as that would mean that everybody has the same skill.

 

All people complaining: I have yet to see one solution that would FIX the so called MM problem.

So far, the only fix I can come up with is learn people how to play. If everybody knew how to play (even just the basics) than we'd have a lot less landslide victories / losses.

 

This is not a valid argument. 2 groups of ppl have to carry X amount of weight to the finish line, 1 group gets 7 weaker members in terms of strength and your repeated opinion is that the remaining members of the team have an attitude problem when they complain that the opposing team has a clear strength advantage and that the outcome is predetermined. Sry i couldn't disagree more, that is not logical and above all not fair. 

@Excavatus can you please confirm that 7 out of 10 matches are a guaranteed win/loss and only 3 matches can be influenced by the players skill? 

-------

Another approach that possibly could improve this imbalance is to keep everything as is for people above 40% ? 45% ? . Keep people below 40% / 45% (potential griefers) only being matched up with new players. Raise the newbie limit to 500-800-1000? account battles total and use bots at any time that the playerbase is not sufficient for a full match or have smaller teams. As you have already seen from the MM monitor pics , players below 200 battles are often put in regular randoms partly causing the imbalance discussed.

The facts are that the existing MM system creates unfair matches. These matches should be worked out to prevent frustration / pointless loss of time. The system should be modified to create equal chances for both teams. That is basic for any multiplayer online game. It will not force unicums to 50% if implemented tiers / parameters are set correctly. Games will become challenging and more enjoyable. Personally , i find that a triple super unicum division will still excel if the player account WR selection is from 50% to 99%. Account WR parameters should be applied after the newbie tier. By following this example the best players will fairly have the best stats in the server. Increase the rewards (flags? credits? achievements? ship camos?) per tier.

Avoid unnecessary repetitions please. No one wants to have a fixed match or have someone else pull their weight for him. 

[ColonelPete]

3 hours ago, Darkeid said:

 It will not force unicums to 50% if implemented tiers / parameters are set correctly.

If an unicum can keep his 65% WR, then the teams are not equally strong.

3 hours ago, Darkeid said:

No one wants to have a fixed match or have someone else pull their weight for him. 

But that will happen when the 40% WR player will suddenly have a 50% chance of winning.

[Excavatus]

6 hours ago, Darkeid said:

can you please confirm that 7 out of 10 matches are a guaranteed win/loss and only 3 matches can be influenced by the players skill? 

cannot confirm or deny anything :) 

I made a statistical simulation to show that. you can have your own understanding about that, 

here is the link :) to the work @159Hunter mentioned. 

 

 

Sadly it is locked. 

 

10 hours ago, Greek_Patriot said:

Bro

I couldn't tag you because of the glorious soviet forum software :) 

I thought you may be interested in it as well. 

[DFens_666]

7 hours ago, Darkeid said:

This is not a valid argument. 2 groups of ppl have to carry X amount of weight to the finish line, 1 group gets 7 weaker members in terms of strength and your repeated opinion is that the remaining members of the team have an attitude problem when they complain that the opposing team has a clear strength advantage and that the outcome is predetermined. Sry i couldn't disagree more, that is not logical and above all not fair. 

 

Because a match being predetermined is only in your own head since you are using MMM.

One ranked game i should have lost according to you, but it wasnt. Another game i should have easily won, but it wasnt easy at all, infact it could have been lost very easily, despite 2x 65% players on my team.

One very common claim back in the RTS CV days was, that the worse CV player always loses, and i remember one game, where he had a 30% Saipan against a 60% Hiryu, yet we still won.

 

Lets take this example

 

Why do you complain about this? You only posted it and said "BALANS episode 3"

Whats wrong with that? And i dont want to hear "if you cant see it, you are *insert random insult*.

 

[ForlornSailor]

7 hours ago, Darkeid said:

@Excavatus can you please confirm that 7 out of 10 matches are a guaranteed win/loss and only 3 matches can be influenced by the players skill? 

 

When he talked about this, he didnt mean some sort of matchmaker algorythm, that is behind this. What he talked about, is something, that has to do with logic and has been gathered and discussed here in this forum for years.

 

The logic behind it has to do with the environment, that random matches are. 12 vs 12 (usualy) and you are 1 of the 12 on one team. Its impossible to be that good - or bad for the matter, that you could decide the outcome of each and every game (also, beause logic dictates, that even if you are a top 1% player, the other team will at some point also have a top 1 % player so you cancel each other out). thats where this 7 to 3 ratio comes from. And if you dig into the statistics, you can find confirmation for this. Players with a 60 % WR are already rare. They constantly influence 1 out of 10 games so much, that it turns from a predicted loss into a win.

 

And you can tune this force of impact up. With divisions, with the right ships and the correct line up of ships in a division. This matters a lot. First of all, you need to have a division where each player has the potential to impact the game. Second, they need to pick a ship, which synergizes with the other picks and usualy is of another class. There is a reason why divisions with 3 unicums usualy aint of the same class - unless they want to fool around or dont care. This way, you can bump up the Winrate into the regions of 85...90%. But there still stays a small amout of games, that is unwinable. Simply because in those games, the 9 other dudes on your team are so bad and / or you meet another super good division, so you cancel each other out and its again down to the 9 randoms on both teams.

 

 

[GarrusBrutus]

7 hours ago, Darkeid said:

This is not a valid argument. 2 groups of ppl have to carry X amount of weight to the finish line, 1 group gets 7 weaker members in terms of strength and your repeated opinion is that the remaining members of the team have an attitude problem when they complain that the opposing team has a clear strength advantage and that the outcome is predetermined. Sry i couldn't disagree more, that is not logical and above all not fair.

How is it not a valid argument? If what you say is the case in EVERY match, then why would a team even try to play against TWA in KOTS? All of their players are super unicums.... So statistically they should beat every opponent. 

Why should Fulham play against Manchester City? Every player of MC is better skilled than the ones of Fulham... 

Why should a lower placed tennis player play against Nadal? He will not beat him anyway.... 

Etc. Etc. 

 

It IS an attitude problem when you assume you're gonna lose when the odds are stacked against you. 

In wows, knowing and depending on those odds is a mortal sin and self fulfilling prophecy. 

 

If you watch your precious MMM like it's a future prediction machine, you will always get the outcome it predicts. "Oh my team has more bad players, this will surely be a loss.". 

 

Stop. 

Using. 

MMM. 

[ForlornSailor]

16 minutes ago, GarrusBrutus said:

How is it not a valid argument? If what you say is the case in EVERY match, then why would a team even try to play against TWA in KOTS? All of their players are super unicums.... So statistically they should beat every opponent. 

Why should Fulham play against Manchester City? Every player of MC is better skilled than the ones of Fulham... 

Why should a lower placed tennis player play against Nadal? He will not beat him anyway.... 

Etc. Etc. 

 

It IS an attitude problem when you assume you're gonna lose when the odds are stacked against you. 

In wows, knowing and depending on those odds is a mortal sin and self fulfilling prophecy. 

 

If you watch your precious MMM like it's a future prediction machine, you will always get the outcome it predicts. "Oh my team has more bad players, this will surely be a loss.". 

 

Stop. 

Using. 

MMM. 

we could also ask @Darkeid this: does he think, the matchmaking monitor looks different for players like you and me? If we take ourselfs out of the picture, the odds are stacked against us as often as they are in our favour, looking at the stats of players. We could basically fill the same complaints with some fancy screenshots, since we get the same stuff all the time.

[SkipperCH]

Removed 1 post.

[Semmleri]

How does random battle matchmaker consider balancing teams according to skill?

 

My average WR was 57% and I only have 1340 games with most games having 70-90k damage, and 1-3 kills at mostly 5,6,7,8 tiers. I have lost literally 60+% of games over the last week. I keep getting team after team after team of people who know nothing about the game and manage to lose it in the first 8 minutes (Broadsiding so no angling, not using correct ammo type, no map avareness, no objective or teamplay, abandoning one side of the map, not even managing to get kills etc.) .. So every game is basically a 5v10 after considering the lost ships after a few minutes of play.

 

This is getting beyond annoying, I mean how is this possible? I am not even that good of a player, but come on please give me at least an average team to have chance to win.. these over 10 game loss streak make me question why I even bother to play this game.

Can anybody provide me with any info regarding this kind of imbalance and the matchmaker being broken?

 

[ColonelPete]

24 minutes ago, Semmleri said:

How does random battle matchmaker consider balancing teams according to skill?

It does not. Which you could have known if you had read one page of the discussion.

 

Btw. I suggest a reality check about your performance. You are far from doing mostly 70-90k damage and 1-3 kills.

I also suspect that you are playing on higher Tier than before.

[Semmleri]

1 minute ago, ColonelPete said:

Btw. I suggest a reality check about your performance. You are far from doing mostly 70-90k damage and 1-3 kills.

I have done this consistently over the last week (not talking about long term). Mostly playing Baltimore, New Mexico, Prinz Eugen, Tirpitz B, Nagato etc. (of course I get those 50-60k damage games too with no kills, although It is not my fault entirely because I get 1v3 because team crumbles)

 

What I mean is the horrible state of imbalance in the teams that matchmaking is generating. I know that I am just an average player, I too make mistakes (mostly eat torps). But the amount of piss poor teams I have gotten is truly amazing.

Why there is no balancing? It would make this game so much better. I just don't understand why players have to suffer these constant steamrolls and stomps, with slim to none chance of winning even if you get a kraken..

[Triplexmazi]

19 hours ago, ColonelPete said:

That is why new players have newbie protection and veterans cannot do most missions at Tier I to IV.

new players can buy t8 premium ships. They appear in t9 and t10 matches.. Players who have not experienced firing have t8, t9, t10 ships

 

After that, there is no team in the first 5 minutes 

 

 

[ColonelPete]

3 minutes ago, Semmleri said:

I have done this consistently over the last week (not talking about long term).

Doubt it or very small number of games.

4 minutes ago, Semmleri said:

What I mean is the horrible state of imbalance in the teams that matchmaking is generating. I know that I am just an average player, I too make mistakes (mostly eat torps). But the amount of piss poor teams I have gotten is truly amazing.

It is a sign that you are learning. You are now spotting all the mistakes your teammates do, which you did not spot before. Players are like that for over half a year now and were not much better before.

6 minutes ago, GargencUA said:

new players can buy t8 premium ships. They appear in t9 and t10 matches.. Players who have not experienced firing have t8, t9, t10 ships

 

After that, there is no team in the first 5 minutes 

But not all new players do that.

And I encountered teams that lasted for 20 minutes.

 

And translating my post does not make is very easy for me to understand what you are refering to...

[Darkeid]

In the meantime this is what happens

[159Hunter]

@Darkeid , just stop. You've made your point. 

If you want to keep posting pictures of MM monitor, then have the intellectual honesty to post :

- screenshots from ALL your games;

- add the results from all those games.

Right now, all you do is clutter the forum with screenshots that add no value to this discussion.

[DFens_666]

3 minutes ago, 159Hunter said:

@Darkeid , just stop. You've made your point. 

 

I also like, how he wants an open discussion, and when i ask him a normal question - nothing happens

[159Hunter]

20 hours ago, GargencUA said:

Are you not going to matchmaking to the average rating of the player ?

The newcomer and the player with 5 years of experience enter the same match, this is not fair

So then I'll just make a new account to go club seals?

Quote

I think these should be taken into consideration

-Battles

-Win rate

-average damage

-Experience

-Kills / deaths

So you want these values to be taken as a( and please explain why ):

- global account value;

- account value over the past month;

- shipclass value;

- shipclass value over the past month;

- ship value;

- ship value over the past month;

- tier based value;

- tier based value over the past month;

- grouped tier value;

- grouped tier value over the past month;

- division value;

- solo value;

- games played with flags and camo;

- games played without flags and/or camo;

...

 

Month / 2 weeks / 2 months / current patch ....