Jump to content

State Of Match Making - Feedback/comments


1142 replies to this topic

#481 Cael Rhythyr

    Member

  • PipPip
  • The Widow Maker
  • The Widow Maker
  • 22 posts

Posted 04 September 2015 - 01:50 PM

Group Queues as stands is a cluster f&ck, allowing smaller multiple groups to face-off against the 10-12 man diehards.  I know there's an argument for keeping the queue's as is; however there is just as valid (if not more so) an argument to limiting the group size.
I would vote to see the MM break the Pug Group Queues into two types:  1.)  2-4 man teams grouped in Queues against other 2-4 man groups; and 2.) 5-12 man teams grouped in a separate Queu to face-off against other 5-12 man teams.
Not all casual players want to be thrown into the mix against the competitive crowd (where there is 4 different groups of no more than 3 to a group dropping against an organized 10-12 man group); and honestly, the seal clubbing does absolutely nothing to promote this game for the casual players who just want to play with their friends.  Actually, the player base is sure to dwindle from such imbalanced team construction.
The 10-12 man teams would still have their own Queues, Private Lobbies and CW to flex their muscles on; and the casual crowd wouldn't be bashed into disconnecting with this game because of the imbalance.

#482 Dinochrome

    Member

  • PipPipPipPipPip
  • Survivor
  • Survivor
  • 164 posts
  • LocationSan Diego, Ca.

Posted 04 September 2015 - 01:50 PM

Hailz Mechwarriors;

Well having burned a couple of hours reading all this, I usually get some sort of general impression of what is going on, but not this time. Discounting all the personal attack posts and you never listen posts, except for the mention of 4 versus 4, nothing new.

- I agree with Kjudoon, 4 versus 4 in LFG play kind of lame unless it is designed as a stepping stone to CW (League play)

- I agree with the statement solo play has significantly improved, much more enjoyable. Russ scored with the recent changes to solo matches.

- LFG should be a bridge to 12 man team play mode, not a dedicated place for highly practiced 12 man groups to compete as some would have it.

- Lastly there is no way a small number of skilled players/ large groups should dictate what the game should be, make your suggestions, continue to play or not as you see fit. Give the developers some time to develop CW as you 12 man uber mode.

V/R Dinochrome

#483 Fire for Effect

    Member

  • PipPipPipPipPipPipPip
  • Mercenary Rank 5
  • Mercenary Rank 5
  • 583 posts
  • LocationGermany

Posted 04 September 2015 - 01:53 PM

View PostDarkblood, on 04 September 2015 - 01:45 PM, said:

Okay, this game needs to stop going in circles...

We had smaller team caps before, did it work back then? If yes, why did we change away from it? If not, why the hell would it do better now? You guys seriously need to think outside the box instead of going back and forth all the time (same thing with quirks, which are now being rolled back).

My personal experience with this was (back in the time it was around): the smaller team cap was killing the social aspect of the game. The chit-chatter between matches, the coordination during matches, all of these are why people get into TS and form UNITS. You kill that and you are killing units. They were in fact dwindling... I went through two mergers at that time, with small dying groups being swallowed by bigger ones (Warriors of Glory -> Mechronomicom -> Comstar Irregulars).

The moment the cap was removed and we had ANY SIZED GROUPS AGAIN (no point forcing twelve either) a lot of old timers actually came back. There was literal CHEERING in the channels. Now you want to take it away again as a solution.

The match making is working as well as it will for such a small pop, you get stomped by Empyreals and Jaguars now and them, but that´s part of life (those guys need matches, you know). What you guys at PGI should be worrying about is improving new player experience (if the solo matchmaking is fine, that´s where new players go) and adding some motivation to this game (read: FIX CW and make it meaningful). With more people playing the matchmaker will fix itself


I remember it dimly, it was nothing worth remembering, you are right we ALREADY had the limit to group size and it was terrible we lost many members at that time and it had accomplished NOTHING.

The whiners that were resposible for destroying the game were STILL not satisfied and again pointed their finger at anything but themself when they again lost a game....

View PostAH Vulture Vlad, on 04 September 2015 - 01:50 PM, said:

Group Queues as stands is a cluster f&ck, allowing smaller multiple groups to face-off against the 10-12 man diehards. I know there's an argument for keeping the queue's as is; however there is just as valid (if not more so) an argument to limiting the group size.
I would vote to see the MM break the Pug Group Queues into two types: 1.) 2-4 man teams grouped in Queues against other 2-4 man groups; and 2.) 5-12 man teams grouped in a separate Queu to face-off against other 5-12 man teams.
Not all casual players want to be thrown into the mix against the competitive crowd (where there is 4 different groups of no more than 3 to a group dropping against an organized 10-12 man group); and honestly, the seal clubbing does absolutely nothing to promote this game for the casual players who just want to play with their friends. Actually, the player base is sure to dwindle from such imbalanced team construction.
The 10-12 man teams would still have their own Queues, Private Lobbies and CW to flex their muscles on; and the casual crowd wouldn't be bashed into disconnecting with this game because of the imbalance.


we already had 2 of your 3 ideas and it did not work....what is an 8-11 person group supposed to do in your field of ideas by the way?

Edited by Fire for Effect, 04 September 2015 - 01:56 PM.


#484 Ramsess

    Member

  • PipPipPipPipPip
  • The Pharaoh
  • The Pharaoh
  • 106 posts
  • LocationBrossard

Posted 04 September 2015 - 02:06 PM

No its a bad idea and as it has been pointed out by others... You lost a lot of players last time you did this.
Seriously why is it so hard to freaking put 1-2 max pugs into group queues to make wait times lower.
You probably had less players leave your game back in open beta because they were unhappy ending up with groups when they were pugging then when you decided to force groups into 4 mans only.

#485 Tsula

    Member

  • PipPipPipPipPipPipPip
  • The Wolf
  • The Wolf
  • 516 posts
  • LocationNew Alavon

Posted 04 September 2015 - 02:10 PM

View PostRuss Bullock, on 04 September 2015 - 11:49 AM, said:

well looking at MM command center it looks like the last 24 hours have gone fairly well for the group queue at least from a wait point stand point.

After the Labor Day weekend I will talk with Neema regarding some of the tweaks we have been discussing.

Question: if we went back to a game mode voting system - how would you do it differently than last time? You might need a review or dig up old posts to remember. But in short it wasn't top voted mode that you got but that mode if it had 80% of the vote had an 80% chance. So you could still get one of the other modes including for example a 5% chance for one of them.

Again this would be great because it would open the door to adding more game modes, something w have been adverse to doing as it would create more buckets.


Just make it a majority vote get a match together then all players vote majority wins simple. I don't mind any other the game modes

#486 Peiper

    Member

  • PipPipPipPipPipPipPipPip
  • The Dragoon
  • The Dragoon
  • 1,444 posts
  • Google+: Link
  • LocationA fog where no one notices the contrast of white on white

Posted 04 September 2015 - 02:12 PM

View PostDarkblood, on 04 September 2015 - 01:45 PM, said:

Okay, this game needs to stop going in circles...

[a bunch of other well-thought-out stuff based on the experience us islanders had.]



Bumping your words for great justice! To add, the tutorial, I think, with killing all the AI stuff will help the new user experience immensely too. And I agree, I'd rather be stomped by CSJ and MS once in awhile than not play with my friends at all. And yeah, I remember the mergers - caused by fear and survival - and I remember the cheers too, when casual and non-masochistic players returned to the fold. Sure, CW is where competitive groups go, but even competitive guys use the public drops to test new mechs, grind C-bills, play with friends, train up new guys, and participate in weekend events.

#487 Lugh

    Member

  • PipPipPipPipPipPipPipPipPip
  • The Widow Maker
  • The Widow Maker
  • 3,910 posts

Posted 04 September 2015 - 02:25 PM

I guess PGI's memory is too short. We already tried the 4 man limit in group queue and it DID NOT WORK.

I predict it will not work here either.

#488 Scandinavian Jawbreaker

    Member

  • PipPipPipPipPipPipPipPip
  • Overlord
  • Overlord
  • 1,251 posts
  • Twitch: Link
  • LocationFinland

Posted 04 September 2015 - 02:40 PM

Playing whole evening in the group queue. It's a joke. No matchmaking whatsoever, or so it seems. The solo queue provides much better matches in quality. I understand that people want to play in big groups - that's fine. Ýou really need to look in the mirror to your selfish eyes and think that is it really worth it.

#489 Clay Pigeon

    Member

  • PipPipPipPipPipPipPipPip
  • Mercenary Rank 3
  • Mercenary Rank 3
  • 1,121 posts

Posted 04 September 2015 - 03:04 PM

Retroactively cap unit sizes instead. Max pilots per unit == 100.
Freeze new invites to the units over the cap, and display offline timers so the people with the power to remove from unit can remove inactives.

While this will help with CW ghost drop zerg rushes, it won't really help much with group queue pubscrub games, but who cares about those? The only way to fix them is to increase player count, which means more content, and maps that don't always involve ring around the rosie around the citadel/tryhardmountain/mount doom/the caldera in autistic valley. Why make a map 100% larger if everyone just circle jerks around the obvious terrain feature?

Edited by Clay Pigeon, 04 September 2015 - 03:10 PM.


#490 Nainko

    Member

  • PipPipPipPipPipPipPip
  • IS Exemplar
  • IS Exemplar
  • 815 posts

Posted 04 September 2015 - 03:06 PM

I do not support the 4 mech group queue idea. It caused a lot of frustration to the players in our unit and some players stopped playing because of that. But what about the compromize and limit it to 8 mechs?

But I support that the 12 mech groups get a queue of their own, but without any ELO thing. Who comes into the 12 mech group has to fight against anybody they get.

Which means, we need a 12 mech group matchmaker.
For this I would suggest to extend the overlay that shows the percentages of mech classes used by a total number of 12 mech groups PLAYING & SEARCHING per Server. If aTeam searches for NA & EU the it increases both numbers by one.

Edited by Nainko, 04 September 2015 - 03:24 PM.


#491 Big Tin Man

    Member

  • PipPipPipPipPipPipPipPip
  • Rage
  • Rage
  • 1,957 posts

Posted 04 September 2015 - 03:13 PM

View PostRuss Bullock, on 04 September 2015 - 11:49 AM, said:

well looking at MM command center it looks like the last 24 hours have gone fairly well for the group queue at least from a wait point stand point.

After the Labor Day weekend I will talk with Neema regarding some of the tweaks we have been discussing.

Question: if we went back to a game mode voting system - how would you do it differently than last time? You might need a review or dig up old posts to remember. But in short it wasn't top voted mode that you got but that mode if it had 80% of the vote had an 80% chance. So you could still get one of the other modes including for example a 5% chance for one of them.

Again this would be great because it would open the door to adding more game modes, something w have been adverse to doing as it would create more buckets.


Personally I didn't have an issue with the previous preference/voting system, but I also run all 3 game modes checked most of the time anyway. Again, if eliminating player hard selection of game mode leads to the development of MORE GAME MODES OF WIDER VARIETY, I'm all for it 100%. We need more game modes besides deathball, deathball with one base and deathball with 5 bases.

My wish for a voting system is what I had suggested this previously, but will do so again since you asked; provide a vote as follows once the groups are matched:

Game Mode A on Map X
Game Mode B on Map Y
Random

Simple majority vote in the lobby wins. Your vote acts as the ready button, cannot change vote once it is cast.

#492 Fire for Effect

    Member

  • PipPipPipPipPipPipPip
  • Mercenary Rank 5
  • Mercenary Rank 5
  • 583 posts
  • LocationGermany

Posted 04 September 2015 - 03:13 PM

View PostNainko, on 04 September 2015 - 03:06 PM, said:

I support the 4 mech group queue idea.



we already had that and it did not work....

#493 Scandinavian Jawbreaker

    Member

  • PipPipPipPipPipPipPipPip
  • Overlord
  • Overlord
  • 1,251 posts
  • Twitch: Link
  • LocationFinland

Posted 04 September 2015 - 03:15 PM

View PostFire for Effect, on 04 September 2015 - 03:13 PM, said:



we already had that and it did not work....

For 12DG, 228 and other bloated units.

#494 Big Tin Man

    Member

  • PipPipPipPipPipPipPipPip
  • Rage
  • Rage
  • 1,957 posts

Posted 04 September 2015 - 03:17 PM

View PostClay Pigeon, on 04 September 2015 - 03:04 PM, said:

Retroactively cap unit sizes instead. Max pilots per unit == 100.
Freeze new invites to the units over the cap, and display offline timers so the people with the power to remove from unit can remove inactives.


What would this fix? Mixed units within the same faction can still group up in CW, so now there would be MS-1, MS-2, MS-3 and MS-4 on the same team instead of just MS.

#495 Nainko

    Member

  • PipPipPipPipPipPipPip
  • IS Exemplar
  • IS Exemplar
  • 815 posts

Posted 04 September 2015 - 03:19 PM

View Postugrakarma, on 04 September 2015 - 03:15 PM, said:

For 12DG, 228 and other bloated units.


228th and 12DG represent together 400 active MWO player. So I think our voice should be heared and has the same value as 400 single player which have the opposit meaning. But please do not call us bloated.

#496 Tiamat of the Sea

    Member

  • PipPipPipPipPipPipPipPip
  • Guardian
  • Guardian
  • 1,326 posts

Posted 04 September 2015 - 03:24 PM

View PostRuss Bullock, on 04 September 2015 - 11:49 AM, said:

well looking at MM command center it looks like the last 24 hours have gone fairly well for the group queue at least from a wait point stand point.

After the Labor Day weekend I will talk with Neema regarding some of the tweaks we have been discussing.

Question: if we went back to a game mode voting system - how would you do it differently than last time? You might need a review or dig up old posts to remember. But in short it wasn't top voted mode that you got but that mode if it had 80% of the vote had an 80% chance. So you could still get one of the other modes including for example a 5% chance for one of them.

Again this would be great because it would open the door to adding more game modes, something w have been adverse to doing as it would create more buckets.


I'll be honest, I don't recall the game mode voting system happening at all. Possibly because I've always taken whatever I'm driving into 'any' mode regardless- For all that some players will whine about Conquest being no good in an Assault-class or how Skirmish is terrible or what have you, I like to challenge myself and stretch my experiences. I've had some very good Conquest games in slow Assault-class 'mechs, for instance.

That said, I don't think voting is a good idea. What's going to happen, more often than not, is that players who frequently wind up in modes they didn't vote for will jump ship at the start of the match- which is not good for their teammates.



Frankly, with the removal of the turrets from Assault mode, I don't see that there's much point to having both Assault and Skirmish- players in either mode can and will get upset for various reasons relating to that mode not being the other mode. Pick one and keep it (probably Assault), and either replace the other with a mode that actually has significantly different goals (get more than half your team from X to Y while the opposition tries to stop you doing so, destroy these two/three targets that are too far apart to reach with the whole team while the opfor tries to keep all three intact, etc.) or just have two modes for the PSR queue.

As things stand right now, even Conquest is iffy on how different it is, since it can be won straight-up by blowing up the entire enemy team. It would stand out more as a different mode if destroying the entire enemy team didn't end the match then and there- players would care more about capping during the match, and a number of matches would come out a win for the 'mechless team via resources.

If the game modes available become more divergent, then it encourages players to either broaden their mindset or think more about what's required differently for the smaller variety of game modes they're willing to engage in, which should (with some work on rewards and such) encourage better 'mech diversity, which is part of what the whole 3/3/3/3 thing was about to begin with.



On an additional note, concerning the multiple target location destruction concept, it stands to reason that the mode would have to grant completion based bonuses at the end of match, rather than declaring a straight up win/loss. This would also encourage the players to reach for as much success as they can, even if that amount isn't 'all of it'.

It would be a good starting point for some CW ideas as well, since the result of a battle typically doesn't consist only of a win/loss, but also things like territory gained and resources spent.

View Postugrakarma, on 04 September 2015 - 03:15 PM, said:

For 12DG, 228 and other bloated units.


...or for anyone who had more than three friends that play the game.

#497 Todo Sandybanks

    Member

  • PipPipPipPipPipPip
  • Talon
  • Talon
  • 269 posts
  • Location....

Posted 04 September 2015 - 03:42 PM

Instead of going back to that crappy version of grouping, you guys should make it 8v8 again. or make 8v8 an option along with 12v12,

#498 Gnatter

    Member

  • PipPipPipPipPip
  • The CyberKnight
  • The CyberKnight
  • 140 posts

Posted 04 September 2015 - 03:46 PM

How much has the solo queue benefited from players that would normally play in groups but stopped doing so because of the quality of the group queue? How will solo queue be affected once the group queue has the same quality as the solo queue? A lot to read here, but not sure this has been mentioned.

#499 BerserX

    Member

  • PipPipPipPipPipPip
  • The Hammer
  • The Hammer
  • 424 posts
  • LocationHere

Posted 04 September 2015 - 03:52 PM

Making a maximum group size of 4 people with a 1/1/1/1 format is not a good idea. Wait times aren't that big of a deal - so someone has to wait 3 minutes to get a match, "Oh my gosh, that's three minutes I'll never get back!" The wait times are that bad, and as many option as players have (those options are what make the game so enjoyable for casual gamers like myself) available to select from (group size, server, gamemode, etc.) why would you want to restrict players?

Not placing more restrictions on players is an even more valid argument when the upcoming Steam release is considered. Theoretically, we are about to experience an influx of players from the Steam community who haven't necessarily experienced the awesome amounts of personalization available in MWO (not just MM, but mech paint and paint schemes, weapon loadouts, even engine sizes for crying out loud). When personalization/customization is a major selling point, it should not be sacrificed for "faster queue times".

The skill imbalance may be a bit of an issue, but that shouldn't dictate group sizes. My group (The Brethren Periphery Pirate Mercenary Corp) is by no means competitive. We frequently have 6+ size groups on in the evenings (varying time zones), and even when we have a 12-man group intending to be "feared", we oftentimes lose to smaller groups. Group size does not dictate skill level! (just to make sure that previous assertion is invalidated). Nerfing (yes, this is a "nerf" we're talking about here) group sizes is a band aid solution to rofl-stomps. With our Steam debut around the corner, we don't want anymore band ****.

A better solution would be to allow MM to change battle sizes from 10-12 players to accommodate skill discrepancies. It is well known by anyone who has played for any substantial amount of time that a 1-2 man lead can (doesn't necessarily always) make a humongous difference in the tide of a battle.

If MM could say:

"Team A has an overall skill rating of 4500 and 9 players, and Team B has a skill rating of 5300 and 7 players. An 800 skill point discrepancy exists. Team B needs 3 more players to satisfy minimum group size requirements. I should find a group with a skill rating lower than 1500 (3 times the average skill level of a Team A member) and at least 3 players for Team B, while minimizing Team B's group size. Team A needs 1 more player to satisfy minimum group size requirements, but a significant skill discrepancy exists; therefore I should maximize the Team A group size as best as possible, and find a group with an average player skill rating of at least 700 (just under the average player skill level in Team B)."


This method - though bulky in appearance - would provide a more balanced solution that would enable players to keep their freedom to choose the server, gamemode, and weight class they want to play. Allowing MM to make decisions based on the overall team sizes and skill ratings - instead of just individual group sizes - and allowing it to alter a more heavily-skilled team's size (within the limits of a couple mechs) to help balance both teams' overall skill ratings is a much better solution prior to a Steam debut.

**NOTE: I am aware that the information presented in my theoretical MM scenario may not be accurate or reasonable. I don't know the actual values that PGI uses, but was just posing a hypothetical major team imbalance.



In short:

DON'T PUT A BAND AID WHERE IT DOESN'T BELONG! USE AN ALTERNATIVE METHOD THAT DOES NOT MESS WITH A MAJOR FEATURE THAT ATTRACTS PEOPLE TO THE GAME!


Also, allowing MM to choose team sizes based on skill levels can be extended to comparing the mech values (that have been talked about for rebalance purposes) of each team to better balance a match. It can also be extended to canonical 12v10 IS v Clan battles.

#500 Kaeseblock

    Member

  • PipPipPipPipPipPip
  • WC 2018 Top 12 Qualifier
  • WC 2018 Top 12 Qualifier
  • 258 posts
  • LocationEU / Deutschland

Posted 04 September 2015 - 04:09 PM

View PostRuss Bullock, on 04 September 2015 - 11:49 AM, said:

Question: if we went back to a game mode voting system - how would you do it differently than last time? You might need a review or dig up old posts to remember. But in short it wasn't top voted mode that you got but that mode if it had 80% of the vote had an 80% chance. So you could still get one of the other modes including for example a 5% chance for one of them.


Tough question. It's really nice to choose your game mode, e.g if you just need some few games of a single game mode to complete a challenge. So I'd probably go for a slider for each game mode where you can weight the game modes against each other. Groups could either consider the weights chosen by the group leader only or the average normalized weights of all players. Tbh I don't know which option is better here.

The top voted mode among the groups in a match is the game mode the match will be played in.

The MM decides to kick off matches depending on the votes of the players in a possible match and their wait time in the queue.

So leanging towards a specific game mode would basically be a tradeoff: Equal weighting of game modes means shorter wait times. Leaning heavily towards one game mode increases the chance of playing this mode but will probably increase wait times.

Some examples:

The Sliders

Let's say the weight sliders go from [1] to [10]. ( Note: Assigning a weight of 0 to a game node is not possible. It's important for the player to see this offset on the slider. The exact values assigned to the slider would be hidden --> tunable by developer if necessary.)

Casting your vote

A player/group chooses he wants/they want to play Skirmish only. The resulting weights would be:
Assault: 1
Conquest: 1
Skirmish 10

Scaled with the total number of points distributed the normalized weight for all game modes are:
Assault: 1/12 = 8.3%
Conquest: 1/12 = 8.3%
Skirmish: 10/12 = 83.3%

Like this, the players maintain a good amount of visible control about which gamemode they like to play and which not without limiting the MM too much.

Finding a match

For simplicity let's assume we are in the group queue.

Team 1:
Group 1: 8 players, normalized Weights: Assault: 50% = 0.5, Conquest 25% = 0.25 , Skirmish 25% = 0.25
Group 2: 4 players, normalized Weights: Assault: 20% = 0.2, Conquest 10% = 0.1, Skirmish 70% = 0.7

Team 2:
Group 1: 12 players,normalized Weights: Assault: 60% = 0.6, Conquest 10% = 0.1 , Skirmish 30% = 0.3

Computed over all players the game mode weights would be:
Assault: 8*0.5 + 4*0.2 + 12*0.6 = 12
Conquest: 8*0.25 + 4*0.1 + 12* 0.1 = 3,6
Skirmish: 8*0.25 + 4* 0.7 + 12*0.3 = 8,4

The resulting game mode would be the top voted mode. So the game mode for the match would be Assault.

The MM will then take the normalized weights of the groups to decide if he starts the match right away or leaves the players in the queue a bit longer to find a better match for them. E.g. Group 2 in Team 1 doesn't like to play Assault. So the MM might wait a bit to look for other players that like to play Assault. If the wait time gets too long, it drops the players into the match.


To be honest: I have no idea if this works or would just inflate wait times. So I'd make sure to run a simulation with past queue data before dropping this onto the players. But the idea described above would be my prefered way on tackling the voting problem.


Oh, and on a sidenote:
Please don't take groups of any size away. I was really happy when my friends and I could finally drop in groups of 5 or 6 and nobody had to be the odd man out playing alone anymore. I could live with restrictions on Mech choice in groups though :)





9 user(s) are reading this topic

0 members, 9 guests, 0 anonymous users