Jump to content
You need to play a total of 5 battles to post in this section.
CheapWine

MM methodology

34 comments in this topic

Recommended Posts

21
[B-K-N]
Members
30 posts
8,126 battles

I have seen posts regarding the imbalance of teams but I have not seen a post that indicates if WG has ever explained how it works. If this has already been addressed I apologize. I have been in a battle where the other team had an aggregate 7 tiers higher ships and we had the same number of each type of ship. There also seems to be no accounting for the number of premium ships per team.  Tonight I was in a battle were the other team had a 7 to 1 advantage in premium ships. While these are the most extreme cases I have experienced I struggle to understand why such imbalances are ever allowed. Sooooo have they ever bothered to explain how the MM works?

Share this post


Link to post
Share on other sites
659
[5D]
Members
2,215 posts
13,978 battles

Yes, MM accounts for the tiers and class (BB, DD, etc.). It does not look at ship type, premiums or anything else.

 

Imbalance is a player issue due to not working together. Once things go wrong for one side it seems to quickly multiply against them.

Share this post


Link to post
Share on other sites
2,764
Members
9,864 posts
5 minutes ago, 1SneakyDevil said:

Yes, MM accounts for the tiers and class (BB, DD, etc.). It does not look at ship type, premiums or anything else.

 

Imbalance is a player issue due to not working together. Once things go wrong for one side it seems to quickly multiply against them.

 

Yep, always the players fault that the MM is bad....:Smile_facepalm:

  • Cool 3

Share this post


Link to post
Share on other sites
1,418
[HINON]
Supertester
7,520 posts
7,578 battles
6 minutes ago, 1SneakyDevil said:

Yes, MM accounts for the tiers and class (BB, DD, etc.). It does not look at ship type, premiums or anything else.

:Smile_great: This.

Share this post


Link to post
Share on other sites
423
[1IF]
Alpha Tester
4,252 posts
8,226 battles
1 minute ago, awiggin said:

 

Yep, always the players fault that the MM is bad....:Smile_facepalm:

Yep, that's what the WG Zombies say .... :Smile_great:

Share this post


Link to post
Share on other sites
2,016
[PVE]
Members
5,315 posts
17,798 battles

There are plenty of posts with screenshots of MM that is pretty sucky. Multiple radar ships on one team, none on other just for instance. MM only takes account of class and tier, capabilities of ships and captains are not included.  IMHO matchmaking is a real weak spot in the game that leads to player frustration. Frustrated players don't spend money or continue to play. 

  • Cool 2

Share this post


Link to post
Share on other sites
1,493
[GREPT]
[GREPT]
Beta Testers
6,739 posts
7,090 battles
4 minutes ago, awiggin said:

 

Yep, always the players fault that the MM is bad....:Smile_facepalm:

Occam's razor, yes it is the players fault, a bad MM setup can be won still if the players are smart and work together. The fact it doesn't so often is because the playerbase as a whole is quite poor. 

Share this post


Link to post
Share on other sites
Members
3,554 posts
6,867 battles
9 minutes ago, 1SneakyDevil said:

Imbalance is a player issue due to not working together.

Yes and I will make my usual point that MM could benefit in matching up teams based upon individual ship WR (NOT overall WR). The imbalance comes from this and frankly while losing because you are put onto a clearly doomed potato team is not that much better than winning because your opposition was pathetic.

  • Cool 1

Share this post


Link to post
Share on other sites
2,016
[PVE]
Members
5,315 posts
17,798 battles
Just now, Raptor_alcor said:

Occam's razor, yes it is the players fault, a bad MM setup can be won still if the players are smart and work together. The fact it doesn't so often is because the playerbase as a whole is quite poor. 

Yeah that is it <insert eye roll here>. If we quit losing veteran players maybe player capabilities would improve. If MM was better maybe we wouldn't bleed players like a severed carotid artery.... 

  • Cool 1

Share this post


Link to post
Share on other sites
1,493
[GREPT]
[GREPT]
Beta Testers
6,739 posts
7,090 battles
Just now, Taylor3006 said:

Yeah that is it <insert eye roll here>. If we quit losing veteran players maybe player capabilities would improve. If MM was better maybe we wouldn't bleed players like a severed carotid artery.... 

Adjust or die, I strongly suggest you leave at this rate since you are clearly more fit to complain than to figure out how to adjust to adversity. This isn't farmville kid, you won't always win but you have to actually give it a shot rather than give up on the moral front immediately. 

Share this post


Link to post
Share on other sites
2,016
[PVE]
Members
5,315 posts
17,798 battles
24 minutes ago, Raptor_alcor said:

Adjust or die, I strongly suggest you leave at this rate since you are clearly more fit to complain than to figure out how to adjust to adversity. This isn't farmville kid, you won't always win but you have to actually give it a shot rather than give up on the moral front immediately. 

Not complaining at all, I am pretty indifferent about it honestly. Just stating my opinion. Could care less if Wargaming closes up shop tomorrow. BTW kiss my furry buttocks you nasty butt sniffing twerp. I suppose I can remember to insult as well. 

Edited by Taylor3006
  • Cool 1

Share this post


Link to post
Share on other sites
7,214
[NMKJT]
Beta Testers
21,196 posts
3,871 battles
23 minutes ago, Raptor_alcor said:

Adjust or die, I strongly suggest you leave at this rate since you are clearly more fit to complain than to figure out how to adjust to adversity. This isn't farmville kid, you won't always win but you have to actually give it a shot rather than give up on the moral front immediately. 

There is so much wrong with this post I don't even know where to begin.

 

Assumptions. Insults. A lot of comments tailored in a way as to make yourself seem superior to someone else entirely due to their opinion differing from yours. Holy smokes, and you have the nerve to critique the way I say things?

 

You wanna say I can't handle a "hard" game, go play Castlevania on the NES and get back to me. I played that thing every day and adore it, when most of the millennials you're leaning on the stereotype of couldn't get past the first level without quitting. Don't sit there and spout opinions as fact, and especially don't make assumptions about what kind of game people are looking for.


Warships has a lot of obtuse and frustrating mechanics and design decisions that don't have to be obtuse of frustrating. They just are. And people defend their right to be obtuse and frustrating for no reason other than because they can.

Edited by AraAragami
  • Cool 2

Share this post


Link to post
Share on other sites
2,764
Members
9,864 posts
41 minutes ago, Raptor_alcor said:

Occam's razor, yes it is the players fault, a bad MM setup can be won still if the players are smart and work together. The fact it doesn't so often is because the playerbase as a whole is quite poor. 

 

Yes, and since everyone knows who the player base is...why wouldn't a company recognize it and adjust? :cap_hmm:

 

Share this post


Link to post
Share on other sites
[SYN]
Members
509 posts
5,971 battles
2 hours ago, Stauffenberg44 said:

Yes and I will make my usual point that MM could benefit in matching up teams based upon individual ship WR (NOT overall WR). The imbalance comes from this and frankly while losing because you are put onto a clearly doomed potato team is not that much better than winning because your opposition was pathetic.

If the teams were balanced by skill, half the higher skilled players lose every time due to lower skilled teammates. What does it solve?

Share this post


Link to post
Share on other sites
21
[B-K-N]
Members
30 posts
8,126 battles

If MM accounts for ship tier and class then why are their battles where one team has a decided advantage in tiers?

Share this post


Link to post
Share on other sites
Members
2,409 posts
5,422 battles

It's too bad that WOW does not consider each ship with its custom rating depending on the build. options, Captains skills, flags and camo as a rating much like they rate the ships when you click on them but those are naked with no customization.  Then theres the Player..rate him for the average Win Rate. Add those two numbers together and get more of a approximation of one players and then MM can evenly distribute (as well as possible ) to setup the match.   Just an idea I'm throwing out.   NOW if doing that..and one team really gets steam rolleres..well thats the way it is..MM has done its ob. 

 

If such a thing was possible its also too bad that you cannot have a 10 or more second view of the opponents you will play against and load out what you don't want and load in what you think will work depending on the opposition setup. I was thinking of if there are no CV's..why have a strong AA suite when you could have chosen another alternate setup. And visa versa. 

Share this post


Link to post
Share on other sites
Members
3,554 posts
6,867 battles
53 minutes ago, Ensign_Pulver_2016 said:

If the teams were balanced by skill, half the higher skilled players lose every time due to lower skilled teammates. What does it solve?

It's misread every time I say it. Not overall win-rate--Individual ship win-rate. No idea where the dividing line would be, but basically you have a potato league based on individual ship WR say of 48% or less (and I have many ships I am that poor on) - no one higher in that battle based on ship WR. The rest are in the major leagues of higher ship WR. Not perfect, but radically better than what we have where 38%-ers are battling 70%-ers which is pure f*ckery MM.

Edited by Stauffenberg44
  • Cool 1

Share this post


Link to post
Share on other sites
1,470
[-K-]
[-K-]
Members
3,038 posts
12,524 battles

You'd have to go back and dig through the patch notes to see exactly what the rules are (the Wiki page seems to be missing a few that I think are in there), but here's what I remember:

  • Top-tier and mid-tier battleships are equal between teams
  • Top-tier cruisers are equal between teams
  • CVs are equal between teams AND...
    • CVs are max 2 per team in T4-T7, 1 per team in T8-T10
  • Number (not size) of divisions per team are equal unless there's an odd number, in which case somebody gets the extra one
  • Battleship count, cruiser count, and destroyer count will match within 1 unless a division makes that impossible
    • For example, if there are 4 destroyers in a game but 3 of them are divisioned together, you'll get 3 DDs on one team and 1 on the other
  • Total team tier (add up all the ship tiers) will match within 2-3
  • The MM sort of tries to balance nations within ship types, but the other rules take precedence
  • Cool 1

Share this post


Link to post
Share on other sites
575
[PLPTV]
Members
1,457 posts
9,077 battles
4 hours ago, aethervox said:

Yep, that's what the WG Zombies say .... :Smile_great:

 

Hey, according to them the matchmaker is working "as intended". So, apparently their method of fixing the MM and REMOVING seal clubbing is by INTRODUCING seal clubbing of tier 5s by tier 7s and tier 8s by tier Xs. 

 

I wonder whats the IQ level of the people who come up with these braindead ideas.

  • Cool 1

Share this post


Link to post
Share on other sites
2,016
[PVE]
Members
5,315 posts
17,798 battles
2 minutes ago, Ulthwey said:

I wonder whats the IQ level of the people who come up with these braindead ideas.

Whatever you do, don't say they should be fired.. That is considered an attack..... Just a PSA, don't taze me bro, don't taze me.

Share this post


Link to post
Share on other sites
Members
3,638 posts
13,487 battles
15 minutes ago, Ulthwey said:

 

Hey, according to them the matchmaker is working "as intended". So, apparently their method of fixing the MM and REMOVING seal clubbing is by INTRODUCING seal clubbing of tier 5s by tier 7s and tier 8s by tier Xs. 

 

I wonder whats the IQ level of the people who come up with these braindead ideas.

But they never explain what the intent is. I wonder if they can explain it.

Share this post


Link to post
Share on other sites
Members
426 posts
6,213 battles
2 hours ago, Edgecase said:
  • The MM sort of tries to balance nations within ship types, but the other rules take precedence

I'm (reasonably) sure this got removed by a smallish patch sometime before the 6.# patches started.

 

6 hours ago, awiggin said:

Yes, and since everyone knows who the player base is...why wouldn't a company recognize it and adjust?

Definitely; I may average a low to even mid-50s WR over time, but the perhaps two (yet typically only one) sessions a week that aren't a two-thirds if not three-quarters or more ''loss-fest'' is why I barely play anymore.

No one enjoys almost non-stop losing until suddenly the game decides to work and/or let folks do as well as they can play because ''reasons''; WG however never seems to quite understand this with any of their games.

 

When I am playing badly, I know - but when I am repeatedly doing that poorly yet still finishing in the upper half if not outright top of a team I can tell when the game just doesn't want me to be playing that session.

That I am so often finding myself in that situation, or seeing other user names I know ending up with the same thing, I truly wonder if the company even understands how poorly their MM/coding/etc works as implemented.

Edited by Soylent_Red_Isnt_People

Share this post


Link to post
Share on other sites
5
[ARRSE]
Members
29 posts
1,019 battles
9 hours ago, Raptor_alcor said:

Occam's razor, yes it is the players fault, a bad MM setup can be won still if the players are smart and work together. The fact it doesn't so often is because the playerbase as a whole is quite poor. 

 

Thats not logical, though. If the teams are picked at random then the chances of each team having players that are "smart and work together" will be the same on both sides. Ergo, the team that fortuitously has the best equipment is more likely to win (obviously over a large sample size).

 

If WG state that MM is working as intended, then it probably is - but then WG most likely has a game management strategy that does not necessarily align with players' assumptions or expectations. WG needs to keep the majority of the playerbase within a certain WR band, and uses certain techniques to achieve this - the obvious one being in-game RNG of shots. "Bad MM" would be advantageous to this process, as it is essentially a different application of RNG.

Share this post


Link to post
Share on other sites
[SYN]
Members
509 posts
5,971 battles
10 hours ago, Stauffenberg44 said:

It's misread every time I say it. Not overall win-rate--Individual ship win-rate. No idea where the dividing line would be, but basically you have a potato league based on individual ship WR say of 48% or less (and I have many ships I am that poor on) - no one higher in that battle based on ship WR. The rest are in the major leagues of higher ship WR. Not perfect, but radically better than what we have where 38%-ers are battling 70%-ers which is pure f*ckery MM.

I misread that, but the Law of Large Numbers will still apply.  

 

Okay, so in that case, the >60% are driven down to a lower league and the >45% are driven up to a higher league. They still will meet each other until they once again achieve above or below the mean, or whichever point MM has been programmed to recognize. 

Three leagues: (1) Below average in a particular ship, fighting to reach average, (2) Above average in a particular ship, fighting to stay there, and (3) average in a particular ship. Group # consists of the players that have been bounced out of the first two groups seeking to find their WR level based on individual skill (W/O the intervention of this SB MM.) 

 

My Hypothesis: 

Most players will progress or regress into group 3 cyclically. Therefore, group 3, which will see the lion's share of battles, will be as the MM is today. 

 

Share this post


Link to post
Share on other sites

  • Recently Browsing   0 members

    No registered users viewing this page.

×