Jump to content

Psr Is The Most Broken Ranking System Ever


33 replies to this topic

#1 MasterTBC

    Member

  • Pip
  • 13 posts

Posted 16 November 2024 - 02:55 AM

Posted Image

PSR is broken since only damage seems to matter.
If you killed 8 mechs with headshots you would still lose PSR because you didn't deal enough damage apparently.
The game punishes players for aiming well, because you deal less damage to kill a mech.

On weekends the Tier isn't the same as during the week.
Pretty much every game is a stomp of 2/12 or 12/2 something like that.

It takes too long to get from one Tier to another therefore most players aren't in the correct Tier.
PSR is kinda evaluating the mech build more than the player.
PSR should be different for each mech otherwhise it limits choices.

There also isn't enough Tiers to make balance reasonable.

Why don't you just use Elo or OpenSkill?

#2 martian

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 8,364 posts

Posted 16 November 2024 - 05:02 AM

View PostMasterTBC, on 16 November 2024 - 02:55 AM, said:

Posted Image
Could you post both the Mission Summary Team Table and the Mission Summary Player Table, please?

Without those tables, it is impossible to judge your performance relative to the other players.

Posting only one cropped part of only one table is not too useful.

View PostMasterTBC, on 16 November 2024 - 02:55 AM, said:

PSR is broken since only damage seems to matter.
If you killed 8 mechs with headshots you would still lose PSR because you didn't deal enough damage apparently.
The game punishes players for aiming well, because you deal less damage to kill a mech.
Do you often kill 8 enemy 'Mechs with headshots in the course of one game?

View PostMasterTBC, on 16 November 2024 - 02:55 AM, said:

On weekends the Tier isn't the same as during the week.
Pretty much every game is a stomp of 2/12 or 12/2 something like that.
You do not have to believe me, but the PSR works exactly the same way during the weekdays and weekend.

View PostMasterTBC, on 16 November 2024 - 02:55 AM, said:

It takes too long to get from one Tier to another therefore most players aren't in the correct Tier.
You mean that you are not in the higher Tier that you feel you deserve to be in, right?

View PostMasterTBC, on 16 November 2024 - 02:55 AM, said:

PSR is kinda evaluating the mech build more than the player.
From what I have seen in the game, you can put a bad player in a 'Mech used by the top players and the bad player will still suck.

Also, there are other important factors than just the 'Mech loadout: Proper tactics, situational awareness, map knowledge, etc. ...

The top players have all these skills too.

View PostMasterTBC, on 16 November 2024 - 02:55 AM, said:

PSR should be different for each mech otherwhise it limits choices.
I see no problem here: Play any 'Mech that you like and let PSR do the rest. Eventually, you will end up in a Tier populated by the players who use the same kind of 'Mechs and builds as you use.

View PostMasterTBC, on 16 November 2024 - 02:55 AM, said:

There also isn't enough Tiers to make balance reasonable.
It does not matter if you create five Tiers or ten Tiers, if the player base is limited (such as is the case of MWO). In the end, the matchmaker must assemble two teams from the players that are available.

View PostMasterTBC, on 16 November 2024 - 02:55 AM, said:

Why don't you just use Elo or OpenSkill?
Because Elo had already been tried. As for any other system, you must sort the players according their skill anyway. If some players sucks in the game, he will suck no matter what skill rating you would use.

Edited by martian, 16 November 2024 - 05:12 AM.


#3 LordNothing

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 17,141 posts

Posted 16 November 2024 - 10:54 AM

its not based on absolute performance its based off relative performance based on what the players in the match did. sometimes you see a match where everyone on both teams did exceptionally well, fought to the last scrap of armor. the red arrow will still be handed out to half the players.

#4 Mad Mech

    Member

  • Pip
  • 15 posts

Posted 16 November 2024 - 12:27 PM

View PostLordNothing, on 16 November 2024 - 10:54 AM, said:

its based off relative performance damage


Fixed that for you.

I agree that "performance" implies more than just doing "damage"; it should probably reflect each person's "teamwork" performance, especially since these are 12v12 TEAMS.

But PGI has shown no interest in teamwork dynamics; just ever-skyrocketing weapon damage output, and the PSR reflects that, which means it's not broken; it's working as intended.

Edited by Mad Mech, 16 November 2024 - 12:29 PM.


#5 LordNothing

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 17,141 posts

Posted 16 November 2024 - 03:22 PM

View PostMad Mech, on 16 November 2024 - 12:27 PM, said:


Fixed that for you.

I agree that "performance" implies more than just doing "damage"; it should probably reflect each person's "teamwork" performance, especially since these are 12v12 TEAMS.

But PGI has shown no interest in teamwork dynamics; just ever-skyrocketing weapon damage output, and the PSR reflects that, which means it's not broken; it's working as intended.


performance, damage, same thing.keep in mind the context, this is mwo were talking about.

what is this ...teamwork... people keep speaking of? certainly not in this game. im happy when my team mates are shooting at stuff.

Edited by LordNothing, 16 November 2024 - 03:25 PM.


#6 Meep Meep

    Member

  • PipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 2,938 posts
  • LocationBehind You

Posted 16 November 2024 - 03:32 PM

Four kills doesn't mean you did any meaningful amount of damage to them. You got one solo kill which means you did ~90% of the damage that killed it and a kmdd which means you got the majority of the damage applied to a mech that died which might be a low percentage overall. Who did more? The guy with 1200 damage and no kills who stripped all the armor off of mechs or the guy who laser washes the red ct a few times and gets the kill? Psr is working just fine when you take the entire match performance into consideration.

#7 crazytimes

    Member

  • PipPipPipPipPipPipPipPip
  • Ace Of Spades
  • 1,363 posts

Posted 16 November 2024 - 10:20 PM

The only change for going up in PSR is you will play against better people in better builds.

If you are struggling with your current tier and builds, you probably wouldn't enjoy getting farmed every match.

#8 MasterTBC

    Member

  • Pip
  • 13 posts

Posted 17 November 2024 - 03:46 AM

View PostMeep Meep, on 16 November 2024 - 03:32 PM, said:

Four kills doesn't mean you did any meaningful amount of damage to them. You got one solo kill which means you did ~90% of the damage that killed it and a kmdd which means you got the majority of the damage applied to a mech that died which might be a low percentage overall. Who did more? The guy with 1200 damage and no kills who stripped all the armor off of mechs or the guy who laser washes the red ct a few times and gets the kill? Psr is working just fine when you take the entire match performance into consideration.


This was just the worst example i could find.
7 components destroyed.
There is nothing more meaningful than destroying an enemy so they can't deal damage anymore.
This also reflects in winrate...
The ranking system is saying that killing a third of the enemy team is bad and useless.
You can't say that i didn't contribute to my team.
PSR went down because my team maybe dealt more damage but they didn't get enough kills so it resulted in a loss.
I made the mistake of quickly killing a mech from behind instead i should have farmed them by stripping off their arms apparently.

This promotes unhealthy playstyle were instead of taking out the enemy you need to deal more damage by shooting off the enemies arms and legs first before you kill them and you also aren't allowed to take risks because you can't afford to get your mech destroyed.

Sacrifing yourself and tanking for the team to win is punished.

I looked up the leaderboard for the same average matchscore and the winrate is massively fluctuating with some having as low as 0.55 up to 1.81 just a quick check of 2 pages reveals.

If matchscore was a reliable measure of skill then how is this even possible.

#9 MasterTBC

    Member

  • Pip
  • 13 posts

Posted 17 November 2024 - 03:53 AM

View PostMeep Meep, on 16 November 2024 - 03:32 PM, said:

Four kills doesn't mean you did any meaningful amount of damage to them. You got one solo kill which means you did ~90% of the damage that killed it and a kmdd which means you got the majority of the damage applied to a mech that died which might be a low percentage overall. Who did more? The guy with 1200 damage and no kills who stripped all the armor off of mechs or the guy who laser washes the red ct a few times and gets the kill? Psr is working just fine when you take the entire match performance into consideration.


I don't think that the game records whether or not the damage actually contributed...
If you deal 80 damage to CT and the CT component gets destroyed you contributed a lot.
While 80 damage on the arm which doesn't get destroyed and then the CT gets destroyed basically means wasted damage that contributed absolutely nothing...

The theoretical example of 8 headshots still holds.
If you get 8 headhots you probably didn't deal more than 200 damage but if the average team damage is 500 you will probably lose PSR.

#10 MasterTBC

    Member

  • Pip
  • 13 posts

Posted 17 November 2024 - 04:04 AM

View PostLordNothing, on 16 November 2024 - 10:54 AM, said:

its not based on absolute performance its based off relative performance based on what the players in the match did. sometimes you see a match where everyone on both teams did exceptionally well, fought to the last scrap of armor. the red arrow will still be handed out to half the players.


Relative performance is bad.
Allies will walk into your firing line so you can't deal damage or use you as a meat shield. You are also fighting your allies.

#11 MasterTBC

    Member

  • Pip
  • 13 posts

Posted 17 November 2024 - 04:12 AM

View Postcrazytimes, on 16 November 2024 - 10:20 PM, said:

The only change for going up in PSR is you will play against better people in better builds.

If you are struggling with your current tier and builds, you probably wouldn't enjoy getting farmed every match.


I'm already getting stomped in losses.
Stomps are not fair. How can you deal enough damage when your team or the enemy is getting vaporized.

#12 Gasboy

    Member

  • PipPipPipPipPipPipPip
  • The Spear
  • The Spear
  • 598 posts

Posted 17 November 2024 - 04:45 AM

Kills only reward a miniscule amount of XP and likely rewards a miniscule amount of psr. Damage done is the metric mainly used. And it is different depending on the weight class.

I'm guessing by the damage taken that you were likely in an assault. Possibly a heavy. Those mechs require a higher amount of damage done than mediums and lights. 500ish is a good target to aim for.

#13 foamyesque

    Member

  • PipPipPipPipPipPipPip
  • Ace Of Spades
  • 748 posts

Posted 17 November 2024 - 10:13 AM

View PostMeep Meep, on 16 November 2024 - 03:32 PM, said:

Four kills doesn't mean you did any meaningful amount of damage to them. You got one solo kill which means you did ~90% of the damage that killed it and a kmdd which means you got the majority of the damage applied to a mech that died which might be a low percentage overall. Who did more? The guy with 1200 damage and no kills who stripped all the armor off of mechs or the guy who laser washes the red ct a few times and gets the kill? Psr is working just fine when you take the entire match performance into consideration.

Solo Kill doesn't mean you did 90% of the damage. It means you got the killing blow on a mech you also did the most damage to.

#14 LordNothing

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 17,141 posts

Posted 17 November 2024 - 10:58 AM

View PostMasterTBC, on 17 November 2024 - 04:04 AM, said:

Relative performance is bad.
Allies will walk into your firing line so you can't deal damage or use you as a meat shield. You are also fighting your allies.


i think its based on the idea that you will only drop with your own tier, which frankly hasn't been the case for a very long time. perform outside of your tier's performance bracket and you will bubble up/down accordingly.

the relief valves are open for most of the day, except for a few peak hours where they work. so you get a lot of bad match making that makes your psr go all over the place.

in either case one game is kind of irrelevant, just one datapoint. it all integrates over time. you will find for every great game where you get the red arrow you might get a mediocre stomp game with a green arrow, and not really care because it was mediocre.

that said, dont expect it to change any time soon. getting pgi to put in any effort towards development is definitely a losing red arrow game.

#15 martian

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 8,364 posts

Posted 17 November 2024 - 11:30 AM

View PostMasterTBC, on 17 November 2024 - 03:46 AM, said:

This was just the worst example i could find.
7 components destroyed.
There is nothing more meaningful than destroying an enemy so they can't deal damage anymore.
Yes.

View PostMasterTBC, on 17 November 2024 - 03:46 AM, said:

The ranking system is saying that killing a third of the enemy team is bad and useless.
You got one honest Solo Kill.

Do you realize that somebody - some other player - had to actually rip those three other enemy 'Mechs open, so you could touch their cherry red CT?

View PostMasterTBC, on 17 November 2024 - 03:46 AM, said:

You can't say that i didn't contribute to my team.
Oh, yes, you did contribute - but not enough, obviously.

View PostMasterTBC, on 17 November 2024 - 03:46 AM, said:

PSR went down because my team maybe dealt more damage but they didn't get enough kills so it resulted in a loss.
Without seeing the Mission Summary Team Table, we do not know what damage other players did. Why do not you post it?

View PostMasterTBC, on 17 November 2024 - 03:46 AM, said:

I made the mistake of quickly killing a mech from behind instead i should have farmed them by stripping off their arms apparently.

This promotes unhealthy playstyle were instead of taking out the enemy you need to deal more damage by shooting off the enemies arms and legs first before you kill them and you also aren't allowed to take risks because you can't afford to get your mech destroyed.
Your playstyle is your choice.

I almost never bother myself with "shooting off the enemies arms and legs first", unless it is an easy task and it gives the team some tactical advantage.

View PostMasterTBC, on 17 November 2024 - 03:46 AM, said:

Sacrifing yourself and tanking for the team to win is punished.
Eating a lot of damage is not always a sign of a high skill. Actually, it is often a sign of a low skill, because it suggests that you were caught badly out of position.

View PostMasterTBC, on 17 November 2024 - 03:46 AM, said:

I looked up the leaderboard for the same average matchscore and the winrate is massively fluctuating with some having as low as 0.55 up to 1.81 just a quick check of 2 pages reveals.

If matchscore was a reliable measure of skill then how is this even possible.
Could you post both the Mission Summary Team Table and the Mission Summary Player Table, please?

Without those tables, it is impossible to judge your performance relative to the performance of the other players.

Posting only one cropped part of only one table is not too useful.

We can not even see your Match Score that you keep talking about ...

#16 Vxheous

    Member

  • PipPipPipPipPipPipPipPipPip
  • CS 2019 Gold Champ
  • CS 2019 Gold Champ
  • 3,826 posts
  • Location2 Time MWO World Champion

Posted 17 November 2024 - 12:10 PM

LOL, if you were to headshot kill 8 mechs, there would be very little damage left for your team to do, which means your relative performance in match would be higher than the rest of your team ---> would most likely result in up arrow.

#17 martian

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 8,364 posts

Posted 17 November 2024 - 12:40 PM

View PostMasterTBC, on 17 November 2024 - 03:53 AM, said:

I don't think that the game records whether or not the damage actually contributed...
If you deal 80 damage to CT and the CT component gets destroyed you contributed a lot.
While 80 damage on the arm which doesn't get destroyed and then the CT gets destroyed basically means wasted damage that contributed absolutely nothing...
No, this is not a wasted damage.

Some other friendly player could take the advantage of it and remove the enemy 'Mech's weakened arm, with the weapon system that could be mounted there.

View PostMasterTBC, on 17 November 2024 - 03:53 AM, said:

The theoretical example of 8 headshots still holds.
If you get 8 headhots you probably didn't deal more than 200 damage but if the average team damage is 500 you will probably lose PSR.
And the example is still absurd as when you posted it for the first time. Have you ever had a normal Quick Play game in which you managed to get 8 headshots?

View PostMasterTBC, on 17 November 2024 - 04:04 AM, said:

Relative performance is bad. You are also fighting your allies.
Actually, it is good.

MWO had a PSR system with the firm thresholds ... and it did not work. Therefore, it was replaced with the current PSR system in 2020.

View PostMasterTBC, on 17 November 2024 - 04:04 AM, said:

Allies will walk into your firing line so you can't deal damage or use you as a meat shield.
Protip: Your BattleMech is equipped with legs. Use them to move a few steps to the left or right.

#18 CFC Conky

    Member

  • PipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 1,840 posts
  • LocationThe PSR basement.

Posted 17 November 2024 - 01:27 PM

@OP,

Despite the down arrow, I doubt your tier level changed very much.

Good hunting,
CFC Conky

#19 Tarl Cabot

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Tai-sho
  • Tai-sho
  • 7,773 posts
  • LocationImperial City, Luthien - Draconis Combine

Posted 17 November 2024 - 03:40 PM

This is why we need the team stats. You are being rated again the other 23 players on the field. Damage converts approx 45% of its points into matchscores, thus that 354 dmg pts = approx 159 MS. It is your MS that is inserted into a formulary to see how you ranked again the other 23 players. The formulary gives the winning team earns a slight boost in that final PSR calculation.

https://mwomercs.com...ost__p__6519483

Depending on how your MS ranks with the other 23 players, that down arrow may have scored you a negative 1 PSR point. And the further you are away from the team, more so from your lance, then fewer triggers. The second link below, AMS no longer generates MS points.

https://mwomercs.com...ity-version-10/
https://mwomercs.com...ost__p__6337261

In the first link, the theorical teams, they had matching MS. The two players who had matching 250 MS, the player on the winning side earned 3 PSR points while the player on the losing side lost 7 PSR points. PGI only using Win/Loss and Matchscore in its formulary, which is zero-sum setup for that match. For you to get an equal or an up arrow, it would have required players on one side or the other, or both, to have earned a smaller Matchscore number.

Could PGI adjust how the up/down arrows are displayed, i.e. one up/down arrow for small change (5pts), two for 6-10 then 3 for 11+ in an attempt to keep people from getting their panties twisted.. :)

#20 crazytimes

    Member

  • PipPipPipPipPipPipPipPip
  • Ace Of Spades
  • 1,363 posts

Posted 18 November 2024 - 12:58 AM

View PostMasterTBC, on 17 November 2024 - 04:12 AM, said:

I'm already getting stomped in losses.
Stomps are not fair. How can you deal enough damage when your team or the enemy is getting vaporized.


You have a +1 W/L and +7000 games over 8 years. You were there for the "PSR is just an XP bar" time, and through the two PSR resets.

Surely you're tier 1 or 2 at worst- what do you expect to change? Do you want more easy people to play against?





2 user(s) are reading this topic

0 members, 2 guests, 0 anonymous users