Jump to content

Group Queue Update 2020


297 replies to this topic

#121 Sniper09121986

    Member

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

Posted 22 April 2020 - 08:58 AM

View PostHorseman, on 22 April 2020 - 08:44 AM, said:

Adjusted PSR thresholds - yes, we needed that for years. Possibly different change thresholds based on your current tier.
As for tier realignment... either dramatically increase the magnitude of PSR changes for a while or just go with a global PSR reset. That or have PSR decay over time - however it's done, rising above T3 should require effort and commitment, and staying in T1 should require consistent performance.
Currently we have 38-percenters who were floated to T1 by the PSR algorithm - that negatively impacts matchmaking for EVERYONE, them included. Playtime isn't indicative of experience, and players who can't cut it in higher tiers SHOULD be able to drop down in tiers so that they're matched with oppositon that's more on their level.


I mean, there's a solution for merging the queues that's already in place in FP and that in theory could work in QP - bulk out both teams with groups and then balance with highest-performing solos.

The problem is that currently QP matches based on the tier system, and the upward bias in PSR calculations makes it so that as much as 60% of the playerbase makes it to T1. That would have to be fixed before we can talk about any meaningful matchmaking in QP.


Maybe not decay but periodically reset the PSR for everyone. That way everyone gets placed at their bracket and stays there instead of taking a break for seal clubbing after his PSR sinks down. Pretty much what we have in S7 already, and it has been suggested before they implemented tier system in the first place.

#122 PotatoCrunch

    Member

  • Pip
  • Ace Of Spades
  • 15 posts

Posted 22 April 2020 - 12:00 PM

Ah I should add:

If 8v8 group queue is the final decision, I think group sizes for that mode should be limited to 2 or 3.
I personally think 12v12 combined queues is better considering that if matchmaking takes as long as you say for 8v8, it won’t be worth waiting 6.5 minutes for a short game. I know for sure that my friend won’t be interested in 8v8 for those queue times as well as potentially getting stomped by a good group.

Basically, I think that something HAS to change, and 12v12 combined queues is worth a try.

Edited by PotatoCrunch, 22 April 2020 - 12:22 PM.


#123 Agent Super Chicken

    Member

  • PipPipPip
  • The Widow Maker
  • 70 posts

Posted 22 April 2020 - 12:55 PM

I'm excited to experiment with a new system when it's ready.
While doing some adjusting, will the group finder make an effort to match comparable groups?
If a group of 4 truly tier-zero players all get together on the same team (like, not just tier 1, but every member 99th percentile and hugely imbalanced KDR and win/loss ratio), and a group of 4 tier-3 players are on the other team, will that be the end of the balancing effort?
Or will there be some kind of combat capability score be generated from the overall quality of the members of each group, to try to match groups of comparable skill - or at the very least, match a hugely capable group of 4 with 8 solos in opposition to a coordinated 3 groups of 4, to give them a fighting chance?
Without such a thing, that works at least a portion of the time, I could see this being detrimental to enjoyment of solo queue when the wrong circumstances arise.
Otherwise, I think it's a great idea and I look forward to the end result.

#124 Sniper09121986

    Member

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

Posted 22 April 2020 - 12:56 PM

View PostPotatoCrunch, on 22 April 2020 - 12:00 PM, said:

Ah I should add:

If 8v8 group queue is the final decision, I think group sizes for that mode should be limited to 2 or 3.
I personally think 12v12 combined queues is better considering that if matchmaking takes as long as you say for 8v8, it won’t be worth waiting 6.5 minutes for a short game. I know for sure that my friend won’t be interested in 8v8 for those queue times as well as potentially getting stomped by a good group.

Basically, I think that something HAS to change, and 12v12 combined queues is worth a try.


Group queue is specifically intended for group action of all sizes, so no need to turn it into another solo queue, only for small groups. Not sure if the LFG tool allows for groups to search for another group in order to merge with it, but a better solution for your buddy might be if one of you finds a group in LFG and then simply invites the other one into that same group. That way the two of you will have more players to coordinate with and the other guys can fill up the group faster. Sounds like win-win to me.

View PostAgent Super Chicken, on 22 April 2020 - 12:55 PM, said:

I'm excited to experiment with a new system when it's ready.
While doing some adjusting, will the group finder make an effort to match comparable groups?
If a group of 4 truly tier-zero players all get together on the same team (like, not just tier 1, but every member 99th percentile and hugely imbalanced KDR and win/loss ratio), and a group of 4 tier-3 players are on the other team, will that be the end of the balancing effort?
Or will there be some kind of combat capability score be generated from the overall quality of the members of each group, to try to match groups of comparable skill - or at the very least, match a hugely capable group of 4 with 8 solos in opposition to a coordinated 3 groups of 4, to give them a fighting chance?
Without such a thing, that works at least a portion of the time, I could see this being detrimental to enjoyment of solo queue when the wrong circumstances arise.
Otherwise, I think it's a great idea and I look forward to the end result.


Legit concern. The problem with this group PSR though is that groups are created by different players and disbanded all the time, so it would be nearly impossible to evaluate so many people with any accuracy. A similar system exists for S7 duos, but I have yet to see any feedback about its effectiveness.

Edited by Sniper09121986, 22 April 2020 - 01:02 PM.


#125 Brauer

    Member

  • PipPipPipPipPipPipPipPip
  • 1,066 posts

Posted 22 April 2020 - 02:55 PM

View PostAgent Super Chicken, on 22 April 2020 - 12:55 PM, said:

I'm excited to experiment with a new system when it's ready.
While doing some adjusting, will the group finder make an effort to match comparable groups?
If a group of 4 truly tier-zero players all get together on the same team (like, not just tier 1, but every member 99th percentile and hugely imbalanced KDR and win/loss ratio), and a group of 4 tier-3 players are on the other team, will that be the end of the balancing effort?
Or will there be some kind of combat capability score be generated from the overall quality of the members of each group, to try to match groups of comparable skill - or at the very least, match a hugely capable group of 4 with 8 solos in opposition to a coordinated 3 groups of 4, to give them a fighting chance?
Without such a thing, that works at least a portion of the time, I could see this being detrimental to enjoyment of solo queue when the wrong circumstances arise.
Otherwise, I think it's a great idea and I look forward to the end result.


Tiers means very little when tier one players can be in the 99.99th percentile or in the 60th percentile, so it'll be stompy as group modes have always been in this game.

#126 PotatoCrunch

    Member

  • Pip
  • Ace Of Spades
  • 15 posts

Posted 22 April 2020 - 03:23 PM

View PostSniper09121986, on 22 April 2020 - 12:56 PM, said:


Group queue is specifically intended for group action of all sizes, so no need to turn it into another solo queue, only for small groups. Not sure if the LFG tool allows for groups to search for another group in order to merge with it, but a better solution for your buddy might be if one of you finds a group in LFG and then simply invites the other one into that same group. That way the two of you will have more players to coordinate with and the other guys can fill up the group faster. Sounds like win-win to me.


I agree with your sentiment regarding that group queue should be intended for groups of all sizes. I suppose that the problem with groups is more of a problem with skill-based matchmaking. I don't know how difficult it is to tweak it, since I'm not sure what the gap is between like 4 great players who aren't in a unit together versus 4 great players who are in a unit together.

Regarding LFG, we've tried it before and it didn't get us anywhere into any match. Even if we could get into matches, I personally would rather play and have voice comms with people I know well, I'm not particularly interested in grouping up with randoms regardless of how good they are. Coordination isn't nearly as much of a priority to me than simply having fun by playing with people who I like being around. This isn't to say that I don't want to win, it's more like, I want to have a balance between trying to win and actually having fun.

#127 Kamikaze Viking

    Member

  • PipPipPipPipPipPip
  • 384 posts
  • LocationStay on Topic... STAY ON TOPIC!!!

Posted 22 April 2020 - 05:02 PM

So after a few days of this it I'd like to address Pauls OP. and bring everything back on topic.

Primary options:

1) Combining Solo and Group queue. - there is a significant split in the community and it seems 50/50 on this. I'm personally not a fan

2) Max group size of 4 - IF you combine the queues then yes limit it to max group of 4 (preferably less)

3) Group team balancing - 100% required even if no other changes are made

4) Experience/Economic Balancing - yes required if group queue goes to 8v8

Alternate option:

5) 8v8 Group Queue - this seems to be the most popular option.

6) Solo Opt in to Group Queue - This seems very popular and doesn't seem to have a downside.

Community Requested Priority:

7) Fixing PSR - it seems that many people agree that removing the upward bias, and some form of Tier redistribution or periodic reset is a good idea. And that NOT doing this will undermine any other changes you make, as just changing the queues around doesnt improve the quality of matches. If the foundation the MM/Tier/PSR system is built on innacurate skill measurement, then the the results of the MM are always going to be a mess.

Cheers and Good luck.

Edited by Kamikaze Viking, 22 April 2020 - 05:06 PM.


#128 Laser Kiwi

    Member

  • PipPipPipPipPipPip
  • Leutnant-Colonel
  • Leutnant-Colonel
  • 271 posts

Posted 22 April 2020 - 05:39 PM

View PostdenAirwalkerrr, on 20 April 2020 - 07:05 PM, said:

It's just average between 3-5min on NA primetime and 20-30 mins of just wait time whoever outside of it tried searching and never got a drop.

Oh I also will be down for some top tier seal clubbing after groups get merged.


Well as long as there is 2 seal clubbing groups and they are on opposite sides no worries, and also whenever you are in a group right now, you club seals, and whenever you drop alone in tier 1, plenty of seals to club......so in reality it won't matter what they do, elite players such as yourself will always be elite.

#129 Larsh

    Member

  • PipPipPipPipPipPip
  • The Lanner
  • The Lanner
  • 272 posts
  • Facebook: Link
  • Twitter: Link
  • Twitch: Link
  • LocationYinz all going to EnP at PGH n'at?

Posted 22 April 2020 - 05:39 PM

After sifting through this topic, I think the best idea stated here is:

FIXING THE PSR.
I'll be the first to admit, I'm Tier 1, but in no way should I belong there. I always mention that my fondest time playing MWO was back when I was around T3 - T2. Varied mech loadouts, and various strategies were seen without people getting upset over it. But, that is beyond the point really.

If the Devs don't test their current suggestion, then that is where we should start.

I think if the Devs are smart enough, they will notice the discussion here in the forums, and realize how much of an important issue this is to the community.

And if they do test it, and they do not use the PTS, they better back up any word they use if it doesn't work properly and reverse the motion after their "scheduled" test period.

#130 Nearly Dead

    Member

  • PipPipPipPipPipPip
  • Philanthropist
  • 274 posts

Posted 22 April 2020 - 06:00 PM

If teams come into QP with their own private comm channels and dominate the gameplay by treating their own solo teammates as cannon fodder, the lights will be out in a few months and PGI can move on to whatever their next gig is.

#131 Alloh

    Member

  • PipPipPip
  • The Privateer
  • The Privateer
  • 60 posts

Posted 22 April 2020 - 06:01 PM

Finally!!! Thank you !

But why not a mid-term, groups of 6 ?

Six is better than 4, less than 8... but a very good number of friends playing as a Unit!

#132 Laser Kiwi

    Member

  • PipPipPipPipPipPip
  • Leutnant-Colonel
  • Leutnant-Colonel
  • 271 posts

Posted 22 April 2020 - 06:07 PM

View PostMisterSomaru, on 21 April 2020 - 12:29 AM, said:

guess it's time to farm the people asking for this for years to remind them how wrong they are once again.
As usual.


If there are only 4 farmers grouped up at one time then the game is already dead. I can see you all rubbing your hands together as you plot your 4 man groups (and then complaining about how **** everyone is), but, you go out farming and lo, if the MM could possibly be reassigned by matchscore, we would see Vxheous v Somaru, not Vxheous AND Somaru. Farming is easier on say Faction because you pick a side, in the group queue you would "prearrange group sizes to maximise tonnage" ie drop 4 groups of 3 with max tonnage to screw a couple of 6'ers or a 12'er. As long as there are at least 2 competent groups running at the same time it should hopefully balance out. Or at the very least make no difference as i've been on both ends of a 12-0 stomp in regular QP anyway.

Thinking about MrSomaru, say you have 8 WKA boys, pretty good top tier players, top 8 in team all well in the top 6% of players by score and at the highest tier. By rule, those 8 players would be split into 2 opposing sides if they played at a the same time. You can't exactly farm yourself now can you?

#133 Laser Kiwi

    Member

  • PipPipPipPipPipPip
  • Leutnant-Colonel
  • Leutnant-Colonel
  • 271 posts

Posted 22 April 2020 - 06:14 PM

View PostNearly Dead, on 22 April 2020 - 06:00 PM, said:

If teams come into QP with their own private comm channels and dominate the gameplay by treating their own solo teammates as cannon fodder, the lights will be out in a few months and PGI can move on to whatever their next gig is.


Solo players need to learn how to avoid that, the easiest way is this, don't rush off ahead into the middle just to die like normal QP, if a group hangs back, then that is the group you want to hang with. Take advantage of their superior aim and ability and pick up the scraps. I actually get annoyed at the way solo players string themselves out, its not even nascar that bothers me its the mindless rush to the same damn place every damn match, and this is tier 1 i'm talking about, where all us mechdads end up after a couple of years.

#134 justcallme A S H

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • CS 2020 Referee
  • CS 2020 Referee
  • 8,987 posts
  • LocationMelbourne, AU

Posted 22 April 2020 - 06:25 PM

View PostBrauer, on 22 April 2020 - 02:55 PM, said:


Tiers means very little when tier one players can be in the 99.99th percentile or in the 60th percentile, so it'll be stompy as group modes have always been in this game.


60%??

That's generous.

Players in the 30% range are well known to have "made it" and acquired that coveted Tier 1 marker.

#135 Laser Kiwi

    Member

  • PipPipPipPipPipPip
  • Leutnant-Colonel
  • Leutnant-Colonel
  • 271 posts

Posted 22 April 2020 - 06:28 PM

Finally, (i promise), i'm tier 1. My bar is maxed out. It doesn't matter how **** i play, i'm gonna be tier 1 forever and ever, no way to change that, i could play 500 sub 100 match score matches and i still wouldn't be T2. What the hell is up with that. No way a psr based on tier is going to work. i've played against the good players in faction and in QP and the difference between good and average T1 players is obvious and startling.

Oh Ash must have been typing at the same time, what he said X2

Edited by Laser Kiwi, 22 April 2020 - 06:28 PM.


#136 Larsh

    Member

  • PipPipPipPipPipPip
  • The Lanner
  • The Lanner
  • 272 posts
  • Facebook: Link
  • Twitter: Link
  • Twitch: Link
  • LocationYinz all going to EnP at PGH n'at?

Posted 22 April 2020 - 06:34 PM

Also, can the idea of solo pugs being scraped off the bottom of the elite's shoe's be scrapped? When solo players see this mantra it really is discouraging and demeaning.

MWO is a team game after all. If there are elite players that are looking to stomp other teammates, can we agree that they should be the champions of us and lead us into glorious, bloody battle? I know when I try comms people tend to listen.

Edited by Larsh, 23 April 2020 - 04:07 AM.


#137 Tyrsen

    Member

  • PipPip
  • The Altruist
  • The Altruist
  • 27 posts

Posted 22 April 2020 - 07:19 PM

Step 0: Keep Group Queue and Solo Queue separate. Set Groups to 8-max or 12-max, whatever works .

Step 1: Opt ALL players (except maybe Cadets?) into a 'Group Queue Reinforcement Pool', TELL THEM IN-GAME (ffs stop making major announcement on the forums and website that 80%+ of your players never look at). Give them an option to EASILY opt out at any time and EASILY opt back in at any time, and make it REALLY obvious in-game whether they're currently opted in or out.

Step 2: Use the 'Reinforcement Pool' to let MM add up to 2 ungrouped players to form a 8-man team or 3 ungrouped players to form a 12-man team (depending whether you set Group Queue to 8-max or 12-max). Tune the numbers as needed. Here's the important part: make it REALLY obvious when a match starts whether it's a group match or a solo match so they can make an informed decision about staying opted in.

Step 3: If a player's in solo queue, then it's because they don't want to be in group queue for WHATEVER reason. Incentivize being opted in to the 'Reinforcement Pool', in the form of a rewards multiplier for ungrouped players pulled into Group queue by MM. Set the multiplier low (like 1.2x) and increase/decrease as needed until your opt-in ratio is at a level that keeps MM sufficiently fed. Make sure the current multiplier is REALLY obvious in the UI next to that opt-in/opt-out toggle.

Step 4: Watch the queue size problem solve itself, with the only seals getting clubbed being the aquatic BDSM crowd.

And here's the MOST important part. If you do nothing else, do this:

Step 5: For the love of god, START SUSTAINABLY MONETIZING THE GAME. There are countless mobile games that do this better than you are. Games that sell 5$ monthly packs that do nothing more than slightly boost resource gain over the month get a guaranteed minimum income to run the servers, because a large chunk of modern players are generally willing to pay 5$ a month for a game they enjoy. They also have more expensive cash shop items designed for dolphins (who spend in the 10-50$ range) and then there are the obligatory whale-level items for players with bursting wallets (usually cosmetics, because 90% of players don't care how much whales spend on cosmetics in a Free-to-play game. The best money-makers are the ones with a sliding scale of how much a player spends each month with increasing perks as they spend more.

It's not difficult. Here's an easy example if you need ideas. Have a 5$ 'resource booster' that runs for one calendar month (e.g. buy on the 5th of this month, and it lasts until 4th of next month). Have it boost resource gain the way Premium Time does (25% boost maybe? Just pick a number your Finance person is happy with). But wait, there's more! Sell multiple booster sizes. If your base booster is 20% (for example) offer a 50% booster at 10$ and a 120% booster at 20$ (bigger booster = better ROI-per-dollar for the buyer = better revenue for seller). Make it so only one booster can be active at a time (so they can't be stacked). When the booster runs out in a month's time, buyer can get another booster. You might even give them an an email reminder when it's a few days away from running out. Discount boosters at launch and then again once every 3-4 months. And consider making the boosters obvious on the pre-match and post-match screens so people know who's contributing to the game and how much. There's nothing like seeing an x% earnings multiplier tag next to a player's name to make newer players realize they can speed up the cbill/xp grind.

Yes, I get that Premium Time already exists. Stop trying to sell one-off products to maximize up-front revenue (You still have 12-month premium time blocks in an out-of-development game? Really?) Start focusing your attention where the easy money is, selling monthly renewable products to support stable ongoing revenue, and make it EASY for the buyer. I'm willing to bet that a lot of long-term players will buy monthly boosters just to keep the servers and support going so the game doesn't fall off a cliff, and a not-insignificant share of new players to stick around because there's an easy and obvious way to trade dollars for time.

Worst case scenario, the game is financially sustainable for longer. Best case scenario, you make enough money to start funding developers and creating new cosmetics to sell to the whales. What do you have to lose?

Edited by Tyrsen, 22 April 2020 - 07:21 PM.


#138 Larsh

    Member

  • PipPipPipPipPipPip
  • The Lanner
  • The Lanner
  • 272 posts
  • Facebook: Link
  • Twitter: Link
  • Twitch: Link
  • LocationYinz all going to EnP at PGH n'at?

Posted 22 April 2020 - 07:43 PM

View PostTyrsen, on 22 April 2020 - 07:19 PM, said:

the aquatic BDSM crowd.



I think you summed up my Discord group right here

#139 Kamikaze Viking

    Member

  • PipPipPipPipPipPip
  • 384 posts
  • LocationStay on Topic... STAY ON TOPIC!!!

Posted 22 April 2020 - 07:57 PM

View PostLarsh, on 22 April 2020 - 06:34 PM, said:

Also, can the idea of solo pugs being scrapped off the bottom of the elite's shoe's be scrapped? When solo players see this mantra it really is discouraging and demeaning.



Agreed we need to focus on that less, but the point many are trying to make is that if done incorrectly this is a warning of what will happen.
If done correctly then this will be prevented. By fixing PSR and no other MM changes this would be fixed in an instant because they wouldn't get matched together (never for the t4-5's or more rarely for the t3s).

All the other of pauls proposed MM changes are aimed at lower queue times and revitalizing queues, which is a good thing, i just hope he picks the right one.

#140 Larsh

    Member

  • PipPipPipPipPipPip
  • The Lanner
  • The Lanner
  • 272 posts
  • Facebook: Link
  • Twitter: Link
  • Twitch: Link
  • LocationYinz all going to EnP at PGH n'at?

Posted 22 April 2020 - 08:04 PM

View PostKamikaze Viking, on 22 April 2020 - 07:57 PM, said:

Agreed we need to focus on that less, but the point many are trying to make is that if done incorrectly this is a warning of what will happen.
If done correctly then this will be prevented. By fixing PSR and no other MM changes this would be fixed in an instant because they wouldn't get matched together (never for the t4-5's or more rarely for the t3s).

All the other of pauls proposed MM changes are aimed at lower queue times and revitalizing queues, which is a good thing, i just hope he picks the right one.


Totally agree on what could happen if this goes all horribly wrong. And if things do go south quickly, we will need to let the Devs know quickly here, or on Twitter, or what ever means we can use to communicate properly.

I think going forward, the players in the forums need to take the seal / pug stomp and push it to the side. Instead, lets focus on helping others up instead of tearing them down.

If I was exceptionally skilled in this game, I would love to have some kind of badge of honor next to my name to show that I've had what it takes to fight with the big league. Be some kind of inspiration to the other players, and with leadership they could trust over comms
--- In before others neg this down because of that NASCAR match where the solo pugs didn't listen to their advice. Yea, we know, it happens.





8 user(s) are reading this topic

0 members, 8 guests, 0 anonymous users