Jump to content

Combined Queues - Discoveries Week 1


344 replies to this topic

#61 Slothasaurus

    Member

  • PipPip
  • The Hunter
  • The Hunter
  • 27 posts

Posted 05 May 2020 - 05:42 AM

View PostMrMadguy, on 04 May 2020 - 11:23 PM, said:

My opinion - funny thing, but quality of matches was much higher, when there wasn't weight class balancing. Because, you know, for example I'm bad Assault pilot. Therefore for me playing in Heavy > playing in Assault. This assumption, that higher weight always > lower weight - is just misconception. Skill - is only thing, that really matters.


While it is true skill is one of the most important, the most vital, aspects of play there still needs to be restrictions for some sort of balance. To say otherwise is not smart.
You are good with heavies. Say for example you had a team that was all good with heavies vs. a team that was equally skilled players with assaults. With the skill being equal on both sides the team with the assaults would more than likely win. They might not win all of them, but they would definitely win the majority of matches.
In QP matches there are many types of players and skill levels being thrown in together. Giving one side a +5 or +6 assault advantage gives the team that advantage before the match even starts.
Does the team with that much of an assault advantage win every time? No, but I am sure the data shows that those teams won the majority of the time.

#62 Warning incoming Humble Dexterer

    Member

  • PipPipPipPipPipPipPipPip
  • Bridesmaid
  • 1,077 posts

Posted 05 May 2020 - 05:43 AM

I think "Tier Balancing" and "Weight Class Balancing" could be replaced by "Match Balancing", specially because Tiers aren't reliable.

Then instead of a 3D Triangle you have a simple 2D line, between "Match Maker Speed" and "Match Balancing".


But another suggestion would be "Fun Balancing" or "Win Balancing" : Regardless or Tier or Weight, the more a player is winning, the harder you make it for him to win again, if only by opposing him to other player who keep winning.

Because in the end who cares what the Tier Balance, Weight Balance or even Match Balance is, as long as you're getting enough wins to keep you playing : Most players won't care how incredibly perfectly well balanced everything is, if striking that perfect balance means they're personally stuck in a lose streak where they're not getting a single win.

Bad players just want some wins to keep them happy, and perhaps good players want a proper harder challenge instead of a balanced one, even if that means unbalancing the matches on purpose to make them easier to win for some, and harder to win for others.

#63 Knight Captain Morgan

    Member

  • PipPipPipPipPipPip
  • Bad Company
  • 340 posts

Posted 05 May 2020 - 05:43 AM

View PostSpare Knight, on 05 May 2020 - 04:38 AM, said:


You guys could go to a private lobby and would be together every match. You could choose your maps, etc. It works very well. We do it for training and such. Posted Image

But you can’t farm PUGS in a private lobby. If you can’t do that the WTH is the point of even forming a premade?

#64 C337Skymaster

    Member

  • PipPipPipPipPipPipPipPip
  • Philanthropist
  • Philanthropist
  • 1,450 posts
  • LocationNew England

Posted 05 May 2020 - 05:45 AM

View PostFrigginWaffle, on 04 May 2020 - 08:12 PM, said:

Adjustments to the weight or tier classes or not, at least change the spawn points of the lances to be much closer to their respective team, so that the slow assaults who randomly get put in the alpha lance don't get rushed and caught so easily. Would prevent a lot of frustration in the first place.

This has been an issue in group que for years and nothing happened so I'm hoping these times finally catch the attention.


I wonder why they went with Leopard Dropships in the first place, and never just added Union Dropships that would drop the entire team in one spot. Wouldn't it be more reasonable to have a 'mech company deploy to a single location, rather than be spread all over creation?

#65 Isaiah Kell

    Member

  • PipPipPip
  • Hero of Davion
  • Hero of Davion
  • 68 posts

Posted 05 May 2020 - 05:51 AM

Thank you for stopping the whole "assaults only drop in C Lance" thing.

#66 Zirconium Kaze

    Member

  • PipPipPip
  • Ace Of Spades
  • 92 posts

Posted 05 May 2020 - 06:05 AM

View PostSlothasaurus, on 05 May 2020 - 05:42 AM, said:


While it is true skill is one of the most important, the most vital, aspects of play there still needs to be restrictions for some sort of balance. To say otherwise is not smart.
You are good with heavies. Say for example you had a team that was all good with heavies vs. a team that was equally skilled players with assaults. With the skill being equal on both sides the team with the assaults would more than likely win. They might not win all of them, but they would definitely win the majority of matches.
In QP matches there are many types of players and skill levels being thrown in together. Giving one side a +5 or +6 assault advantage gives the team that advantage before the match even starts.
Does the team with that much of an assault advantage win every time? No, but I am sure the data shows that those teams won the majority of the time.

Who cares? I am a solo player without affiliation or knowledge of whose who in this game and I think qp is fine just the way it is. Stomps are literally impossible to stop. They were happening before the change and happen at what appears to be the same rate to me. Things are fine. It's like some of you can't enjoy your meals and the rest of you think you're doing something good for suggesting such a "balance" but all you do is destroy the diversity an unrestricted qp brings in. If it was up to me, I'd allow drop decks into qp. Part of what made the older mechwarrior online games fun was that people had the freedom of choice and simply learned the mechanics and rules of the game and went from there. There will always be a best and a worst and trying to bring it into the middle ruins it by making things stale.

Also, if teams are equally skilled, I would argue medium and heavies would beat an entire team of assault mechs. They're slow, some of them have weapon placements that's very inconvenient for certain terrain, and they turn sooooo slow. A lance of heavies could literally flank assaults, take their pop shots, and move fast enough to reposition before the assaults even moonwalk to where they last were at. Assaults don't have that much armor otherwise one walking around the wrong corner wouldn't go down in seconds from focus fire. And team coordination deserves to be put on display for all to see, not just a pool of team players. Solo artists just need to learn how the teams operate and work in such a manner that they capitalize on their team mates if they're unorganized. I can't count how many times people keep popping out of the same spot to trade shots until they die and then another takes their place instead of retreating or repositioning. Assaults don't have that luxury. If people are searching for them, they have to stand their ground unless they're the first to move but that requires being situationally aware and even then, you can get caught. Yet something that runs at 81 kph can effectively get away unless a lance of faster mechs chase you down, but if they break from their main group to do that, they risk getting their stuff pushed in if the rest of your team is aware. Ultimately, calling shots and focus fire wins the game. Nothing can balance that and people who refuse to coordinate deserve the loss they get from those who do coordinate. Even then, nothing is stopping lances from breaking off into a three pronged type formation so they can take pop shots from every angle instead of where the murder ball is at where everyone has their guns pointed at. Because if they call charlie and everyone aims and opens fire on charlie and then alpha, beta, and gamma pop up from their flank to take their pop shots and then back off into cover immediately what you're eventually gonna have is a break in formation and if that occurs, then it comes down to pure skill instead of mass targetting. At that point, you have no one to blame but yourself. Those assaults ain't gonna point you from charlie in under 4 seconds. They turn too slow.

Edited by Zirconium Kaze, 05 May 2020 - 06:29 AM.


#67 Brauer

    Member

  • PipPipPipPipPipPipPipPip
  • 1,066 posts

Posted 05 May 2020 - 06:14 AM

View PostZirconium Kaze, on 05 May 2020 - 05:37 AM, said:

That cadet was in my group.


Ah, good to know. I had wondered if they were in a group.

#68 Vidarion

    Member

  • PipPipPipPipPip
  • Philanthropist
  • Philanthropist
  • 102 posts

Posted 05 May 2020 - 06:39 AM

Is there any way to get a rundown on match results (kill totals, damage totals, avg damage per mech, etc)? People keep talking about 12-0, 12-1, etc but, honestly, kill totals don't tell the whole story. I've had a bunch of REALLY good (fun, exciting, whatever) matches that ended with lopsided kills but when you look at the amount of damage each side put out, they were very close. There have also been a bunch of steamrolls where one side puts out 2x (or more) the amount of damage and those are much less fun, but on the whole the matches have been pretty engaging.

It would be interesting to see if PGI can tease out things like:
1. Damage averages / totals (similar numbers on each side would probably indicate a more "fun" match than, say, noticeably different numbers)
2. Hit location averages (coordination of fire, are multiple players hitting the same hitbox on a target)
3. Avg number of simulateous shooters per target (concentration of fire, are multiple players engaging the same target at the same time)
4. Avg number of assists per kill (are multiple players engaging the same target at different times)
5. Avg TTK after taking initial enemy fire (are targets falling quickly after initial engagement or are they able to disengage and re-engage)

These metrics would allow people to give feedback on what they find to be engaging / fun and maybe there would be more / different dials to turn to achieve them.

Edited by Vidarion, 05 May 2020 - 09:45 AM.


#69 L1f3H4ck

    Member

  • PipPipPipPipPipPipPip
  • 738 posts

Posted 05 May 2020 - 06:39 AM

First of all, and I know that's subjective, I feel an increase in match quality since the merge. There are stomps, but also drawn out, nailbitingly close matches.

Great to see work done and getting updates, that's a breath of fresh air all in itself.

I still think the biggest hurdle in this is PSR. Especially in Tier 1 the skill gap is enormous, and I don't think there will be a proper matching of these players in particular, without introducing some other metric. I'm not suggesting to go full Jarl's List, but something to that effect, maybe distribute the players in a match based on average matchscore over the last 3 months.

I'm not strictly against having the opposite ends of the skill spectrum meet occasionally btw, it gives new players a chance to observe good players early on.

#70 MrFatBard

    Member

  • PipPip
  • The Terror
  • The Terror
  • 36 posts

Posted 05 May 2020 - 06:58 AM

so you arent merging group and normal ques then Posted Image

#71 L1f3H4ck

    Member

  • PipPipPipPipPipPipPip
  • 738 posts

Posted 05 May 2020 - 07:04 AM

View PostMrFatBard, on 05 May 2020 - 06:58 AM, said:

so you arent merging group and normal ques then Posted Image


They ARE merged, that's what this is all about, now turn that frown upside-down.

Edited by Dakkalistic, 05 May 2020 - 07:04 AM.


#72 Cluster Fox

    Member

  • PipPipPipPipPip
  • 104 posts
  • LocationStuck on a rock in Grim Plexus

Posted 05 May 2020 - 07:37 AM

Paul, thanks for the information and transparency about the tweaks and tests you and PGI are doing, it's great to see.

Could you please add something like the graph enclosed for the different cases discussed? Your post still doesn't quite give an idea of the effect on overall balance and I think a graph like this could help visualize how the parameters are really affecting the balance. Providing one for how it was in SQ before the merge would give a good baseline too, but that might be too much to ask.

It could simply be a bell curve if you don't want to expose exact numbers. And if it's still too much data exposed, may I suggest "average damage difference" or "average team score difference", which would compare sum of damage and sum of player score for both teams.

Posted Image

Edited by Cluster Fox, 05 May 2020 - 08:01 AM.


#73 Horseman

    Member

  • PipPipPipPipPipPipPipPipPip
  • The Mercenary
  • The Mercenary
  • 4,694 posts
  • LocationPoland

Posted 05 May 2020 - 08:48 AM

View PostKnight Captain Morgan, on 05 May 2020 - 05:43 AM, said:

But you can’t farm PUGS in a private lobby. If you can’t do that the WTH is the point of even forming a premade?
For a full private lobby match, you'd have to gather 12 players on each side. And for all the time and effort it took to arrange that, you wouldn't get any XP, C-Bills or event progress... which is a bit meh.

View PostZirconium Kaze, on 05 May 2020 - 06:05 AM, said:

Stomps are literally impossible to stop. They were happening before the change and happen at what appears to be the same rate to me.
According to PGI, stomp frequency increased - so no, they are not happening at the same rate.
While there will always be stomps, a properly tuned matchmaker would reduce their frequency. PUGs don't enjoy being stomped, and most competitively-minded players don't enjoy stomping pugs - they're cleaning out the match so that they can queue up again for another shot at getting matched against worthwhile opponents.

Quote

There will always be a best and a worst and trying to bring it into the middle ruins it by making things stale.
The best matches are clutch games where the final outcome is never certain until the last moments. Equalizing teams in the matchmaker makes it more likely for such tense games to result.

Quote

Ultimately, calling shots and focus fire wins the game. Nothing can balance that and people who refuse to coordinate deserve the loss they get from those who do coordinate.
True, but again - properly balanced teams would have a reasonably equal proportion of each of those types in them. Not a spontaneous hivemind on one side and a disorganized horde of solo berserkers on the other.

Edited by Horseman, 05 May 2020 - 08:49 AM.


#74 Gakuseinozen

    Member

  • PipPipPip
  • Giant Helper
  • 55 posts

Posted 05 May 2020 - 09:06 AM

Players should not be measured by PSR. It is well known that PSR is not a good metric to compare players skill level; anyone can get to Tier 1. Can we try replacing it with another readily available, and more reliable, metric? Such as Average Match Score?

Thank you

Edited by Gakuseinozen, 05 May 2020 - 09:06 AM.


#75 Sniper09121986

    Member

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

Posted 05 May 2020 - 09:30 AM

View PostFRAGTAST1C, on 04 May 2020 - 07:52 PM, said:

While it's a grand accomplishment that you've been tweaking numbers here and there, I'd suggest for you, Russ and Chris to play some matches as well to get first-hand experience of how things are going.


You know not of what you speak...



#76 UnkerZ

    Member

  • PipPipPip
  • 69 posts

Posted 05 May 2020 - 09:34 AM

The MOST important metric is forever conveniently passed. Were the matchups balanced?

Because if they were by any slim margin I'd surely be swinging my balls in public. But the fact they are quietly sideskipped quoting matchmaking time is a perfect attempt to deflect the issue.

I will wait 5 minutes if that is what it takes to get a balanced matchup - but clearly match quality isnt the goal till today. So while you continue to tweak around the triangle that clearly DOES NOT apply the key metric of what all this effort is (i.e. quality balanced matchups) - im loading up my potato roller 100 damage alpha face slapper.

I just rolled out of a game with a MCII trying to climb a 45 degree slope until he got mauled to death (inb4 the nascar nearly finished another lap around him). Not to add he had no idea what was going on, did not know what is the button 'Q', and did literally <50 damage. Thats Tier 1 PSR for you with a 50 second wait time. Clearly the matchmaker didn't run out of players and was having a great time setting the matchup.

Edited by UnkerZ, 05 May 2020 - 09:37 AM.


#77 Pz_DC

    Member

  • PipPipPipPipPipPipPipPip
  • Staff Sergeant
  • 1,100 posts

Posted 05 May 2020 - 11:00 AM

Well, my 50 cents.
1st - great job, dear devs and great explanation dear Paul.
2nd - with all those experiments I've noticed drastcly reduces of nascar, and a bit less reduction of steamrolls.
3rd - If, IF You will ask me - then in triangle of "mm time-weight balance-tier balance" those nodes are not equal, i.e. tier balance affect gameplay more then weight balance, while weight balance affect gameplay more then mm time. So finding a "golden middle" is not an option, and even middle of "weight balance" and "tier balance" is not an option. IMHO.

4th - current PSR system works... with some problems, mean it's MUCH more easy to farm uptier then get it down, fix it and then we can talk about balance in triangle etc.

P.S. sorry for all grammatic mistakes and so on, hope You can understand my message well :)

#78 Alreech

    Member

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

Posted 05 May 2020 - 12:19 PM

View PostPaul Inouye, on 04 May 2020 - 12:52 PM, said:

There are times where the Match Maker runs out of players. When this happens and the queue is waiting to fill beyond release valve timers, there will be the odd circumstance where all valves are open and you will see a full open Tier/Weight Class match formed. While this is fairly rare, it is possible.


How about setting up matches with smaller teams if the match maker runs out of players?
Like 8 vs 8 on smaller maps?

Edited by Alreech, 05 May 2020 - 12:21 PM.


#79 Alreech

    Member

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

Posted 05 May 2020 - 12:35 PM

View PostKnight Captain Morgan, on 05 May 2020 - 05:43 AM, said:

But you can’t farm PUGS in a private lobby. If you can’t do that the WTH is the point of even forming a premade?

And you can't farm PUGs in Solaris either, that's the reason the Solo Players play the 12 vs 12 game mode on big maps, build around 3 Lances of players that should workt together. ;)
And because the maps in 12 vs 12 are to big and the players don't work together most matches in quickplay depend on simple tactics like nascaring around and mech builds without any team work in mind.

#80 Alreech

    Member

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

Posted 05 May 2020 - 12:46 PM

View PostPaul Inouye, on 04 May 2020 - 12:52 PM, said:

In the image above, I tried to find a happy medium where time to get a match, Tier/PSR matching and Weight Class matching all had their place. The combination of Tier and Weight matching is additive so no matter what the settings are set to, the global wait time is going to increase.

If all players have the same Weight you can take that factor out of the Match making.

Same Weight can be done with tools yet in MWO: Drop Decks.
A drop deck of 3 Mechs with for example 165 tons would give all players a Weight of 165.





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users