Jump to content
Sign in to follow this  
You need to play a total of 50 battles to post in this section.
13Ruan13

A solution to stomps : having a MMR ?

22 comments in this topic

Recommended Posts

[ORCA]
Players
5 posts
6,281 battles

Hello there.

 

I'm posting this and there hoping to start a manner discussion about how MatchMaking can maybe impact the stomps.

 

Let's face it : stomps are not fun, even if you are on the winner side. You don't have time to really do something and get fun in this game, which should be the main objective. Stomps have become so much usual it's even a surprise when you play a close game or have fun playing WoWs. The fact is that no players with any conscience and a minimal knowledge about the game find fun to be stomped. 

 

So, what about having a MatchMakingRanking to face stomps and finally, recover some fun ? What do you guys think about that ?

 

To clarify, I don't mean getting a straight MMR where unicum would only face themselves. And I know MMR need something to rely on to be able to rank players by skills. Well, i'm just proposing an idea, i'm not a magician.

But, first, MM actually use nothing to mix players and make 2 teams. It's often unbalanced teams, and it's lovely to be 7 vs 11 in the first 4min of the game. I have to make sacrifices to the god of blood to get at least some good players and I run out of cats, dogs and virgins. So, to rely on something that can rank players, maybe try to get a stat like PR ? Not the same formula, but something close ? 

 

Then, in term of picking players depending on skills, maybe getting a MMR selecting in a certain pool of players corresponding to a % of actuals players ? Just for exemple, if at Time H, 60% of players have a PR between 800 and 1200, just match them. If 50% have between 1000 and 1500, match them. Etc... It would make queue a little longer. But would you wait 10s to get stomped, or would you wait 30s to get to try having the most balanced teams possible ?

 

Tho, I know WG don't care about balancing team, like they said during the Questions/Answers on discord : "deal with it, too hard to balance, no time, we are focus on soviet paperships"  (the last 6 words arent maybe true). 

But what do you, players, think about the need to get a MMR to avoid stomp and get more balanced teams ? Or how would like you to reduce stomps, if you want to ?

 

 

It's just an idea and question, asking for your thought. And sorry for my bad english. 

PS : Posting in a new thread to not monopolise the main MatchMaker discussion, because it's a suggestion too, and invite (hoping) to long discussion. Sorry if I had to post it somewhere else. 

 

 

  • Cool 3
  • Boring 4

Share this post


Link to post
Share on other sites
[-LA-]
Players
687 posts
5,152 battles

i dont think it would help ,  the stomps happen because of how much power offence has compared to defence and the cluster /camp nature of gameplay it leads to ships getting focused and deleted in far too fast a time given the speeds involved.

 

edit: the stomps are by design its an intentional choice of the devs

 

plus MMR tends to lead to everyone having a 50% win rate, and if you propose that on this forum your going to be mopping up a hurricanes worth of neurotic piss thats going to get sprayed all over the forum at even that suggestion

Share this post


Link to post
Share on other sites
[P7S]
Players
186 posts

This would not work, on top of creating severe issues with assembling teams for the match.

 

But discussing MM is pointless for other reason. WG MM is not random, and winning and loosing streaks serve the purpose known to WG. MM is the last thing they would consider to change.

  • Cool 6
  • Boring 1

Share this post


Link to post
Share on other sites
[CATS]
Players
28,107 posts
14,890 battles
43 minutes ago, Padds01 said:

edit: the stomps are by design its an intentional choice of the devs

To what purpose? Research times are shorter than years ago.

 

49 minutes ago, 13Ruan13 said:

So, what about having a MatchMakingRanking to face stomps and finally, recover some fun ? What do you guys think about that

Skill based MM could reduce the stomps a bit, but will not eliminate them.

Skill based MM should be tested in a defined ennviroment, like a League system.

 

I do not see how skill based MM could work in the random enviroment with hundreds of ship in four classes at 10 Tiers.

 

A system like TruSkill™ needs 91 matches per player for calibration just for 8 man teams. For 12 man teams, this number is higher. Then you have to add Tiers, different ships and classes.

  • Cool 1

Share this post


Link to post
Share on other sites
[POP]
Players
637 posts
12,481 battles

With equally skilled teams stomps happen more frequently than with unequally skilled. This is a real world observation. I bet someone chimes in and tells what the name for it was, I've forgotten.

Share this post


Link to post
Share on other sites
[-LA-]
Players
687 posts
5,152 battles
2 hours ago, ColonelPete said:

To what purpose? Research times are shorter than years ago.

 

Honestly after some descisions they have made i really wouldnt hazzard a guess. or even bet they have one, i just mean its design choices that are causing stomps , it would be easy enough to design them out if they were so inclined.

Share this post


Link to post
Share on other sites
Beta Tester
939 posts
10,196 battles

Yes, please.

The average potato won't notice any difference in queue time, and I don't have fun with snowball teams.

Share this post


Link to post
Share on other sites
[CATS]
Players
28,107 posts
14,890 battles
15 minutes ago, Padds01 said:

Honestly after some descisions they have made i really wouldnt hazzard a guess. or even bet they have one, i just mean its design choices that are causing stomps , it would be easy enough to design them out if they were so inclined.

How?

Share this post


Link to post
Share on other sites
[SM0KE]
Players
6,117 posts
14,594 battles
3 hours ago, 13Ruan13 said:

asking for your thought

I'm not sure that tinkering with MM would have a significant impact on the incidence of roflstomps (unless you can have some algorithm that can spot people that know to focus fire), as the tendency towards them is built into the fundamental nature of the game - whichever team can generate focused fire is more likely to trigger the snowball effect that results in a roflstomp.

 

The only way to reliably eliminate (or maybe just reduce) roflstomps is probably to change the fundamental mechanics of the game, such as by introducing respawns, or battle royale type modes (there's one of the latter running in WOT at the moment)...

  • Cool 2

Share this post


Link to post
Share on other sites
[ORCA]
Players
5 posts
6,281 battles
Il y a 2 heures, ColonelPete a dit :

To what purpose? Research times are shorter than years ago.

 

Skill based MM could reduce the stomps a bit, but will not eliminate them.

Skill based MM should be tested in a defined ennviroment, like a League system.

 

I do not see how skill based MM could work in the random enviroment with hundreds of ship in four classes at 10 Tiers.

 

A system like TruSkill™ needs 91 matches per player for calibration just for 8 man teams. For 12 man teams, this number is higher. Then you have to add Tiers, different ships and classes.

 

Goal is surely not to eliminate them, but to reduce the possibilities of happenning. 

Yes, MMR should be tested on this game. Do you mean a new game mode ?

 

 

Ships played have only the impact of class and number in random. The game seems to stop there (maybe someone who know how MM is made in this game can correct me). MM just take 1cv, 5bb, 5 cruisers and 1 DD each teams and it stops there. No matter if cruisers have radars, are CA or CL .... So I don't see how ships would influence MMR. Even more if we just talk about player skill, expect if you mean skill on particular ships. 

 

I donno TruSkill, but it's look like to match players in a straight and close MMR. If you take a % of the pool, it's just cancel the need of calibration. For exemple, if you have right now 15 000 players who have between 1200 and 1700 players, lets go, and MM just try to make both team having the same average of PR. For exemple.

 

 

 

@Vbeest"serve the purpose known to WG"  And what is this purpose ? Harvest money ?

Share this post


Link to post
Share on other sites
[CATS]
Players
28,107 posts
14,890 battles
1 minute ago, 13Ruan13 said:

So I don't see how ships would influence MMR. Even more if we just talk about player skill, expect if you mean skill on particular ships.

Some people perform well in BB A and bad in same Tier BB C, while being average in same Tier DD Z.

How should a player be matched? According to BB A, BB C or DD Z?

Share this post


Link to post
Share on other sites
Beta Tester
175 posts
10,474 battles

I think it is a good idea .... but ... 

Coming from the game industry I know how hard it is to ballance just a simple game let alone a game as complicated as WoWs. 
All these different ships, radar, DD's, Hydro, CV's. It's impossible to even ballance the same amount of radar ships in one match, let alone the rest, and than you ask this on top of that? 
This is not going to happen, it's like asking for wold peace with Trump as abassador, not going to happen. 

It's hard enough to try and ballance a 3 tier difference between ships and all these variables. And i hear you think 'How hard can it be to take also PR in account?'. 

Verry hard because if you do that you will multiply the possible MM by hundreds of times if not thousands. Ballance that .... 

Ballancing is test by trial and with adding more and more content to keep the playerbase interested you make ballancing WoWs even harder. 
The only ballane there is in the gaming world of free to play games is 'Is the effort to create content and make the game profitable'. 
Thats the only calculatible ballance there is. The rest is fiction allthough I know they do their verry best. 

Share this post


Link to post
Share on other sites
[BATES]
Alpha Tester
3,072 posts
12,027 battles

The only balance I ever wanted in the mm in Wot and Wows was one where it evened out player skill as best as possible. For example, why put a team together with predominantly less skilled players versus a team with predominantly more skilled players..why not just best effort even out the skills between the two teams. It might not solve everything but would be a damned side better than some of the grossly uneven team setups.

Share this post


Link to post
Share on other sites
[-LA-]
Players
687 posts
5,152 battles
3 hours ago, ColonelPete said:

How?

right now the offence/defence slider relative to speed is set very heavily towards offence, ie very few mistakes can be made , very little time exists for help to arrive , camping a corner is always easier than pushing and there is nothing wrong with that , but this game is so static because of how far it is pushed. 

the high damage means a loss also has a larger impact on that teams offence , its all a recipe for a game that is static and snowbally. 

 

just push it all the other way make ships less lethal lower range, higher conceals, slower rates of fire, lower pen values, nerf fires and floods more, increase speed , turning acceleration , map design , more corners to cover lles lines of sight covering multiple corners etc etc. less ships per team , more teams , resets , reinforcements, theres going to be hundreds of little mechanical tweaks you can do to push the needle over to it taking longer for ships to be killed basically. it all allows more "rounds" of combat ,lowering variance and allowing more chances for recovery.

 

basically it comes down to making each ship less lethal , raising the time to kill and lowering the time to get help 

 

or you can make big changes that acccpet the imbalance and work with it , Verdun would be a great example , similar settings for the game, but by having it asymetric and having frequent resets they make the game less passive and reduce the snowballing. obviously those specific things wouldnt work here just an example .

 

thats just off the top of my head , with a few weeks and a team im sure a lot could be hashed out.

Share this post


Link to post
Share on other sites
[THESO]
Weekend Tester, In AlfaTesters
3,469 posts
8,619 battles

I doubt this Topic will exist long cuz MM Topic is not allowed.

 

 

But I.ll post it anyways.

 

Any Type of Skillbased MM. No matter on which Stat or Formula it works.

Will always Destroy itself unless it is a Clan Team vs Clan Team or 1 vs 1 Setup.

 

Because if it works. You get Teams that are somewhat Equal.

Which means that by Default the Results of Bad Players get Buffed. The Results of Good Players get pulled down.

 

After all. That is the Target.

You want Games that are more Even.

 

And that means. Good Players will have a Harder Time to actually get Results and Bad Players will have an easier time.

 

Hence. The System Equalizes itself and then is useless cause the Results used for Matchmaking are no longer Valid due to Bad and Good Players having Similar Results due to having Faced Enemies of their own rough Quality Estimate.

 

Worse. If you make it too rough you end up Mass Creating Roflstomps.

Cause Bad Players repeatedly Matched against Bad Players suddenly get Great Results. And then get Counted as Good Players and are on Purpose Matched against Good Players even tough they are Potato.....

Likewise. Good Players which got repeatedly matched against Good Players end up with Bad Results and will be Sharks thrown into the Fishtank.....

  • Cool 2

Share this post


Link to post
Share on other sites
[CATS]
Players
28,107 posts
14,890 battles
5 hours ago, Padds01 said:

right now the offence/defence slider relative to speed is set very heavily towards offence, ie very few mistakes can be made , very little time exists for help to arrive , camping a corner is always easier than pushing and there is nothing wrong with that , but this game is so static because of how far it is pushed. 

the high damage means a loss also has a larger impact on that teams offence , its all a recipe for a game that is static and snowbally. 

 

just push it all the other way make ships less lethal lower range, higher conceals, slower rates of fire, lower pen values, nerf fires and floods more, increase speed , turning acceleration , map design , more corners to cover lles lines of sight covering multiple corners etc etc. less ships per team , more teams , resets , reinforcements, theres going to be hundreds of little mechanical tweaks you can do to push the needle over to it taking longer for ships to be killed basically. it all allows more "rounds" of combat ,lowering variance and allowing more chances for recovery.

 

basically it comes down to making each ship less lethal , raising the time to kill and lowering the time to get help 

 

or you can make big changes that acccpet the imbalance and work with it , Verdun would be a great example , similar settings for the game, but by having it asymetric and having frequent resets they make the game less passive and reduce the snowballing. obviously those specific things wouldnt work here just an example .

 

thats just off the top of my head , with a few weeks and a team im sure a lot could be hashed out.

But these are things that were always in the game. Why are they a problem now?

Share this post


Link to post
Share on other sites
Players
3,962 posts
5,268 battles

This PR thing was my idea for fixing Ranked, and PvP in general but I think people didn't like it much.

Fun part is, this is probably not so hard to add into game.

Share this post


Link to post
Share on other sites
[LEEUW]
Players
2,612 posts
10,265 battles
6 hours ago, Shaka_D said:

The only balance I ever wanted in the mm in Wot and Wows was one where it evened out player skill as best as possible. For example, why put a team together with predominantly less skilled players versus a team with predominantly more skilled players..why not just best effort even out the skills between the two teams. It might not solve everything but would be a damned side better than some of the grossly uneven team setups.

So each game I'd get a 32%wr player on my team to even the balance? Because that is basically what you're saying. 

Yeah no, thanks. Keep it random please. 

Share this post


Link to post
Share on other sites
[21DCS]
Beta Tester
283 posts

the stomps are intentional, matchmaker is set to produce shorter games. shorter games = less exp = longer grind = more likely to use premium = more money.

 

when wows started, and for about 2 years, it was rare to have any kind of one sided stomps, "snowballing" games. not because the average players were better, but matchmaker was really random.

 

and with every matchmaker tweak, with every algorithm that was surely taken from WoT matchmaker, games became shorter and more one sided.

 

2-3 years ago a game below 10 minutes was rare. 2:10, 1:11, 0:12 defeats just didnt exist. losing streaks 5+? yeah happened, but not as common as for the last 1-2 years.

 

and no, that's not tinfoilhat theory. shorter turnaround time for match relates to more money spent, because the grind gets more tedious.

 

WoT had the same "phenomenon"...it was the point when I stoppen playing it. You could tell if a game was a win/loss - in most cases - after 1 minute into the game, just by playerdeployment on the map. In WoT games lasted around 7 minutes, at the end (when I stopped playing) it was around 3 minutes...and for sure the reason wasnt a sudden influx of (more) bad players (than usual).

Share this post


Link to post
Share on other sites
[CATS]
Players
28,107 posts
14,890 battles
9 minutes ago, Kutfroat said:

the stomps are intentional, matchmaker is set to produce shorter games. shorter games = less exp = longer grind = more likely to use premium = more money.

As stated above, research times are shorter than ever. 

 

Share this post


Link to post
Share on other sites
[THESO]
[THESO]
Moderator
4,183 posts
11,264 battles
15 hours ago, 13Ruan13 said:

Hello there.

 

I'm posting this and there hoping to start a manner discussion about how MatchMaking can maybe impact the stomps.

Hello op, 

MM discussion is a hot topic so we try to keep it in one place so people can read all of the arguments & counterarguments before responding. 

This topic will be locked, I suggest you go take a look at there and join the discussion. 

 

 

Share this post


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

×