Jump to content

Combined Queues - Discoveries Week 1


344 replies to this topic

#21 Dee Eight

    Member

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

Posted 04 May 2020 - 06:37 PM

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

In the above image, the Tier tightening uses the average PSR of the teams and limits how wide the MM will go before expanding to include a wider skill set of players.


Well the fact you continue to cling to the belief that tier represents skill is why whatever you do isn't going to work. The majority of us have known for YEARS now that tier has almost nothing to do with skill... its a game tenure measurement representing grinding match scores from a formula that rewards poor players with ever raising tier ranks that they don't deserve. Its just not possible for you to admit you were wrong with going to a tier/psr system years ago is it ?

Edited by Dee Eight, 05 May 2020 - 08:11 AM.


#22 Anomalocaris

    Member

  • PipPipPipPipPipPipPip
  • Bad Company
  • 671 posts

Posted 04 May 2020 - 07:01 PM

View PostBrauer, on 04 May 2020 - 04:35 PM, said:

Info on match quality would be helpful even though that can be tough to judge using just numbers like the distribution of different results (12-0, 12-1, etc.).

Would also be helpful to learn why these release valves have to be opened up so wide at this point. I had never noticed them being pulled so far open under the old solo queue system, flawed as it was.


Really need to see match quality. The wait times, even at the worst don't seem noticeably different than what I was experiencing in solo queue before 4/28 on the NA server. I can only assume these wait times include EU and Oceanic which tend to be worse than NA. I would like to see what NA wait times look like because even with the complete opening of all valves it doesn't seem to different than 12p-9p wait times used to be on NA.

Agree with the others about Tier level. You are not measuring player skill here.

I think that pursuing wait times as a primary goal of this merge is one of the least important metrics. Yeah, groups used to have to wait forever, but even 5 min would seem pretty short to those guys now. Anything under 3 min is pretty much a wash IMO. Match quality has got to be addressed first.

View PostKnight Captain Morgan, on 04 May 2020 - 05:11 PM, said:

[Redacted]


I fear you're right but I hope you're wrong because I still really like this game. Even taking a week off so far has sucked.

#23 Cemanicus

    Rookie

  • Little Helper
  • Little Helper
  • 1 posts
  • LocationAdelaide

Posted 04 May 2020 - 07:05 PM

I think one of the other issues with PSR is that (afaik) it doesn't take the mech you are piloting into account either - I'm OK in lights, mediocre in mediums and heavies and an absolute liability to my team in assaults but my PSR is the same regardless of which weight category mech I have selected.

#24 kawaiiChiimera

    Rookie

  • The Marauder
  • The Marauder
  • 9 posts
  • LocationTier 413

Posted 04 May 2020 - 07:12 PM

Haven't played MWO in a few weeks, but these changes look positive enough. Not sure if this has been attempted before by PGI, and if not, why did it take almost a decade? Better late than never. Looking forward to trying it out sometime soon.

#25 crazytimes

    Member

  • PipPipPipPipPipPipPipPip
  • Ace Of Spades
  • 1,316 posts

Posted 04 May 2020 - 07:29 PM

View PostkawaiiChiimera, on 04 May 2020 - 07:12 PM, said:

Haven't played MWO in a few weeks, but these changes look positive enough. Not sure if this has been attempted before by PGI, and if not, why did it take almost a decade? Better late than never. Looking forward to trying it out sometime soon.


Why wasn't it done earlier? Not sure. I mean, watching premade comp teams in meta builds waffle stomp cadets in trial mechs is good for a bit.of a laugh, I can't see it discouraging new players at all.


#26 kawaiiChiimera

    Rookie

  • The Marauder
  • The Marauder
  • 9 posts
  • LocationTier 413

Posted 04 May 2020 - 07:52 PM

View Postcrazytimes, on 04 May 2020 - 07:29 PM, said:

Why wasn't it done earlier? Not sure. I mean, watching premade comp teams in meta builds waffle stomp cadets in trial mechs is good for a bit.of a laugh, I can't see it discouraging new players at all.


To be fair, I've done plenty of roflstomping myself. Its part of the fun for me. Was hoping to replace it with fighting actual competent players at some point, but was always too busy with work to do try comp more than once; played in ISC in the Awful Waffles. This may do, though. Can't wait to fight Bowser and Derek every match in my ****** Cicada haha.

#27 FRAGTAST1C

    Member

  • PipPipPipPipPipPipPipPipPip
  • The Fighter
  • The Fighter
  • 2,869 posts
  • LocationIndia

Posted 04 May 2020 - 07:52 PM

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.

#28 Tatula

    Member

  • PipPipPipPipPipPipPip
  • Liquid Metal
  • Liquid Metal
  • 683 posts
  • LocationSF Bay Area

Posted 04 May 2020 - 08:06 PM

Enjoying the combined queue, but this stupid matchmaker needs to be fixed. Just had a match where we had no assaults vs 7 on the enemy team. That's plain stupid and lazy.

#29 FrigginWaffle

    Member

  • PipPip
  • Death Star
  • Death Star
  • 29 posts
  • LocationBelgium

Posted 04 May 2020 - 08:12 PM

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.

Edited by FrigginWaffle, 04 May 2020 - 08:13 PM.


#30 ManOfAthens

    Member

  • PipPip
  • Ace Of Spades
  • 31 posts

Posted 04 May 2020 - 08:14 PM

View PostAloha, on 04 May 2020 - 08:06 PM, said:

Enjoying the combined queue, but this stupid matchmaker needs to be fixed. Just had a match where we had no assaults vs 7 on the enemy team. That's plain stupid and lazy.


I think I was in that game too. The other side (one with 7 assaults) also had a group on it of at least 2 or 3. Guys I had seen earlier who have been running LRM Awesomes together. Not sure how that factors in, but you would think that groups should end up with lower tonnage due to the limitations, no?

Edited by ManOfAthens, 04 May 2020 - 08:15 PM.


#31 Korvus Knull

    Member

  • Pip
  • The Hunter
  • The Hunter
  • 11 posts

Posted 04 May 2020 - 08:59 PM

We appreciate the effort that is being made in the area of wait times. Two or three friends getting together to play has been my norm for awhile, and this new system has got us some fun and diverse drops this last week or so. Thanks again for taking the time to fix things as best as possible. I have been playing a BattleTech or MechWarrior game since 1985, when you had to have friends to even play! I know some newcomers may find the learning curve a bit steep, but listening to your teams comms will help to get you up to speed. Also many tube videos to get some experience on builds and tactics. Thanks to Russ and team for listening to our issues and working on a solution. MechWarrior forever!

#32 BARRY SHlTPEAS

    Member

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

Posted 04 May 2020 - 09:21 PM

[redacted]

[mod]for unconstructive and offensive content
Please keep the discussion constructive, folks![/mod]

Edited by GM Patience, 06 May 2020 - 12:26 PM.


#33 Swamp Ass MkII

    Member

  • PipPipPipPipPipPip
  • 333 posts

Posted 04 May 2020 - 09:50 PM

I have been in quite a few matches tonight, were one team had Assaults, meaning 4 to 6, while the opposing team had none. This is a big assault problem... :(

Could we break the groups down so they would be in the appopiate lance for weight class?

Alpha: Light and Medium
Bravo: Light (When it's a very light drop), Medium, Heavy, and Assault (When it's a very heavy drop).
Charlie: Heavy and Assault

This is my only complaint...

And as far as wait times, well, I am tier 3, and I don't think I have ever waited for more than 2 minutes. If I did, I select a different weight class and boom, got a round... :D

#34 Zirconium Kaze

    Member

  • PipPipPip
  • Ace Of Spades
  • 92 posts

Posted 04 May 2020 - 09:54 PM

I still think unrestricted qp is the way to go. If anything, allow the use of trial mechs in drop decks and allow drop decks into matches. The chaos is actually good. I have even seen matches where the majority of the team are assaults and they lose against lights and mediums. I love how qp has become. Actual skill gets displayed. Plus, people ain't gonna get better if they play with the same based on some arbitrary restriction. Keep restrictions for actual comp play, but for just having a romp ol' time, let qp be free. I personally think it would be super swell if we could have drop decks for qp. Same weight restrictions as fp but allow the mixing of inner sphere and clan mechs.

#35 Tarogato

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Civil Servant
  • Civil Servant
  • 6,557 posts
  • LocationUSA

Posted 04 May 2020 - 10:12 PM

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

Hey MechWarriors,

As I mentioned last week, I'd be playing with a few of the match making numbers on the live servers throughout the weekend and I'd share info that was discovered in doing so.

[...]

It became very clear that there was a heavy trade-off in wait times doing even minimal adjustments to the release valves and hard limits on Tier/Weight Class. Essentially it boils down to the same principle of economics:

Basically we are putting a point (set of MM values/restrictions) somewhere in the above triangle. Unless the point is dead center, there's a give and take effect happening. If that point leans toward any given corner of the triangle, the other two properties will take a hit in balance.
[...]

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. It now comes down to how much time is acceptable to get the Tier balancing and Weight balancing having a positive impact.


What interests me is how altering the MM preference actually influences the quality of matches.

As we know, the value of each slot on a team is a function of the pilot's skill, the effectiveness of the mech (of which tonnage is only a small component), and the combination of the pilot in that given mech.



If you tweak the MM to prefer evenly matched tonnage across teams, how much does that improve the evenness of the matches? You can look at how even the matches were when they started as you did in the OP, but what I really want to know is how evenly they ended. Or rather, how much does tonnage imbalance influence the result of a match? How strong is that influence? If tonnage imbalance to a certain degree can be shown to have little effect on the actual winner of the match, then it is not worth sacrificing MM queue time in order to achieve a perceived tonnage balance, because other factors are having a stronger influence.

Same argument for Tier balancing. When there is a tier imbalance, how much does the tier-favoured team actually come out as the victor? Is it a very strong correlation? By answering this question, you can decide how valuable it is to spend MM queue time to ensure more even Tier balancing.

What I suspect, is that within certain limits, these correlations are not extremely strong - really only readily apparent in more extreme imbalance. Match results depend on so many factors, that even with perfect tonnage and tier balancing, there is a lot of variance in whether the players perceive the matches to be even themselves, so striving for better weight and tier balance would lengthen queue times for very little benefit.

This is the type of research I began here, with limited resources available:
https://mwomercs.com...is-of-the-12-0/

Ie., are bad results (matches that ended imbalanced) correlated to some basic starting conditions?




For instance, I found that tonnage actually had a negative correlation. When a match was a stomp, it was very slightly more likely that the lighter tonned team was the victor. Now, my research was limited in scope, but that this result even occurred *at all* I think is significant, and begs a more thorough investigation.

And the strongest correlation I noticed was predicted by W/L ratio. When there was a stomp, it was most likely that the team who won, was already comprised of players who averaged a higher W/L ratio than the losing team. This component is not measured by PSR in its current form. Nor is a player's average match score (which is roughly a continuous record of the direct value of a player). Sure, these components influence the rate of change in a player's PSR, but PSR ultimately does not accurately measure how valuable a player for the purposes of fair matchmaking. And PSR, not being a zero-sum system, has a cap.

So among the upper echelon of players, Tier 1, there is a humongous variance in the actual skill that these players wield, and PSR does not currently account for that variance in any way whatsoever.

So how valuable is it, truly, to sacrifice any MM expediency in favour of more balanced PSR matching, when so many Tier 1 players are really Tier 2 players in actual skill... and so many Tier 2 players are simply people who haven't grinded enough PSR to inevitably reach Tier 1 yet?



Before we can achieve a proper balance between MM time, and quality of matches, we need to improve both our means of assessing the actual quality of a match, and our means of measuring the value of an individual player (how influential they routinely are on the outcome of a match). Until then, I daresay it would be safer to simply lean towards quicker matches.

#36 kawaiiChiimera

    Rookie

  • The Marauder
  • The Marauder
  • 9 posts
  • LocationTier 413

Posted 04 May 2020 - 10:28 PM

View PostTarogato, on 04 May 2020 - 10:12 PM, said:

What interests me is how altering the MM preference actually influences the quality of matches.


My boy Taro taking the time to write out this wall of text. Is it too late to add something similar to Elo to the game? Or attempt to make Tiers actually a useful metric? Say, for starters, giving it a proper curve, so that there's actually a fifth of the current playerbase in each tier instead of it being a metric of time played for most players. Of course, you probably know more about PGI's matchmaking than I do Posted Image

#37 IanDresarie

    Member

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

Posted 04 May 2020 - 10:43 PM

First if all, thanks for the data and great explanation!
Good communication is addicting, keep it up and give us even more data!

#38 Kamikaze Viking

    Member

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

Posted 04 May 2020 - 10:45 PM

View PostkawaiiChiimera, on 04 May 2020 - 10:28 PM, said:

Is it too late to add something similar to Elo to the game?


We don't need to make more work than is necessary. Also we used to have Elo, it was replaced by PSR. its unlikely that they will rebuild the system again, so try to work within that system.

The Key problem here is that PSR has an upward bias, That is all.
As PSR is the Foundation for the Tiers, And Tiers are used by the Matchmaker. It is the equivalent of bad building foundations.

BUT the system Paul made was designed well (if I understand it correctly). The only issue being the default values chosen were never adjusted.

This is what we currently have:
Posted Image
This is what my version would be: (note, no programming required, only value changes)
Posted Image
Or Alternately this (which does require a small amount programming)
Posted Image


And I believe this minor change fits with everything Tarogato said above.
By fixing this SKILL issue here at the PSR level, then Tiers are set correctly, & it doesn't need to be fixed at the MM Triangle shown in Pauls OP. Hence Paul can focus the MM on getting fast matches, because the Skill component is dealt with elsewhere.

Note: this change would not have an instant effect on the matchmaker but it would correct Skill vs tier issues over time. But if a PSR reset or re-distribution was also done afterwards then its benefits would show up straight away.

Edited by Kamikaze Viking, 04 May 2020 - 10:51 PM.


#39 Magic Pain Glove

    Member

  • PipPipPipPipPip
  • CS 2023 Gold Champ
  • CS 2023 Gold Champ
  • 138 posts

Posted 04 May 2020 - 10:54 PM

First day was absolutely dreadful and this explains why pretty much as the flood gates have been widely opened . After the tiers have been "tightened" my group and my solo experience has been improved slightly . I mean as much as it can be improved considering how flawed the tier system is.

View PostTarogato, on 04 May 2020 - 10:12 PM, said:

Before we can achieve a proper balance between MM time, and quality of matches, we need to improve both our means of assessing the actual quality of a match, and our means of measuring the value of an individual player (how influential they routinely are on the outcome of a match). Until then, I daresay it would be safer to simply lean towards quicker matches.


I am glad that this Stat Analysis thread got linked by someone who probably did more research than PGI itself when it comes to match outcomes and their dependency on various metrics .

Well , I am here waiting for people to start arguing that they don't care about match outcomes / balanced matches ... And how its "mission impossible" to incorporate more valuable ranking system even withing confines of the current one by simply changing metrics which we use to separate tiers from one another and how one climbs or drops a tier.

#40 Voice of Kerensky

    Member

  • PipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 473 posts

Posted 04 May 2020 - 11:08 PM

And you conduct such experiments on living people, ordinary players? Although I suspect that such very strange steps are caused by a very small number of players in this game.
I want to say the following for our dear PGI:
On a crooked foundation (skill level system and matchmaker) you will never be able to build a good, reliable and comfortable home.
Your matchmaker has never been good, but you decided to conduct such an extensive and bold test on it (which he obviously can not cope with) using all the players. This is a very wise decision.
Here are some bright and memorable moments, see how the matchmaker works:
Conquest match: a team in which there is not a single light mech is fighting a team in which there are five (!) light (fast) mechs and three (!) fast medium mechs are added to five light mechs in this team. I think this was done so that the game balance was maintained.
Posted Image

And this is just a screenshot of the battle with a typical score of 12: 2. In one of the teams, six (!) people failed to get 100 damage.

Posted Image

I have many more similar screenshots.

Not so long ago, I participated in a battle in which 4 (!) players died in my team in the first two (!) minutes of the battle, in the next five seconds the fifth died. Is that a matchmaker? No, it always seemed to me that a matchmaker should create interesting matches.





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users