Jump to content

Psr Community Formula V1.0


144 replies to this topic

#61 RRAMIREZ

    Member

  • PipPipPipPipPip
  • FP Veteran - Beta 1
  • FP Veteran - Beta 1
  • 183 posts
  • LocationIn the Blob

Posted 08 July 2020 - 09:40 AM

Can we expect an update of the player distribution graph after one week?

Edited by RRAMIREZ, 08 July 2020 - 09:41 AM.


#62 OneTeamPlayer

    Member

  • PipPipPipPipPipPip
  • Ace Of Spades
  • 399 posts

Posted 08 July 2020 - 12:24 PM

View PostGagis, on 08 July 2020 - 05:24 AM, said:

That tendency is why most people who play MWO never come anywhere close to the forums. These are a bad medium.


The reason most people who play MWO don't come to the forums is the same as the reason most people who play any game don't come to their related discussion areas, people generally only invest such effort in cases of dedication or frustration.

It's the problem with online polling as well, the only people who show up are the people who are fanatical or deeply invested, or those who are absolutely fed up with the product.

90% of people who use any product/game will almost never take the time out of their day to discuss it online.

#63 General Solo

    Member

  • PipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • 3,625 posts

Posted 08 July 2020 - 10:10 PM

View PostBig-G, on 08 July 2020 - 03:12 AM, said:

Fixed that for ya...


ThanksPosted Image
Exclamation mark added
For EmPhasis!!!

#64 Big-G

    Member

  • PipPipPipPipPipPip
  • The Mercenary
  • The Mercenary
  • 358 posts
  • LocationFormerly New Capetown, now Kikuyu - Lyran Alliance

Posted 08 July 2020 - 10:47 PM

View PostBowelhacker, on 08 July 2020 - 12:43 PM, said:

[Redacted]

No accounting for the general cesspool that is tiers 3-5...

#65 latinisator

    Member

  • PipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 573 posts
  • LocationGermany

Posted 09 July 2020 - 12:12 AM

I am an on-off-guy for MWO right now. Has PGI somehow ****** up the match maker tonnage wise? I am just coming out of a match where we had 1 assault, 6 heavies, 3 meds and a light. The opfor had 8 assaults, 1 heavy, 2 meds and a light. 660t vs 860t. Despite the team having been very passive and thus losing, this does not seem right to me.

#66 martian

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 8,269 posts

Posted 09 July 2020 - 02:43 AM

View Postlatinisator, on 09 July 2020 - 12:12 AM, said:

I am an on-off-guy for MWO right now. Has PGI somehow ****** up the match maker tonnage wise?

This is not a usual **** up, this is the new MWO standard. It is PGI's deliberate decision.

View Postlatinisator, on 09 July 2020 - 12:12 AM, said:

I am just coming out of a match where we had 1 assault, 6 heavies, 3 meds and a light. The opfor had 8 assaults, 1 heavy, 2 meds and a light. 660t vs 860t. Despite the team having been very passive and thus losing, this does not seem right to me.

PGI is okay with it. Get accustomed to it. Sorry.

#67 crazytimes

    Member

  • PipPipPipPipPipPipPipPip
  • Ace Of Spades
  • 1,352 posts

Posted 09 July 2020 - 04:40 AM

At the lofty heights of 2 now. All it took was not playing like a potato and almost no potato builds. To be honest, it was mostly MPL or SPL builds and paying a bit of attention to words of wisdom from people much better than me. Plus getting carried hard in groups.

I will say, I've been team killed more.in the last hundred games than in the preceding thousand or so, and only one was deliberate that I am aware of. Lots of misplaced arty strikes that aren't called in chat. The downside of playing SPL builds, in conjunction with future tier 5s throwing out strikes without looking. Winning.

#68 Tarl Cabot

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Tai-sho
  • Tai-sho
  • 7,729 posts
  • LocationImperial City, Luthien - Draconis Combine

Posted 09 July 2020 - 05:07 AM

View Postlatinisator, on 09 July 2020 - 12:12 AM, said:

I am an on-off-guy for MWO right now. Has PGI somehow ****** up the match maker tonnage wise? I am just coming out of a match where we had 1 assault, 6 heavies, 3 meds and a light. The opfor had 8 assaults, 1 heavy, 2 meds and a light. 660t vs 860t. Despite the team having been very passive and thus losing, this does not seem right to me.


PGI merged the Solo into Group Queue and using the Group Queue MM. And the current tonnage limits allows each group size the ability to bring 2*100 ton assaults.

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

3-player groups: Min Tonnage 80, Max Tonnage 255

4-player groups: Min Tonnage 120, Max Tonnage 280



#69 latinisator

    Member

  • PipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 573 posts
  • LocationGermany

Posted 09 July 2020 - 05:49 AM

View PostTarl Cabot, on 09 July 2020 - 05:07 AM, said:


[...]


That's cute and all, but why on earth did they allow for SOME (not all!) matches to be that lopsided? Ah...nevermind..I just recall it: They do not play their own damned game. So all good now. Posted Image

#70 selfish shellfish

    Member

  • PipPipPip
  • 66 posts

Posted 09 July 2020 - 06:00 AM

View Postselfish shellfish, on 06 July 2020 - 03:50 AM, said:

In theory the new PSR calculation should stabilize you at a certain Tier.


View PostCluster Fox, on 07 July 2020 - 03:18 PM, said:

What is that theory based on? MM has to mix some level of PSR into games because of low POP. Meaning you never quite play against people of your own PSR. There's no force to stabilize it, just hope.


I can only speak for my own experience of the system.

I am currently stabilized at the middle of Tier3 with a W/L ratio of 0.9. When the better players rise to Tiers 1 and 2 presumably the match maker will have enough players to create better games for them. However at the same time I might start to rise towards tier 2 possibly ending up juggling between Tiers 2 and 3.

#71 Cluster Fox

    Member

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

Posted 09 July 2020 - 06:38 AM

View PostGagis, on 08 July 2020 - 05:24 AM, said:

...


I actually have an important update on it. I removed the assumption that WLR and AvgMS are equally representing the player base and used the natural trend line. And I transformed WLR into additive WL aka: (Wins - Losses) / matches. This is how it's calculated at the end and some might not like the X,Y,C numbers...




Now. There is a much more pressing issue :
Quality of matches, slow PSR movement and player retention !!

Note: The stability formula is still 100% valid. Parameter A that would need slight tuning but that's it.

With players getting poor matches right now, Stomps being reported at a much higher frequency. If PGI wants to retain players we need two immediate improvements:
  • Make the PSR stable. Using the stability formula I provided. It only uses PSR as an extra variable, the same one the MM used to make the match !
  • Seed the first few matches. With my Formula, You just need a larger P-factor constant for the first few matches. Aka 0.1 or 0.05. This will put players very very quickly in the ballpark.
In terms of importance, X,Y,C can wait but something needs to be done for Player retention right now.

Edited by Cluster Fox, 09 July 2020 - 08:23 AM.


#72 GARION26

    Member

  • PipPipPipPipPipPip
  • Giant Helper
  • 301 posts

Posted 09 July 2020 - 10:57 AM

View Postlatinisator, on 09 July 2020 - 12:12 AM, said:

I am an on-off-guy for MWO right now. Has PGI somehow ****** up the match maker tonnage wise? I am just coming out of a match where we had 1 assault, 6 heavies, 3 meds and a light. The opfor had 8 assaults, 1 heavy, 2 meds and a light. 660t vs 860t. Despite the team having been very passive and thus losing, this does not seem right to me.




The issue here is the matchmaker has to balance multiple variables with a low population of players available to matchmake most of the time.
1. Group vs non Group
2. PSR
3. Tonnage
4. Acceptable wait time

PGI can assign different values to those different issues and 'how quickly the gates' open when a perfectly balanced match isn't found.

In a low population you can't perfectly match make on the first three and keep Wait Time low. Prior to the merge of group Que it was just balancing PSR and tonnage vs wait times. Adding group que added a variable.

So should group vs non group be more important by tonnage? I think there are plenty of people who say yes on this forum.
Should PSR be more important then tonnage? I think there is good reason to think it should be as Horseman points out.
Would waiting 5 minutes for perfect tonnage matchup be acceptable to most players? (I don't think so)

But it's not all or nothing it's how much you weight each factor and how long you are willing to allow the matchmaker to spin before you just 'open the gates' and put teams together.

Group's aren't actually messing up the tonnage intrinsically - if someone drops with 4 assaults in a group you could still just tonnage balance the other side with 4 assaults. The issue is group vs non group is apparently prioritized higher then tonnage balancing and possibly PSR balancing is higher then tonnage.

Edited by GARION26, 09 July 2020 - 10:59 AM.


#73 Cluster Fox

    Member

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

Posted 09 July 2020 - 11:28 AM

I finished properly documenting my most important PSR suggestion : PSR Stability solution

It doesn't matter how much WLR and avgMS is weighted for this, it'll be stable no matter what.
This uses PSR itself to make PSR stable, a bit incestuous but efficient and well within scope.

Google Docs (better than a post) explains the formula.
https://docs.google....dit?usp=sharing

Slow PSR movement and PSR that doesn't represent player skill is a huge problem right now for player retention. The PSR reset did this. It's delusional to think people losing 20 games in a row want to continue playing.

If PSR is made stable, it doesn't matter what the starting PSR is.

This formula is for stability only and works regardless of how PSR is calculated at the end of a match.

Posted Image

Edited by Cluster Fox, 09 July 2020 - 03:12 PM.


#74 MadcatX

    Member

  • PipPipPipPipPipPipPipPip
  • Big Brother
  • Big Brother
  • 1,026 posts

Posted 09 July 2020 - 04:51 PM

View PostCluster Fox, on 09 July 2020 - 11:28 AM, said:

Slow PSR movement and PSR that doesn't represent player skill is a huge problem right now for player retention. The PSR reset did this. It's delusional to think people losing 20 games in a row want to continue playing.

This formula is for stability only and works regardless of how PSR is calculated at the end of a match.


Hold up... you mean to tell me the 5th option PGI created and implemented is not working all that well at the moment

I'm shocked, SHOCKED, ok, not that shocked.

#75 General Solo

    Member

  • PipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • 3,625 posts

Posted 09 July 2020 - 05:15 PM

IMO I don't think any MM can balance a merged queued with solo's and groups

#76 Tarl Cabot

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Tai-sho
  • Tai-sho
  • 7,729 posts
  • LocationImperial City, Luthien - Draconis Combine

Posted 09 July 2020 - 07:44 PM

View Posttechnopredator, on 09 July 2020 - 05:27 PM, said:

Stomps: since happen practically 99.9% then the answer is right there, just in reverse, reverse what you have been doing to avoid them, because you're simply favor them and make them happen constantly.

How do we know is not cheating? some players killing way too fast, always hitting the CT no matter how U move, that's pretty much BS


Aiming for CT does not mean simply aiming for the middle of a mechs torso. Aim for the head, for most humaniod mechs it is almost all CT but a sliver for the Cockpit And fo non-humanoid mechs, that is lots of CT going over the top.

https://mwomercs.com...localization-2/

#77 crazytimes

    Member

  • PipPipPipPipPipPipPipPip
  • Ace Of Spades
  • 1,352 posts

Posted 09 July 2020 - 10:52 PM

View Posttechnopredator, on 09 July 2020 - 05:27 PM, said:

Stomps: since happen practically 99.9% then the answer is right there, just in reverse, reverse what you have been doing to avoid them, because you're simply favor them and make them happen constantly.

How do we know is not cheating? some players killing way too fast, always hitting the CT no matter how U move, that's pretty much BS


Compared to most FPS games, MWO has huge hitboxes and very slow player movement, 20 tonners aside. I can't even watch fortnite streams they are that fast paced and they get regular headshots, shooting CT in this game is absolutely glacial by comparison.

You should eventually end up matched against people in the same skill bracket as you, just Keep grinding.

Edited by crazytimes, 09 July 2020 - 10:54 PM.


#78 Nesutizale

    Member

  • PipPipPipPipPipPipPipPipPip
  • The Privateer
  • The Privateer
  • 3,240 posts

Posted 10 July 2020 - 01:15 AM

View Postcrazytimes, on 09 July 2020 - 10:52 PM, said:

I can't even watch fortnite streams they are that fast paced and they get regular headshots, shooting CT in this game is absolutely glacial by comparison.


That is pretty much why I am still playing MWO. Tried Warframe with a friend and while the game is interesting its way to fast for me. Most of the time the other players where through the level while I was still in the middle looking at stuff....like tourist in a warzone ^_^
Also got the "You are no longer it a mech! Use sprinting and bullet jumps good damn it!" from my friend.

----

Anyway back to topic. While I am bad at aiming I still get good hits on a single location but some players here are realy good at aiming and keeping hitting the one allready damaged zone over and over. Watch some of their streams. They don't cheat, they are just that good.

Also teamwork. Lock your target with R before firing will tell you where to hurt him most effectively. When you have players who understand that you will find out that they fire together at that single spot...or at least try to.
That way even a group of PUGS can drop a target realy fast. Just with an extra bit of information, going with your lance or at least 2-3 other people of the same speed and just looking who the first guy is firing upon and just adding your firepower to his.

While MWO is in general slower paced dosn't mean that you can be taken down in a second. Go around a corner that you haven't scouted with UAV or by another team member and you can run into a few mechs or the entire other team and you will be dead in a blink of an eye.

Go with your team, be happy about every well placed UAV, cause information is survival and peak around corners very carefully. MWO is unforgiveing. No respawns, no healthpacks or health regeneration. You get one chance and thats it.

#79 RRAMIREZ

    Member

  • PipPipPipPipPip
  • FP Veteran - Beta 1
  • FP Veteran - Beta 1
  • 183 posts
  • LocationIn the Blob

Posted 10 July 2020 - 06:04 AM

View PostRRAMIREZ, on 08 July 2020 - 09:40 AM, said:

Can we expect an update of the player distribution graph after one week?

I guess it's a "no"?

#80 Spare Knight

    Member

  • PipPipPipPipPip
  • 119 posts
  • LocationAlso StankDog the damp

Posted 10 July 2020 - 06:09 AM

The new PSR is functional. I find myself moving up and down. Mostly down. However, there are some tweaks that need to happen next.

MS Kickers need to be adjusted. As an assault, dealing out massive damage, I go up. If I have to ton down for my group and play a support mech, I don't dish out as much damage and I go down. I am keeping my assaults alive, but my score suffers. Even if I feel that I did a good job, I get the dreaded red down arrow, or at best the equal sign.

Tonnage balancing. Wednesday, I was one of two assaults on my team. The other team had 10. We were quickly rolled and it was brutal. We did not have enough armor to compete.





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users