Matchmaker Discussion Thread & MM Balance

[Dutchy_2019]

11 minutes ago, DFens_666 said:

 

That only happens when both sides have a faildiv with a T6 CV, because CVs are always mirrored. So it doesnt provide a disadvantage for one team only, but in general, i could live without faildivs. Most people only do it to get some kinda advantage. I mean, you could allow faildivs only from T8-10, since there it doesnt matter anymore.

 

I am not sure about that. Not sure whether I myself was in one, or I saw it online. And IIRC the CV tiers were NOT mirrored.

[ForlornSailor]

1 hour ago, Dutchy_2019 said:

And IIRC the CV tiers were NOT mirrored.

 

Thats not possible and almost never was possible (unless in really really early days of this game, which is irrelevant to any discussion at this point).

 

CVs are always matched absolutly exactly in tiers and numbers.

[Dutchy_2019]

- I know they are matched in number

 

- And as for whether there were 2 fail division CV's, I do not remember for certain, that is why I said 'IIRC'.

[Garry_Flood]

match making is utter garbage its worse now than it ever has  been im not going to explain  because you know this already 

[DFens_666]

4 minutes ago, Garry_Flood said:

match making is utter garbage its worse now than it ever has  been im not going to explain  because you know this already 

 

Only because more potatoes play due to Corona-virus. Its not MM fault that the players suck.

[FAP_pt]

I think MM is ordering players not only by tier or ship randomly but as a function of what rank they are in the game. This means when you start the client in the first game you are in a low ranked team (because you usually dont play just one game), the next you are on a better ranked team (so you dont log out after desperation of losing 2 in a row), this is intended to keep us playing as much of the time possible. With so many information on individual players i think its obvious that they will use that in order to balance not the match itself but our persistent on the server.

This is why battles are so uneven, its a easy win or a dramatic loss, battles to the last second on points or ships are extremely rare.

But if they didnt use this system it would be worse, like in the old days when you had 9 or 10 games where you lose or win in a row.

But they should be clean around this.

Remember, its the not the ships, its the players who play them that can make a difference.

[ColonelPete]

1 minute ago, FAP_pt said:

I think MM is ordering players not only by tier or ship randomly but as a function of what rank they are in the game. This means when you start the client in the first game you are in a low ranked team (because you usually dont play just one game), the next you are on a better ranked team (so you dont log out after desperation of losing 2 in a row), this is intended to keep us playing as much of the time possible. With so many information on individual players i think its obvious that they will use that in order to balance not the match itself but our persistent on the server.

This is why battles are so uneven, its a easy win or a dramatic loss, battles to the last second on points or ships are extremely rare.

But if they didnt use this system it would be worse, like in the old days when you had 9 or 10 games where you lose or win in a row.

But they should be clean around this.

Remember, its the not the ships, its the players who play them that can make a difference.

Do you have proof?

Why do people regularly win in these low ranked teams?

Why do so many people lose twotimes in a row?

Why implement a system to prevent something that happens anyway?

 

[FAP_pt]

I dont have proof, its just my opinion, you can have the same ships on each team  but if you don't balance the players it won't matter.

If you are in a low ranked team its because you won the last one (its just my experience, try it by yourself and check the ranked position on the players and add up either sides, usually you lost because you are in the team with a lower ranked aggregate) and usualy (not always) it doesnt matter what you do you end up losing.

I tend to use the best ships (wich whom i do best) on the games after i just won.

Of course with this kind of MM its not exact science but it will help maintain the players longer in the server, its the best policy from the game managers... The MM its not about the player satisfaction ... or better, its about the player satisfaction as long as it continues to play...

 

[ColonelPete]

19 minutes ago, FAP_pt said:

I dont have proof, its just my opinion, you can have the same ships on each team  but if you don't balance the players it won't matter.

If you are in a low ranked team its because you won the last one (its just my experience, try it by yourself and check the ranked position on the players and add up either sides, usually you lost because you are in the team with a lower ranked aggregate) and usualy (not always) it doesnt matter what you do you end up losing.

I tend to use the best ships (wich whom i do best) on the games after i just won.

Of course with this kind of MM its not exact science but it will help maintain the players longer in the server, its the best policy from the game managers... The MM its not about the player satisfaction ... or better, its about the player satisfaction as long as it continues to play...

• Rankings are not a good measurement of skill
• the distribution of Rankings does not follow any pattern, since it is random
• as I explained, your theory is not working, there is no reason to implement something that does not work

 

[snddim01]

3 hours ago, FAP_pt said:

I think MM is ordering players not only by tier or ship randomly but as a function of what rank they are in the game.

I believe there may be an element of truth to this.

 

There's a tool called Matchmaking Monitor from Aslains Mod Pack. It lets you see the stats of every player in both your team and the opposing one, providing no-one has chosen to hide those stats.

 

I've been using it for approximately my last five hundred battles. This has covered my grind of around half a dozen ships.

 

At the start of each battle, I take a note of which team has the highest number of ship experience. I use that as a guide as to which team is the stronger and then record the number of times I get the weak team or the strong one.

 

Thus far at least, a clear pattern has emerged. A pattern which depends on my own performance playing that ship.

 

If I've performed well, (say by getting more kills than average on a ship), then on more occasions than not I've been placed on the "stronger" team. This has proven true for Skane, Oland, Implacable and Roon.

 

If  I've performed badly, then on more occasions than not I've been placed on the "weaker" team. This has proven true for Alsace and Ostergottland.

 

It may be pure coincidence. After all, these are only six ships and I'm only one player. And it could be debated whether picking one stat from a team lineup is the best way to measure matchmaking.

 

However, my experience suggests that the best way to "get lucky" with your teammates is to be good in the ship you're playing. And if you're bad, well, don't expect to get a run of games where the unicums are all on your side.  

[ColonelPete]

26 minutes ago, snddim01 said:

If I've performed well, (say by getting more kills than average on a ship), then on more occasions than not I've been placed on the "stronger" team. This has proven true for Skane, Oland, Implacable and Roon.

 

If  I've performed badly, then on more occasions than not I've been placed on the "weaker" team. This has proven true for Alsace and Ostergottland.

• What should be the purpose?
• It contradicts FAPs theorie
• Why does your theory only work for few ships?

That is no theory. That is random.

[snddim01]

14 minutes ago, ColonelPete said:

• What should be the purpose?
• It contradicts FAPs theorie
• Why does your theory only work for few ships?

That is no theory. That is random.

 

I have no idea what the purpose is. I'm simply agreeing with the suggestion that there may be a link between a players performance and the quality of teammates they get. Averaged out over a large number of battles.

 

It doesn't "work for only a few ships". It applies to all the ships I've been playing in the last month since I started recording information from MMM.

 

Noting a trend from 500+ battles is not "random".

 

 

[Floofz]

1 hour ago, snddim01 said:

I believe there may be an element of truth to this.

 

There's a tool called Matchmaking Monitor from Aslains Mod Pack. It lets you see the stats of every player in both your team and the opposing one, providing no-one has chosen to hide those stats.

 

I've been using it for approximately my last five hundred battles. This has covered my grind of around half a dozen ships.

 

At the start of each battle, I take a note of which team has the highest number of ship experience. I use that as a guide as to which team is the stronger and then record the number of times I get the weak team or the strong one.

 

Thus far at least, a clear pattern has emerged. A pattern which depends on my own performance playing that ship.

 

If I've performed well, (say by getting more kills than average on a ship), then on more occasions than not I've been placed on the "stronger" team. This has proven true for Skane, Oland, Implacable and Roon.

 

If  I've performed badly, then on more occasions than not I've been placed on the "weaker" team. This has proven true for Alsace and Ostergottland.

 

It may be pure coincidence. After all, these are only six ships and I'm only one player. And it could be debated whether picking one stat from a team lineup is the best way to measure matchmaking.

 

However, my experience suggests that the best way to "get lucky" with your teammates is to be good in the ship you're playing. And if you're bad, well, don't expect to get a run of games where the unicums are all on your side.  

 

Im usually all for conspiracy theories in the MM but I dont believe this to be true. I have ships in my port that simply refuse to win no matter what, doesnt matter when I play them. Kremlin I have an under 30% WR with despite having decent overall stats in it for example.

And then I have ships like my Tirpitz that wins even how hard I fail.

If your theory were true then the WR percentage would automaticly even out if you play 1-2 games each day/session.

[ColonelPete]

23 minutes ago, snddim01 said:

 

I have no idea what the purpose is. I'm simply agreeing with the suggestion that there may be a link between a players performance and the quality of teammates they get. Averaged out over a large number of battles.

 

It doesn't "work for only a few ships". It applies to all the ships I've been playing in the last month since I started recording information from MMM.

You named only a handful.

And your theory still contradicts FAPs theory.

[snddim01]

54 minutes ago, ColonelPete said:

You named only a handful.

And your theory still contradicts FAPs theory.

 

Point 1 has already been answered.

 

Point 2  - so what? Does that disqualify me from passing an opinion on part of his post? The part that I quoted rather than the entirety of it?

[ColonelPete]

2 minutes ago, snddim01 said:

 

Point 1 has already been answered.

 

Point 2  - so what? Does that disqualify me from passing an opinion on part of his post? The part that I quoted rather than the entirety of it?

One can also be of the opinion the moon is a big ball of cheese.

[Bear__Necessities]

6 hours ago, Floofz said:

 

Im usually all for conspiracy theories in the MM but I dont believe this to be true. I have ships in my port that simply refuse to win no matter what, doesnt matter when I play them. Kremlin I have an under 30% WR with despite having decent overall stats in it for example.

And then I have ships like my Tirpitz that wins even how hard I fail.

If your theory were true then the WR percentage would automaticly even out if you play 1-2 games each day/session.

Holy s**t I agree with you on something for once 

 

So, I have Midway and Hak on both EU and NA. Lets look at the difference in WR. As @Floofz pointed out, for that theory to work, it would have to replicate the same result no? Well, as you will see, they don't. 

 

 

 

Ohhhh and the rage that my CV's at T10 can't seem to buy wins on NA must warm the heart of CV haters 

 

[159Hunter]

To be fair @Bear__Necessities your sample on NA is a tad low, so maybe get them to 50 - 100 and it could level out ( if you can stomach it)

[DFens_666]

7 minutes ago, 159Hunter said:

so maybe get them to 50 - 100 and it could level out

 

Im pretty sure itll get better after ~30 games. Happened to me on several occasions, and usually after 30 games it starts to get normal. Ive also had the opposite, get easy wins despite doing nothing for same amount of games, then the losing starts at some point.

[snddim01]

6 hours ago, Bear__Necessities said:

Holy s**t I agree with you on something for once 

 

So, I have Midway and Hak on both EU and NA. Lets look at the difference in WR. As @Floofz pointed out, for that theory to work, it would have to replicate the same result no? Well, as you will see, they don't. 

 

 

 

Ohhhh and the rage that my CV's at T10 can't seem to buy wins on NA must warm the heart of CV haters 

 

 

 

I was suggesting a correlation between player skill and the team-mates he gets. Your example refers to player skill and winrate. They're not quite the same thing.

 

Let me give you another stat.

 

I've played seven games this morning. I've lost every single one. On each occasion, the matchmaking has been very similar to this, which is a screenshot taken from my last game. The enemy has a far higher win rate and exp in their ships than my team does, plus the advantage of four good players. Our very best player is only average. The team lineups have looked like this in every single one of my seven games. Totally imbalanced. Unsurprisingly, my team collapsed in virtually all of those games. Only one of them lasted the full 20 minutes.

 

My win rate in Saint-Louis is abysmal. But the other stats aren't too bad. My kills are higher than average for the ship and my damage isn't far off average, despite playing in a stock ship which loses every game after barely ten minutes.

 

Now here's a prediction.

 

I generally like kiting cruisers and play reasonably in them. So my personal stats will gradually improve as I upgrade it and by the time I've finished with SL I'll probably be one of the green (good) players.

 

More importantly, the better I get the more often I'll be placed in the stronger team. That seems nonsensical but I've found it to be true for every ship played since I started monitoring my teams. By the time I finished with Roon I was on the stronger team in almost every match. The opposite also applies. If I don't improve, then more often that not my teammates will be potatoes.

 

Why? I have no idea. But for me personally, it seems to be a recurring theme as monitored over the last 500 or so battles.

 

Hence why I believe that player performance stats are somehow incorporated into matchmaking.

[MementoMori_6030]

There should be a checkbox pop-up after hitting the BATTLE button:

[  ] I do not play this ship to face +2 tiers opponents more often than not, I play this ship to face equal tier opponents. I have patience to wait for that kind of match.

[  ] I do not care into which kind of  you pitch me, give me your worst.

[diegobrando_2019]

Lost 7 games in a row. All except 1 complete stomps. Every single game atleast 3 people die immediately when they see enemy. There is no *MM* in this game. It just balances the ship types and tiers asap and that's it. Then I get the supercontainer from the twitch missions only to get useless flags. Wp game.

[DFens_666]

59 minutes ago, snddim01 said:

Hence why I believe that player performance stats are somehow incorporated into matchmaking.

 

To me, that makes no sense.

 

 

Why was the Smolensk in our team? Why was the Stalingrad in the enemy team? If you only put overperforming players in one team, they couldnt maintain it, since there is a limited amount of HP and kills floating around.

Also in your own example, why are you in the "bad" team, and not the Donskoi, which is worse than you are in that ship? It could swap you and the Donskoi without a problem. Same question for Hipper and Amalfi.

[snddim01]

5 minutes ago, DFens_666 said:

 

To me, that makes no sense.

 

 

Why was the Smolensk in our team? Why was the Stalingrad in the enemy team? If you only put overperforming players in one team, they couldnt maintain it, since there is a limited amount of HP and kills floating around.

Also in your own example, why are you in the "bad" team, and not the Donskoi, which is worse than you are in that ship? It could swap you and the Donskoi without a problem. Same question for Hipper and Amalfi.

 

I used this match as an example of a trend. I didn't say it happens every time. That would be impossible for the  reasons you've just explained. 

 

"More often than not" was the phrase I used in a previous post. 

 

[Excavatus]

1 hour ago, diegobrando_2019 said:

There is no *MM* in this game. It just balances the ship types and tiers asap and that's it.

Thats basically what MM does in this game :) 

Ship types and tiers.. thats all :)