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

Is there any way for the match maker system to consider amount of games played in ship?

18 comments in this topic

Recommended Posts

206
[UFFA]
[UFFA]
Members
768 posts
63 battles

I realize the MM system will never be skill based and I'm not asking for this. Let's be honest I make potatoes look good. However it surely can't be that hard to get a balance via games played. The last couple of weeks watching horrendous loss streaks I've noticed one constant. There is generally a lopsided amount of games played in favor of one team or the other. If the Match maker monitor information is to be believed of course. With little fail the team with a larger pool of games played wins. There should be some randomness to this and I'm failing to see it so far.

Just an example: In this picture it looks like the Green team should easily mop up this game based on win rate. However look at the disparity in games played. And indeed the game was your trademark one side tries to kite away from the caps while the other side plays the objective. Ending in a horrible loss for the green team. Which in my case is pretty much 4 out of 5 games these days.

 

yYkM2sV.png

  • Boring 1

Share this post


Link to post
Share on other sites
Members
473 posts
12,505 battles

The level of complexity this would add to MM just isn't worth it. Hard enough to just get 24 boats to even out.

  • Cool 2

Share this post


Link to post
Share on other sites
301
[CK5]
Members
808 posts
6,608 battles

Some people argue that win rate should be used to organize MM which would mean the Red team in your example was at a disadvantage. 

Share this post


Link to post
Share on other sites
294
[O7]
[O7]
Beta Testers
1,254 posts
7,782 battles

Paaaaleeesss.... some people are going to potate regardless of how many battles they play in a ship.

Share this post


Link to post
Share on other sites
Community Department
2,128 posts
520 battles
1 hour ago, Sparviero said:

I realize the MM system will never be skill based and I'm not asking for this. Let's be honest I make potatoes look good. However it surely can't be that hard to get a balance via games played. The last couple of weeks watching horrendous loss streaks I've noticed one constant. There is generally a lopsided amount of games played in favor of one team or the other. If the Match maker monitor information is to be believed of course. With little fail the team with a larger pool of games played wins. There should be some randomness to this and I'm failing to see it so far.

Just an example: In this picture it looks like the Green team should easily mop up this game based on win rate. However look at the disparity in games played. And indeed the game was your trademark one side tries to kite away from the caps while the other side plays the objective. Ending in a horrible loss for the green team. Which in my case is pretty much 4 out of 5 games these days.

Spoiler

yYkM2sV.png

 

Besides of the complexity (I think about impossibility. Noone can predict actions in the battle, but they always matter) It's 100% destructive to make any MM which somehow controls someone's winrate. It's only a question of time, when players could discover trick. And that would be fatal for such project. We understand this well, so there is no need to seek out what does not exist. 
*Аt the same time when I have several defeats in a row I laugh with collegues about maybe it's MM tricking.:cap_haloween: Of course does not.  

  • Cool 1

Share this post


Link to post
Share on other sites
206
[UFFA]
[UFFA]
Members
768 posts
63 battles
11 minutes ago, PrairiePlayer said:

Some people argue that win rate should be used to organize MM which would mean the Red team in your example was at a disadvantage. 

I’ll be honest I was of the same opinion. Then I started noticing my high 50s percentage teams with purple damage counters would crumble when faced with a team of even distribution. 

So I started looking elsewhere and this is my latest windmill. Isn’t it nice? :cap_haloween:

 

  • Cool 1

Share this post


Link to post
Share on other sites
Community Department
2,128 posts
520 battles
3 minutes ago, Sparviero said:

I’ll be honest I was of the same opinion. Then I started noticing my high 50s percentage teams with purple damage counters would crumble when faced with a team of even distribution. 

So I started looking elsewhere and this is my latest windmill. Isn’t it nice? :cap_haloween:

From my point of view victory may depend a little on number of battles for new players. But I don't think that any restrictions for others could matter.

Share this post


Link to post
Share on other sites
Members
567 posts
75 battles
1 hour ago, Sparviero said:

I realize the MM system will never be skill based and I'm not asking for this. Let's be honest I make potatoes look good. However it surely can't be that hard to get a balance via games played. The last couple of weeks watching horrendous loss streaks I've noticed one constant. There is generally a lopsided amount of games played in favor of one team or the other. If the Match maker monitor information is to be believed of course. With little fail the team with a larger pool of games played wins. There should be some randomness to this and I'm failing to see it so far.

Just an example: In this picture it looks like the Green team should easily mop up this game based on win rate. However look at the disparity in games played. And indeed the game was your trademark one side tries to kite away from the caps while the other side plays the objective. Ending in a horrible loss for the green team. Which in my case is pretty much 4 out of 5 games these days.

Can you explain please how number of games played is relevant to skill when many of the worst players in the game have thousands of games played more than skilled players?

Share this post


Link to post
Share on other sites
2,295
[PVE]
Members
9,930 posts
7,492 battles

One side is always going to have more battles than the other as getting a perfectly even number of battles on each side would be very hard. Now getting it closer to the same number of battles on each side would be *possible*, but difficult as you are adding a third item to match up teams with. Right now, tier and ship type are what is matched up to be evenly as possible. Each item you add to match up, increased the difficultly of getting an *even* match greatly.

 

39 minutes ago, turbo07 said:

Besides of the complexity (I think about impossibility. Noone can predict actions in the battle, but they always matter) It's 100% destructive to make any MM which somehow controls someone's winrate. It's only a question of time, when players could discover trick. And that would be fatal for such project. We understand this well, so there is no need to seek out what does not exist. 

 

Having the MM being random is the fairest way to do it, even if occasionally one side or the other is off balance. It will all balance out anyways over the long run.

 

Edited by Kizarvexis

Share this post


Link to post
Share on other sites
206
[UFFA]
[UFFA]
Members
768 posts
63 battles
Just now, A_Crying_Hipster said:

Can you explain please how number of games played is relevant to skill when many of the worst players in the game have thousands of games played more than skilled players?

Skill? Not sure. Experience? Through numbers of games played. :Smile-_tongue:

 

Please look at the preface of the subject title before thinking too seriously about this matter. :fish_boom:

Share this post


Link to post
Share on other sites
206
[UFFA]
[UFFA]
Members
768 posts
63 battles
Just now, Kizarvexis said:

One side is always going to have more battles than the other as getting a perfectly even number of battles on each side would be very hard. Now getting it closer to the same number of battles on each side would be *possible*, but difficult as you are adding a third item to match up teams with. Right now, tier and ship type are what is matched up to be evenly as possible. Each item you add to match up, increased the difficultly of getting an *even* match greatly.

 

 

Having the MM being random is the fairest way to do it, even if occasionally one side or the other is off balance. It will all balance out anyways.

 

I in no way mean an even number of battles. What I'm talking about is reducing the gulf when half a team has, let's say as an extreme, less than 10 games played and the other side has most players in their BFF. The games aren't fun for one side. :Smile_ohmy:

Random is great when you are on the part of the curve that gives out even distribution or even generous.  Then it is really great.  :Smile_glasses: When you start a match and over and over again your team refuses to contest caps, loses multiple ships in the opening minute(!), etc. Always attempts to kite to the back line you start wondering what the hell is going on. I started looking at the match maker monitor and as stated above I figured win percentage had to be the number to watch. Then I started noticing my statistically awesome teams where worse than my fellow 30% potatoes. So I started looking for other factors. Now my sexy new windmill is the amount of times each team has played in their ships.

 

Am I probably wrong, most likely. The way people have to make sure to say it is wrong instead of letting a pointless rant die. Now that's interesting. :Smile_popcorn:

Share this post


Link to post
Share on other sites
2,801
Members
9,996 posts

It's pointless. We have players with 20k games who can't sink a lead weight....number of games simply isn't a good measure.

Share this post


Link to post
Share on other sites
2,295
[PVE]
Members
9,930 posts
7,492 battles
4 minutes ago, Sparviero said:

I in no way mean an even number of battles. What I'm talking about is reducing the gulf when half a team has, let's say as an extreme, less than 10 games played and the other side has most players in their BFF. The games aren't fun for one side. :Smile_ohmy:

I covered that as you can see above and below where I highlighted it.

17 minutes ago, Kizarvexis said:

One side is always going to have more battles than the other as getting a perfectly even number of battles on each side would be very hard. Now getting it closer to the same number of battles on each side would be *possible*, but difficult as you are adding a third item to match up teams with.

 

5 minutes ago, Sparviero said:

Random is great when you are on the part of the curve that gives out even distribution or even generous.  Then it is really great.  :Smile_glasses: When you start a match and over and over again your team refuses to contest caps, loses multiple ships in the opening minute(!), etc. Always attempts to kite to the back line you start wondering what the hell is going on. I started looking at the match maker monitor and as stated above I figured win percentage had to be the number to watch. Then I started noticing my statistically awesome teams where worse than my fellow 30% potatoes. So I started looking for other factors. Now my sexy new windmill is the amount of times each team has played in their ships.

Do you have a 100 or more screen shots with WR for each team to back up your assertions? A 1,000 games would be better to have a significant sample size.

 

In any case, as Turbo07 said, fixing the matches based on player skill would be worse as then you would have a system that takes player skill (whatever metric you use) into account. The accusations of bias would grow exponentially as a portion of the player base that already complains that MM is biased against them would complain worse about a system that introduces bias.

 

Making the MM random, is the fairest way to go as lopsided results will even out over time as sometimes, you will be put on the lopsided team.

 

10 minutes ago, Sparviero said:

Am I probably wrong, most likely. The way people have to make sure to say it is wrong instead of letting a pointless rant die. Now that's interesting. :Smile_popcorn:

Are there any other motives for this rant?

Share this post


Link to post
Share on other sites
66
[RCB4]
[RCB4]
Beta Testers
472 posts

my mama said world of warship is like a box of chocolate...you never know what your gonna get 

)

  • Cool 1

Share this post


Link to post
Share on other sites
Members
208 posts
2,089 battles

Ok I hate to say this but has anyone ever noticed that most of the time the people who complain about MM being "broken" generally are not "good" players in any way shape or form and instead of saying hey I suck and my team sucks(which op did say) so let's go learn game mechanics so we can play better. No that's not the answer its MM fault everytime, honestly it's just sad to me to see so many people complain that MM is "broken" when currently it's the most fair and balanced it's ever been and probably ganna be.  If you lose a lot especially below like tier 6, you need to learn to carry like all good players in ANY multiplayer game. Btw not saying I'm one of them good players I know I'm not but I also know it's not matchmakers fault its mine and the morons on my team when we lose. Also just to be clear wasn't referring to OP just topic in general. 

Edited by Themothius

Share this post


Link to post
Share on other sites
610
[OPRAH]
[OPRAH]
Beta Testers
3,900 posts
12,634 battles

@Sparviero MM has enough trouble as it is. Also number of battles played does not equal skill gained. My battle count is rather high and yet I am hardly a highly skilled player.

Share this post


Link to post
Share on other sites
Members
208 posts
2,089 battles

I think the OP is trying to link battles played with knowing how to play the game.  Sadly there is no way for MM to know if a player took the time to learn game mechanics or not. You can play 10000 battle and still not understand how to cap or that kiting to the back of the map is only situational and more importantly the ability to aim.  I think they where trying to say that more battles equals better understanding of the game but sadly it doesnt.

Share this post


Link to post
Share on other sites
206
[UFFA]
[UFFA]
Members
768 posts
63 battles
7 hours ago, Themothius said:

I think the OP is trying to link battles played with knowing how to play the game.  Sadly there is no way for MM to know if a player took the time to learn game mechanics or not. You can play 10000 battle and still not understand how to cap or that kiting to the back of the map is only situational and more importantly the ability to aim.  I think they where trying to say that more battles equals better understanding of the game but sadly it doesnt.

Indeed. However it just seems like a controllable a little bit better. I am not looking for some Utopian solution.

Example this match was a complete disaster. Despite having a tactical advantage at a and a 3 person division they did not push until the game was decided.

 

rkhlXJw.jpg

 

:Smile-_tongue:

Share this post


Link to post
Share on other sites

  • Recently Browsing   0 members

    No registered users viewing this page.

×