Jump to content
You need to play a total of 50 battles to post in this section.
Genie_of_the_Lamp

Matchmaking is getting even worse

18 comments in this topic

Recommended Posts

Players
142 posts

More and more games; also in Ranks are unfair setup. For the first 4 pictures you can see amount of DD s is not equal. This should not happen but it happens more and more.

In one match we were missing a dd and what we got back for it was a worthless Omaha in a tier 7 game. great.

 

Another example of messed up MM is you play a match fi Ranked match and you have to be up against 2 or 3 radar boats in a domination match while your team has none.

 

Every time WG telling us they improved matchmaking but this is pointing towards the other direction.

shot-21.03.20_17.43.21-0055.jpg

shot-21.03.20_17.50.17-0949.jpg

shot-21.03.19_21.00.45-0619.jpg

shot-21.03.13_22.12.25-0226.jpg

shot-21.03.14_18.32.50-0281.jpg

  • Cool 2

Share this post


Link to post
Share on other sites
Players
434 posts
6,856 battles

A violation of the mm rules can happen if the waiting time becomes too long, which can also affect class distribution. Divisions can make things even worse. That is rare though. 

Also, you fail to understand what wg means with improved mm. The last improvements were a soft-cap for classes I think, regulation of -2 matches and a rule that ensures that in all multi-tier battels a minimal and maximum amount of ships of the same tier are included. All these improvements are working and have improved the mm. When wg says they improved the mm it does not mean that they changed the area you complain about.

Also, the mm rules themselves have not changed in the last couple of month, therefore it's impossible that the mm got worse lately. Only your rng could have been worse recently. 

Share this post


Link to post
Share on other sites
[CATS]
Players
31,103 posts
15,723 battles
38 minutes ago, Genie_of_the_Lamp said:

Every time WG telling us they improved matchmaking but this is pointing towards the other direction.

When you think the old MM is better, it would be interesting to see how you enjoy uneven numbers of CV per team.

  • Boring 2
  • Bad 1

Share this post


Link to post
Share on other sites
[CHEFT]
Players
12,664 posts
9,841 battles

Too many BBs in queue.

Faildivs which just [edited]everyone...

Well the last screenshot - what MM gonna do about it? Split the division up because hurrdurr? They play 2 Radar Cruisers in one div, if there are no other T8 Radar Cruisers in queue - not MMs fault. Well yes, there is also Kronshtadt, but theoratically Neptune can also be Radar so...

  • Cool 1

Share this post


Link to post
Share on other sites
[SERBS]
Alpha Tester, Players, In AlfaTesters
2,625 posts
9,660 battles

c7jlmom916o61.png

  • Cool 14

Share this post


Link to post
Share on other sites
Players
1,972 posts
5,228 battles

If you're in a kamikaze + double Giulio Cesare division you kinda forfeit the right to complain about matchmaker in my opinion :fish_aqua:

  • Cool 3
  • Funny 5

Share this post


Link to post
Share on other sites
[R3B3L]
Players
1,432 posts
32,703 battles
19 hours ago, Genie_of_the_Lamp said:

For the first 4 pictures you can see amount of DD s is not equal. This should not happen but it happens more and more.

Yeah, I noticed this too. This is from yesterday, random battle

toll.jpg.1ee2b48daeca7b7c8e4652fa5ec532c5.jpg

 

Also, was a blast playing the only cruiser while CV kept me spotted. Although getting used to it, grows a thicker skin :cap_cool:

Share this post


Link to post
Share on other sites
[ASS]
Beta Tester
107 posts
18,756 battles

What WG doesn't tell us. The secret behind random matchmaking! :Smile_trollface:

 

The true and secret philosophy of WG's random matchmaking is easy to understand. WG is not interested in matching equally successfull teams to fight against each other in random games. That would lead to player-win-rates of ca. 50% average for all players over time. No matter how good you play, you will win and lose on the average the same amount of games. Chance will create all win ranks, not your personal playing ability. But WoWs would become very boring, if nobody could boast any significant over average achievments - and players losing interest is surely the last thing WG would want to happen to insure cash-flow! This is where the ingenious matchmaking comes in. The only real aim of WG matchmaking is to keep better player's egos happy, and to keep the poorer players to grinding, hoping to achieve better ranks.

 

Have you ever noticed, that you win more games when you play with ships that have a higher ship-win-rate than 50% - the higher the ship-win-rate, the more games you win! The vice versa is also observable - the lower your ship-win-rate, the more games you lose. After analysing this phenomenon, I believe that the reason why this happens, is do to intended manipulation of matchmaking by WG. WG often matches teams with an average ship-win-rate above 50% (the "winners") against teams below 50% (the "victims"). The higher your ship-win-rate is, the better your team average ship-win-rate will be - and the more chance your team has to win.

 

If you don't believe this, try it out yourself. Play a few random games with one of your most win-rate successfull ships, and you will find yourself much oftener in a "winning team" with an over 50% average team ship-win-rate against "victim teams" with below 50% average ship-win-rates! Or even more convincing - if you are a good player with a high overall win-rate, take out one of your ships with a lower ship-win-rate than 45% and see what happens. You will find yourself in a "victim team", with an average ship-win-rate below 50% against a winning team with an average ship-win-rate above 50%, and you will lose most of your games, no matter how good you play! The madness follows - you now believe, that you can't play this ship very well and will stop playing it. But the truth is, you are losing all those games because your team mates are poor players and they pull down your ship-win-rate with their constant losses as a "victim" team.

 

To check my assumption with stats, just use a monitor like "WoWs Monitor". Set the monitor to show the ship-win-rates of all players in both teams, showing how successfull every player wins games with his chosen ship. On the top of the screen over both team tables the monitor will show the average ship-win-rate of all players in each team. This is a much better rating assessment to judge a game outcome than the use of average player-win-rates, which could be completely misleading, when the selected ship-win-rate of your team ship is far from the overall average player win-rate. Example: A player could have an general average win-rate of 65% over all games, but choses to play in the current game with a ship with only a 45% ship-win-rate. That means, he plays a chosen ship which is lower than his overall average game-win-rate. With the help of this monitor, you can now see the win forcast at the beginning of a game by comparing the average ship-win-rates of both teams at the top of the table. The team with the higher average ship-win-rate will have a higher chance of winning the match - the bigger the difference, the better the chance. And here you begin to realise that the game is not "fairly" matched! The average ship-win-rates should be close to 50% on both teams to allow fair chances of winning for both teams with a mix of good and poor players on both teams.

 

This is the intension of WG coming into play. There should be a winning and a losing team to keep up the moral of the better players and their ability to boast their achievements. A fair mixture of players in both teams (same average ship-win-rates) would only lead to many draw engagements (equal won and lost games) on the long run - killing all the "Top Gun" exitement, and all the cash flow!

 

The bad news is - for all poor players, this means that your chances to climb up the ranking ladder is very unprobable, because you are mostly playing in victim teams with lower win chances...and that is very frustrating!

  • Cool 1
  • Boring 3

Share this post


Link to post
Share on other sites
[JRM]
[JRM]
Players
7,774 posts
9 hours ago, Grandma said:

the higher the ship-win-rate, the more games you win!

DUH m8

Share this post


Link to post
Share on other sites
[CATS]
Players
31,103 posts
15,723 battles
10 hours ago, Grandma said:

... the higher the ship-win-rate, the more games you win!

As Yedwy so eloquently explained, that should be obvious. But congratulations on noticing :cap_like:

 

It should also be obvious that ships with above average winrating raise the average WR of your team and that ships with below average winrating reduce your average WR of your team, influencing the average expected outcome of the match.

If there were some sort of rigging, as you explained, ships with high WR could not get worse and ships with low WR could not get better. And as you explained yourself, bad players could not get better.

 

The reality is that most players started as bad players and got better over time. The stats are are available to see that.

 

My Udaloi had at one time around 48% WR, according to your theory, she could only got worse, but she did not.

Udaloi was one of my oldest high Tier DD and I played much worse back then. Additionally the ship got improved over time. These are things that influence your performance, not some rigging.

Spoiler

Screenshot_20210404-005221.thumb.png.616a375cf1a7fc4b5b3167e964df3a0e.png

 

On the other hand my Harekaze was at one time at around 64% WR. According to your theory, she should not get worse, but she did.

Spoiler

Screenshot_20210404-005238.thumb.png.d365450d1b01afdf3bab313a9856f9ba.png

 

Looks like your theory does not hold up to reality.

  • Bad 1

Share this post


Link to post
Share on other sites
[CHEFT]
Players
12,664 posts
9,841 battles
11 hours ago, Grandma said:

Have you ever noticed, that you win more games when you play with ships that have a higher ship-win-rate than 50% - the higher the ship-win-rate, the more games you win! The vice versa is also observable - the lower your ship-win-rate, the more games you lose. After analysing this phenomenon, I believe that the reason why this happens, is do to intended manipulation of matchmaking by WG. WG often matches teams with an average ship-win-rate above 50% (the "winners") against teams below 50% (the "victims"). The higher your ship-win-rate is, the better your team average ship-win-rate will be - and the more chance your team has to win. 

 

If you don't believe this, try it out yourself. Play a few random games with one of your most win-rate successfull ships, and you will find yourself much oftener in a "winning team" with an over 50% average team ship-win-rate against "victim teams" with below 50% average ship-win-rates! Or even more convincing - if you are a good player with a high overall win-rate, take out one of your ships with a lower ship-win-rate than 45% and see what happens. You will find yourself in a "victim team", with an average ship-win-rate below 50% against a winning team with an average ship-win-rate above 50%, and you will lose most of your games, no matter how good you play! The madness follows - you now believe, that you can't play this ship very well and will stop playing it. But the truth is, you are losing all those games because your team mates are poor players and they pull down your ship-win-rate with their constant losses as a "victim" team. 

 

Not what ive experienced.

F.e. with GK, from the first 13 games i had only 3 wins. With 30-ish games i still had around 40% WR. Now i have 80 games in it with 65% WR.

Same with Nagato. For the first 20-30 games or so i had horrible teams, red WR. Then when i got to 30 games i suddenly started winning and finished the grind with 36 battles and 53% WR.

With Fuso, i also had crap MM attached to it. But i didnt play it more to balance it out, i know that id eventually win more games - it was just bad luck with teams. At any given moment you can have a streak of bad teams, sometimes it happens with a new ship, sometimes it happens after 100 battles with a ship.

Share this post


Link to post
Share on other sites
[-DGH-]
[-DGH-]
Players
424 posts
3,468 battles

I anything the problem is that the matchmaker is not rigged.

 

But it should.

 

It should at least try to match equally strong teams together, especially in regards of divisions, to minimize landslide victories / losses.

 

That is not that difficult according to the stats already monitored.

Share this post


Link to post
Share on other sites
[ASS]
Beta Tester
107 posts
18,756 battles

 

23 hours ago, DFens_666 said:

Not what ive experienced.

F.e. with GK, from the first 13 games i had only 3 wins. With 30-ish games i still had around 40% WR. Now i have 80 games in it with 65% WR.

Same with Nagato. For the first 20-30 games or so i had horrible teams, red WR. Then when i got to 30 games i suddenly started winning and finished the grind with 36 battles and 53% WR.

With Fuso, i also had crap MM attached to it. But i didnt play it more to balance it out, i know that id eventually win more games - it was just bad luck with teams. At any given moment you can have a streak of bad teams, sometimes it happens with a new ship, sometimes it happens after 100 battles with a ship.

You don't see the rigging because you don't know when it really started. Look at your GK - somewhere between 30 and 80 games you reached a Win-Rate of 50%. At that moment you are matched more often with better teams (+50%). But you didn't notice that point because you weren't screening your games. The same with your Nagato - at some point between (probably at game 30) you reached a Win-Rate of 50%. At that moment you write: "I suddenly started winning". That is presumable the moment you reached a 50% Win-Rate. Those following 6 games helped carry you to 53%, because you were matched to "Winning-Teams" more often. Those Winning-Teams carried you to win games, no matter how you played. And you believed you were suddenly getting better because of a skill boast, Maybe you did get better, but the main reason you managed 53% was the Team-Boast, helping you to win.

 

Everybody believes that WG is not using strength stats to match teams in Random Games. I believe they do just that, to keep up the good feelings that winning achievements offer. Look at all the boasting and judging with concern to stats ingame and in the forums. So many players bragging how good they are with regard to other players with lower stats. They don't realise, that WG is permantly helping them to keep their stats up or down.

 

I truly believe nobody could achieve a 65% or better Win-Rate over time if the random team matches were setup fair. Only a few players would be capable to carry some game wins alone - but the balancing force would drag them down over time. All the team-Win-Rates would move toward 50% over time, because the winning probabilty would be balanced, and the 65% players could not hold those high win levels anymore, because all the teams he plays in are also are leveling to a 50% win rate over time. And that would surely become very boring to the community, causing many players to leave the game - and that would concern WG's money flow anticipation very much! That is why a hidden unbalancing algorithm is very probable.

  • Cool 1
  • Funny 2

Share this post


Link to post
Share on other sites
[NWP]
Players
3,718 posts
18,027 battles
22 hours ago, xe_N_on said:

I anything the problem is that the matchmaker is not rigged.

 

But it should.

 

It should at least try to match equally strong teams together, especially in regards of divisions, to minimize landslide victories / losses.

  

That is not that difficult according to the stats already monitored.

You didn't think this through now did you.

 

What happens if you start equalizing teams for WR ( or whatever other metric you want to sue )?

Said metric becomes useless as it doesn't show the playerskill anymore, it'll get skewed by a MM intervention rendering it a useless metric.

And then how do you plan to "balance" the MM when your metric is useless?

  • Bad 1

Share this post


Link to post
Share on other sites
[-DGH-]
[-DGH-]
Players
424 posts
3,468 battles
Vor 14 Minuten, 159Hunter sagte:

You didn't think this through now did you.

 

What happens if you start equalizing teams for WR ( or whatever other metric you want to sue )?

Said metric becomes useless as it doesn't show the playerskill anymore, it'll get skewed by a MM intervention rendering it a useless metric.

And then how do you plan to "balance" the MM when your metric is useless?

 

As 100 times discussed WR cannot be used to equalizing teams.

 

But there are other stats, as game played, average damage, average experience or PR that could be used. Of course, non of them can measure the player skill with 100% accuracy. And you have to introduce some kind of dynamic league levels like in ranked, because a unicum player will perform worse over time if matched with other unicum players. But everything is better than having random matchmaking as currently.

 

Lets consider a European soccer league that randomly mix players ranging from parents, that occasionally play soccer with their kids, together with pro league soccer player and call this "match making". Would be a success, wouldn't it?

Share this post


Link to post
Share on other sites
[NWP]
Players
3,718 posts
18,027 battles
22 minutes ago, xe_N_on said:

 

As 100 times discussed WR cannot be used to equalizing teams. But there are other stats, as game played, average damage, average experience or PR that could be used. Of course, non of them can measure the player skill with 100% accuracy. But everything is better than having random matchmaking as currently.

games played - how is that an indication of skill? this is useless for balancing teams

average damage - useless as well as it is class and even ship dependent. Plus 20k damage dealt to a DD =/= 20k damage dealt to a BB (and there is no way to track this atm)

average experience - useless as premium time is taken into account

PR - useless as it is heavily skewed towards damage (but in relation to a specific ship) but it also uncludes WR (which will flatten out over time )

 

Any other metric you want to include?

22 minutes ago, xe_N_on said:

Lets consider a European soccer league that randomly mix players ranging from parents, that occasionally play soccer with their kids together with pro league soccer player and call this "match making". Would be a success, wouldn't it?

?

Share this post


Link to post
Share on other sites
Guest
This topic is now closed to further replies.

×