Jump to content

Combining Group And Solo Queues - 4 Week Test


1579 replies to this topic

#141 Voice of Kerensky

    Member

  • PipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 506 posts

Posted 27 April 2020 - 09:12 PM

It seems to me that this will finally kill the game.
Your matchmaker for a simple fast game is disgusting (more than half of the fights end with a score from 12-0 to 12-4), your matchmaker in this simple fast game does not know how to assemble balanced teams of equal strength. Do you want to stick in there also groups of two to 4 people? Fear God, take pity on your match maker and our nerves.
What will happen when one team gets a group of four people who play very well, and another team gets a group of four people who are called "noobs"?
Your matchmaker does not take into account anything except the PSR (far from the most objective indicator of the player's level) and the tonnage of robots. Your match maker will not be able to distinguish between a group of strong players and a group of weak players, he will divide them equally into two teams. And what will be the result of the game then?
I'm afraid the only statistics that you can collect using this test is statistics on a sharp decline in the number of players.

#142 BARRY SHlTPEAS

    Member

  • PipPipPip
  • Overlord
  • Overlord
  • 80 posts
  • LocationBrown Sea Buccaneer

Posted 27 April 2020 - 10:21 PM

Yo, don't worry guys, I'll take this one.

View PostFRAGTAST1C, on 27 April 2020 - 05:23 PM, said:

1. What steps are taken to assess the threat value of a group, like I asked you in the previous thread? How do you know the difference between some weekend group vs EMP or JGx or MJ12 or EON or other top comp group?

2. Is the MM still going to make sure that both teams have the same number of mechs across all the weight classes? What changes have you made to ensure that?

3. When are you going to implement a merit-based ranking system? How do you expect a match to be "competitive" when the 1st point can't be achieved and also while solos aren't ranked based on their skill?


1.) None and we have no clue.
2.) No and none.
3.) After we make MW5 into an excellent game of course.

Edited by -Verti-, 27 April 2020 - 10:33 PM.


#143 Snowhawk

    Member

  • PipPipPipPipPipPip
  • Survivor
  • Survivor
  • 433 posts

Posted 27 April 2020 - 10:22 PM

"2-player groups: Min Tonnage 40, Max Tonnage 200"

Steiner Scout lance (2 Atlas D-DC) Incoming.... Posted Image Posted Image

#144 Vxheous

    Member

  • PipPipPipPipPipPipPipPipPip
  • CS 2019 Gold Champ
  • CS 2019 Gold Champ
  • 3,826 posts
  • Location2 Time MWO World Champion

Posted 27 April 2020 - 10:26 PM

View PostSnowhawk, on 27 April 2020 - 10:22 PM, said:

"2-player groups: Min Tonnage 40, Max Tonnage 200"

Steiner Scout lance (2 Atlas D-DC) Incoming.... Posted Image Posted Image


This has always been the case, except on one side, you had 2 Atlas, and the other side you had 2 annihilators that murdered them at 500m

Edited by Vxheous, 27 April 2020 - 10:26 PM.


#145 Vellron2005

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The Blood-Eye
  • The Blood-Eye
  • 5,444 posts
  • LocationIn the mechbay, telling the techs to put extra LRM ammo on.

Posted 27 April 2020 - 10:45 PM

Well I can now conclusively say that I'm about at least 6-12 months ahead of the devs in thinking about what to do with this game Posted Image

When I suggested this last year, my idea got stomped on.. but.. well.. here we are. Nobody's laughing now. Told you so.

My only issue is that a group of 4, while being fun to play in, is still too strong and influential to not steer the course of a match.

What will happen is you will see a lance brake off and do it's own thing, while the 8 other pugs do some other thing. This regularly happens in Faction Play now, so it will just spill over into "The Que"..

And it will also let big units / comp players dominate the last vestige of non-comp play in the game.

For this reason, I suggest the group size be limited to 2. That's all you need to incite people play with their buddy, while not making the quick play scene look like a bad unit-fueled roflstomp fest..

#146 FRAGTAST1C

    Member

  • PipPipPipPipPipPipPipPipPip
  • The Fighter
  • The Fighter
  • 2,910 posts
  • LocationIndia

Posted 27 April 2020 - 11:07 PM

View PostVellron2005, on 27 April 2020 - 10:45 PM, said:

Well I can now conclusively say that I'm about at least 6-12 months ahead of the devs in thinking about what to do with this game Posted Image

When I suggested this last year, my idea got stomped on.. but.. well.. here we are. Nobody's laughing now. Told you so.



So, you are the one responsible for this fiasco. Posted Image

#147 Horseman

    Member

  • PipPipPipPipPipPipPipPipPip
  • The Mercenary
  • The Mercenary
  • 4,737 posts
  • LocationPoland

Posted 27 April 2020 - 11:52 PM

View Post-Tenshi-, on 27 April 2020 - 03:25 PM, said:

And for those who cry that the solo players will die out, or the groups will farm them, I want to ask, have you ever seen a sync drop in quick play? I see and very often even 3-4 players from the same tag cannot make an explicit impact. I also often watch games with a score of 12-2 among solo players, so there has never been and will never be a perfect balance?

View PostVellron2005, on 27 April 2020 - 10:45 PM, said:

My only issue is that a group of 4, while being fun to play in, is still too strong and influential to not steer the course of a match.
(...)
For this reason, I suggest the group size be limited to 2. That's all you need to incite people play with their buddy, while not making the quick play scene look like a bad unit-fueled roflstomp fest..
The reason for lack of balance is that the matchmaking algorithm is garbage.
Had the MM been working as a MM should, there wouldn't be much of an issue with merging the queues because top-tiers would be matched against top-tiers and filled out with competitive solos..
But it's not working, so the result is going to be a trainwreck.

View PostWindscape, on 27 April 2020 - 04:00 PM, said:

Its clear that there is one solution, that would fix a large majority of the problem.
THE STUPID PSR.

If we were to use a more competitive matchmaking system, match quality would be significantly improved, insuring that if PGI did merge the queues under this system, 1 90% team would be matched with another 90% team (or 80-99%, varying with population).

There is one very successful game that uses a modified GLICKO-2 system for this, CSGO. It uses combined matchmaking but uses a skill rating system that better insures that lower skilled players are still matched with their own kind, rather that the current PSR.

The people here are trying to tell you that the reason why this idea is horrible is because PSR IS HORRIBLE. Admittedly, even with a better matchmaking system, population would still be too low for me to want to recommend combining the queues, as even a better matchmaker can only work so well with the population it has, but that would be the biggest step in the right direction.

The solution is, to Change PSR to a better system, even ELO. While I prefer to have solo queue remain solo, fixing the PSR would allow Paul's proposed system to actually work to an extent, providing matches to 2 man groups while ensuring matches remain decently competitive. CSGO allows for a group of solos to face a full 5 man, and due to its matchmaker matches can still remain consistently even. Going this far in MWO though would be very, very stupid though since CSGO actually has the population to ensure that everyone is in the same skill level roughly.

TLDR: Change PSR to a ELO or GLICKO-2 system, and then your idea may work, but just dumping PSR would make the biggest difference.

So much this.

#148 Sniper09121986

    Member

  • PipPipPipPipPipPipPipPipPip
  • The Sickle
  • The Sickle
  • 2,161 posts

Posted 28 April 2020 - 12:05 AM

We have in-game VOIP and an LFG tool. We also have an upward-biased matchmaker and now groups in solo queue. If I try really hard to put all of this into a single sane context it might seem that PGI tries to actually prod players into learning the game and teaming up, so the matchmaker is simply another version of cadet bonus that lets players start at the start until releasing them to what is basically (and now officially) free-for-all.

This is the only method behind this... decision-making that I can think of, the only explanation with a hint of rationale. If that is the case, then the fact is that this train is going where the passengers absolutely do not want to go. If it is not, then I shudder to think about the processes behind this decision. WE HAVE BEEN HERE BEFORE. We know how that ended - by separating the queues. If the majority of (paying) player-base becomes alienated by the minority that will feast on them, then guess what, PGI would lose the majority of their paying customers until there are not enough people left to afford a server for. Everybody loses.

#149 Erik Krieger

    Member

  • PipPipPip
  • Star Captain
  • 77 posts

Posted 28 April 2020 - 12:42 AM

I am a pure solo player that plays A LOT on a nearly daily basis. This is the first time I thnk to quit this game. If the match experience forl MY PERSONAL TASTE will get worse, like i fear it will, i will be gone.

This change will suck for players like me who will oppose now 2-4 players who will much better coordinate against me.

Now mock against me like u want, but I am sad...

#150 ENOVA

    Member

  • PipPipPipPipPip
  • CS 2023 Top 12 Qualifier
  • CS 2023 Top 12 Qualifier
  • 126 posts
  • LocationRussia

Posted 28 April 2020 - 12:43 AM

it will be a HELL of a month!
me likie

#151 Gannycus

    Member

  • PipPipPip
  • 96 posts

Posted 28 April 2020 - 01:08 AM

View Post-Tenshi-, on 27 April 2020 - 03:25 PM, said:

Thank you for adding a group search for quick play. Such an option was sorely lacking.

And for those who cry that the solo players will die out, or the groups will farm them, I want to ask, have you ever seen a sync drop in quick play? I see and very often even 3-4 players from the same tag cannot make an explicit impact. I also often watch games with a score of 12-2 among solo players, so there has never been and will never be a perfect balance?

Therefore, I believe that this is absolutely the right decision to add a group search to Quick Play.


Have you ever see a premade mix in solo before? I do, 5-6 years ago, and it was a nightmare for solo players, best organized premade will simply stomp the enemy team most of the cases.

That was the reason for splitting solo and group queues but seem like PGI and some players (if you were here at the time) never learn from the mistakes of the past.

#152 Alreech

    Member

  • PipPipPipPipPipPipPipPip
  • Little Helper
  • Little Helper
  • 1,649 posts

Posted 28 April 2020 - 01:22 AM

View PostStuffYouFear, on 27 April 2020 - 04:03 PM, said:

Been awhile since I last posted. Thank yall for still trying new things to keep the group system going. Alot of people dont realize the easy answer for PGI is to just get rid of groups and do nothing but solo que. Can't believe they still care after 8 years of this toxic community.

It's also time to get of all that stupid Battletech stuff like Armor, Internal Structure and Heat.
How about regenerating Shields like in Halo?
Also all Mechs should be faster, no Weapons need ammo, ect...
Get rid of the big teams of 12 players, and of the big maps.

MWO should become a highly competetive Shooter like Unreal Tournament or Quake, with skill and not stuff like Mech builds being important. Posted Image

Edited by Alreech, 28 April 2020 - 01:23 AM.


#153 Brizna

    Member

  • PipPipPipPipPipPipPipPip
  • Liquid Metal
  • Liquid Metal
  • 1,363 posts
  • LocationCatalonia

Posted 28 April 2020 - 01:27 AM

View PostBrauer, on 27 April 2020 - 11:34 AM, said:



So what is the logic behind that exactly? Given there's no way for the matchmaker to effectively sort players by ability (based on the info you all have provided and experience) I'm not aware of any reason this would make games more competitive. My experience with group queue, including during the recent revival, is that it's far stompier than solo queue, and I don't see why tossing the queues together would change that (other than to place solo droppers at a disadvantage).


Your concern is correct yet player tier is not the only factor affecting competitiveness of matches. Match Maker's ability to create even games is also heavily influenced by pool of players waiting in queue, this change should help with that theoretically leading to more even games, obviously tier problem remain unaddressed.

#154 Dee Eight

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Big Daddy
  • Big Daddy
  • 6,271 posts

Posted 28 April 2020 - 01:40 AM

View PostClint Steel, on 27 April 2020 - 02:52 PM, said:


Not quite sure I know what you mean, but I do acknowledge that there are several flaws with the way PSR is calculated, aside from my criticism.


PSR is and never was a measure of skill/ability. It has always been a game time grind bar meant to fool bad players into thinking they were getting better, as it continuously raised their tier to the "elite" tier 1... and thus they'd continue to play a game they genuinely suck at. A F2P game depends on such players for the whales who actually are good and spend the most money, to have someone to get matches against... and thus encourage them to continue spending money on new trinkets.

#155 Alreech

    Member

  • PipPipPipPipPipPipPipPip
  • Little Helper
  • Little Helper
  • 1,649 posts

Posted 28 April 2020 - 01:41 AM

View PostWindscape, on 27 April 2020 - 04:00 PM, said:

People cannot get a match in GQ (2-4 man teams), due to low population.

That is only part of the problem.

In 12 vs 12 the matchmaker has to deal with 9 group sizes.
Some group sizes work only in one combination: 10+2, 9+3
In 8 vs 8 it's 5 sizes, and 5+3 / 6+2
Sizes like 11 & 7 did not work at all.

Just by the maths it's stupid to allow bigger group sizes than 4 or 8 or 12 in group play if sizes like 11, 10, 9, 7, 6 & 5 are difficult and sizes like 2, 3, 4, 8 & 12 work best.

The variable group size makes it also difficult to balance the group by the players stats (like K/D, W/L, ELO, PSR, ect...). A bigger group has a better combat value due better teamwork, some thing that can't be broken down to numbers.

IMHO a solution:
Max group size of 4, Solos in Group Queue to fill up vacant slots.
That would mean between 1 an 2 "Solo" Player per Group.
And even that can be avoided if you invite friends or use the LFG to fill up to 4 players.

If the match maker only has to deal with 3 groups of 4 vs 3 groups of 4 it's possible to consider a groups "Skill" (calculated by the player stats in the group) into matchmaking.
That would also allow the match maker to use different match sizes like 4 vs 4 or 8 vs 8 to allow quick matchmaking during times of low group numbers.

Edited by Alreech, 28 April 2020 - 01:50 AM.


#156 Sniper09121986

    Member

  • PipPipPipPipPipPipPipPipPip
  • The Sickle
  • The Sickle
  • 2,161 posts

Posted 28 April 2020 - 02:02 AM

View PostAlreech, on 28 April 2020 - 01:22 AM, said:

It's also time to get of all that stupid Battletech stuff like Armor, Internal Structure and Heat.
How about regenerating Shields like in Halo?
Also all Mechs should be faster, no Weapons need ammo, ect...
Get rid of the big teams of 12 players, and of the big maps.

MWO should become a highly competetive Shooter like Unreal Tournament or Quake, with skill and not stuff like Mech builds being important. Posted Image


Sometimes I am inclined to agree, because in no other franchise players would tolerate the antics that we have seen from PGI repeatedly over the years. Me and other BT fans here literally care too much for our own good. Maybe it is the spell of being the first BT product in a decade (as it was back then) and the only BT product for much of the time afterwards, but this spell is probably the only thing that makes some chunk of the player-base login and pay for the servers every once in a while. Perhaps with that spell lifted it would be for the free market to sort things out and probably for someone else to provide more-than-minimally-viable product... in a decade or so.

#157 ghost1e

    Member

  • PipPipPipPipPipPip
  • CS 2023 Gold Champ
  • CS 2023 Gold Champ
  • 403 posts
  • Twitch: Link
  • Location2023 World Champion

Posted 28 April 2020 - 02:18 AM

well, RIP solo queue

#158 IanDresarie

    Member

  • PipPipPip
  • Veteran Founder
  • Veteran Founder
  • 92 posts
  • LocationGermany

Posted 28 April 2020 - 02:34 AM

View PostAnomalocaris, on 27 April 2020 - 07:17 PM, said:

BTW, reading through the posts on Reddit there seems to be a belief that a select group of people were blowing up Russ' twitter to try and get this implemented. Anyone know who those people are?

View PostLarsh, on 27 April 2020 - 07:25 PM, said:

I can't say for certain, but they may have implied my post to Russ requesting an update and further details to this update they were planning. As far as I can tell, there was no prior discussion about this group queue test before last week on Twitter.

Figure I would place it here before others take it out of context as I've seen happen on Reddit.

https://twitter.com/...535815071436801

I probably belong into that group as well. I make it a point to give people with ideas I agree with some positive encouragement on whatever platform I know I can contact them. Which in this case would be Twitter for Russ and the Forum for Paul.
If someone is able to put everyone who did so as well into a group I'd be interested in learning the names in that group, 'cause I'd probably get along with everyone else in there. But there was definitely no coordinated effort to try and get this implemented, there was merely a encouragement from me, Larsh, Oldbob10025 and others to voice their opinion and put the very negative posts by certain other people in perspective.

#159 Foxylicious

    Rookie

  • Ace Of Spades
  • Ace Of Spades
  • 4 posts

Posted 28 April 2020 - 02:37 AM

Ah yes, let us put the sharks in the guppy tank. That will surely generate more matches. Wait, where did all the guppies go?

#160 Paul Meyers DEST

    Member

  • PipPipPipPipPipPipPip
  • Nova Commander
  • Nova Commander
  • 543 posts

Posted 28 April 2020 - 02:37 AM

Wow, maybe again the most opposite thing of what would make sense. Well done PGI





2 user(s) are reading this topic

0 members, 2 guests, 0 anonymous users