Jump to content

Psr Update And Changes - Jun 2020


490 replies to this topic

#281 evlkenevl

    Member

  • PipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 73 posts

Posted 05 June 2020 - 11:42 AM

Whether this works or not, I'll give it a go. I've always wanted wins to count for more in this game instead of personal stats. If they don't take win/loss into account, players will keep behaving like it's a single-player game. We currently have guild groups running off with no communication. Folks constantly shot-blocking teammates or parking behind them. Anyone who suggests a strategy is more likely to get yelled at or ignored than anything. Lots of other things too, many of which I admit being guilty of.

Just saying that losses have to hurt. It sucks when you've done well just to have the end-game numbers say otherwise. But if they only award single player performance, that's all people will care about.

#282 DeathStalker55

    Rookie

  • The 1 Percent
  • 2 posts

Posted 05 June 2020 - 11:58 AM

I only read the first 2 & 1/23 pages of posts, but about 15 plus pages from other post about reset PSR and and all that. Seems to me most ppl want or need to see Tier # next to their name to feel good about playing a game they supposedly like/love/ or is their favorite game. The rest are for reset and even doing away wit the xp bar. (Cant plz every1 all the time (FOOLS ERRAND!))
As for the stomps, they are not exactly as bad as every1 says, or as the score indicates. Ive been in 12-0, 12-1, 12-3 stomps (on both sides the last week) but they really could of just as easily went the other way! Sometimes its just the luck of the mech being targeted and where its actually being hit. Some of these so called stomps, went for or against some of these teams with quite a few walking dead on the winning side. Just because a game score looks & feels like its close doesn't mean it actually was cause there may have been a very skilled pilot wit an undamaged mech wit plenty of weapons n armor to take on the last 3 or 4 mechs alone with ease and vise versa.
Plz stop Crying on the forums, complaining, making lil kid threats, etc is not very mature or helpful and them kinds of replies makes ppl not want to help or do any extra work to help fix the problems the game we all care about.

#283 Nearly Dead

    Member

  • PipPipPipPipPipPip
  • Philanthropist
  • 274 posts

Posted 05 June 2020 - 12:00 PM

So, if teamwork is the biggest factor in winning matches, and premade groups can assume to be using teamwork and comms group players should, on average, have a higher win ratio than solo players.

Let's assume player "PlayerWithNoName" has two accounts. On account one he plays only solo. On account two he plays only with his unit in a four man premade. He plays the same number of games on each account per month. To keep things simple he took the build code from account one and built an exactly matching mech for account two and only plays those two, identical, meta mechs.

How does the PSR math work out?

Edited by Nearly Dead, 05 June 2020 - 12:02 PM.


#284 Brauer

    Member

  • PipPipPipPipPipPipPipPip
  • 1,066 posts

Posted 05 June 2020 - 12:02 PM

View PostNearly Dead, on 05 June 2020 - 12:00 PM, said:

So, if teamwork is the biggest factor in winning matches, and premade groups can assume to be using teamwork and comms group players should have a higher win ratio than a solo player.

Let's assume player "PlayerWithNoName" has two accounts. On account one he plays only solo. On account two he plays only with his unit in a four man premade. He plays the same number of games on each account per month. To keep things simple he took the build code from account one and built an exactly matching mech for account two and only plays those two, identical, meta mechs.

How does the PSR math work out?


The group play account gets boosted. TBF the current QP meta is to group up with a few strong players and roll the queue. If groups are going to be in the queue I don't think it makes a whole lot of sense to ignore that considering you have players stacking up double digit WLRs right now by playing in groups.

#285 VonBruinwald

    Member

  • PipPipPipPipPipPipPipPipPip
  • The Undisputed
  • The Undisputed
  • 3,460 posts
  • LocationRandis IV

Posted 05 June 2020 - 12:03 PM

View PostERSmurf, on 05 June 2020 - 11:12 AM, said:

@VonBruinwald: I must have bothched the math, because in a sample of 23, this IS not true. If I did not make errors, the match average is roughly 165 (actually 164 and very high decimals)... and the first deviation is about 77,79... out of the samples matches, only SIX fall within that range, nowhere near 70%... *puzzled*


View PostERSmurf, on 05 June 2020 - 11:35 AM, said:

@Brauer, VonBruinwald: Thank you very much both... and here is the sequence: 94, 198, 189, 77, 178, 152, 72 129, 164, 127, 237, 117, 346, 269, 87, 200, 269, 307, 95, 94, 92, 81 and 214.


You got your average and SD right but you're not using your SD right.

You add and subtract your SD from your average to get the range:
Lower limit = 165 - 80 = 85
Upper limit = 165 + 80 = 245

So 70% of your matches should lie between 85 and 245MS. For you that's 16 matches, or 69.5% Posted Image

As an extension 95% of your matches should lie within 2x SDs. And 99.7% within 3x.

Edited by VonBruinwald, 05 June 2020 - 05:13 PM.


#286 Clay Endfield

    Member

  • PipPipPipPipPip
  • Shredder
  • Shredder
  • 172 posts

Posted 05 June 2020 - 12:06 PM

Sounds like the playerbase actually has something some pretty decent input on your PSR calculator: Get rid of the Win-Loss factor entirely, reward PSR based entirely on matchscore.

A question that I'm not seeing asked enough though is redoing matchscore calculation. Damage dealt should not be as hefty a factor in determining matchscore as it presently is. LRMs can dish out something like 700 damage over a minute long engagement to kill a Assault mech; whereas a pinpoint brawler can kill that same Assault mech with 250 damage in a fraction of the LRM boat's TTK.

With the current matchscore metric being what it is, the guy leeching target data in the back for using lock-on tracking missiles is getting almost 3x the matchscore for being less effective and less skilled than the pinpoint brawler. The current matchscore system rewards the less skilled player for mooching, and given how matchscore ties into calculating PSR the less skilled LRM boat player advances through the ranks faster than the more skilled pinpoint brawler player.

Ergo: The current matchscore system defuncts your proposed PSR system.

#287 ERSmurf

    Member

  • PipPip
  • Knight Errant
  • 20 posts

Posted 05 June 2020 - 12:13 PM

@VonBruinwald: Ahhh, thank you. The problem was that I got an impression from an example that deviation is equally split on both sides of the avarage... So I counted scores that were 39 or less up or down, when I should have been counting 78 or less up or down... :D Thank you very much!

#288 Andrzej Lechrenski

    Member

  • PipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 96 posts

Posted 05 June 2020 - 12:42 PM

Just remove the tier number from the interface, or else change it to something without a hierarchy (perhaps random colors or the names of Star League Era lore figures). As long as there is an ordinal number there, people are going to keep thinking this is a system to be gamed for "reward" and "punishment". It is not. This is about improving the quality of match making. Period.

As for people losing PSR on a great match that they happened to lose... bummer. It happens to everybody. But I don't care if you had a great match last night and how much you think you need to be "rewarded" for it when I drop with you. What I care about is your ability to bring something to the table to help me win the match. Your W/L is the best indicator of that, not your sense of entitlement.

#289 Deus X Incognita

    Rookie

  • Philanthropist
  • 2 posts

Posted 05 June 2020 - 12:43 PM

[Redated]

The Majority of people "being fine" with a reset said that it should be changed, to not have the major influence be Win / Loss.
And what do you do? You make it an even bigger influence.

Also it still isn't adressed that match score is depenant on the Mech weight, What the hell.
So in a match where a light plays well and a Assault plays mediocre, the Assault Mech will probably go up more. Meaning Assault Players will rise in tier more than light players. And therefor well Light players will be queued with the less able people.

And it still isn't addressed that match score is too highly dependant on damage (I guess if I want to go up, I just play my Assault Missile Boat) - someone doing 200 Damage to the Center Torso should have better score than someone else who does 300 damage divided on the whole mech.

Or maybe reward playing to objective more too?

I know, its not easy to address everything, but I was expecting at least something, anything, some effort to adress at least some problems.

But you did not even - as several people pointed out - make a win for low scoring players a skill down as a point for making things more even.

No, its just you took away the effort of high scoring losing players as well.
Its as if you even did'nt read your previous thread.

I stand by my statement.

#290 Negat1ve Nancy

    Member

  • PipPip
  • 32 posts

Posted 05 June 2020 - 12:46 PM

View PostAndrzej Lechrenski, on 05 June 2020 - 12:42 PM, said:

Just remove the tier number from the interface, or else change it to something without a hierarchy (perhaps random colors or the names of Star League Era lore figures). As long as there is an ordinal number there, people are going to keep thinking this is a system to be gamed for "reward" and "punishment". It is not. This is about improving the quality of match making. Period.

As for people losing PSR on a great match that they happened to lose... bummer. It happens to everybody. But I don't care if you had a great match last night and how much you think you need to be "rewarded" for it when I drop with you. What I care about is your ability to bring something to the table to help me win the match. Your W/L is the best indicator of that, not your sense of entitlement.


Of course win/loss is the best indicator of a solo's performance in a que with groups, lol.
O Wait, it isn't.

You know the thing that queue's with groups have in common?
They all died, Faction Play even died twice.

Now that this que has groups, I wonder what will happen next?

Edited by Negat1ve Nancy, 05 June 2020 - 12:47 PM.


#291 5th Fedcom Rat

    Member

  • PipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 893 posts

Posted 05 June 2020 - 12:50 PM

I like that "ranking within the team" model best. That's always how I have evaluated my performance in this game, rather than match score.

#292 Brauer

    Member

  • PipPipPipPipPipPipPipPip
  • 1,066 posts

Posted 05 June 2020 - 01:05 PM

View PostNegat1ve Nancy, on 05 June 2020 - 12:46 PM, said:


Of course win/loss is the best indicator of a solo's performance in a que with groups, lol.
O Wait, it isn't.

You know the thing that queue's with groups have in common?
They all died, Faction Play even died twice.

Now that this que has groups, I wonder what will happen next?


I agree, groups are a problem. But comparing players' WLRs over a large enough number of matches will overall show where they should sit. It won't sift between players who exclusively play in decently strong groups and players who solely drop alone but trying to rely more on match score or another artificial indicator opens up a lot of opportunities for the equation to be wrong and/or gamed.

A better solution would be to get groups out of Quick Play, reset PSR, and use a system where the W/L condition has a large impact on PSR as over time that will sort people most effectively.

#293 Andrzej Lechrenski

    Member

  • PipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 96 posts

Posted 05 June 2020 - 01:21 PM

View PostNegat1ve Nancy, on 05 June 2020 - 12:46 PM, said:


Of course win/loss is the best indicator of a solo's performance in a que with groups, lol.
O Wait, it isn't.


Yes. Yes it is. Help the team win. That is all that matters to the 11 other players on your team.

Quote

You know the thing that queue's with groups have in common?
They all died, Faction Play even died twice.

Now that this que has groups, I wonder what will happen next?


Cause and effect:
[color=#FFFFFF]Posted Image[/color]

Group queue did not do this. Group queue died because of it.

#294 Wesxander

    Member

  • PipPipPipPipPipPip
  • Field Marshal
  • Field Marshal
  • 319 posts

Posted 05 June 2020 - 01:44 PM

Match maker scoring system is broken. Example in match where I used no strikes I did more damage than a team mate AND scored 1 kill where they scored none. MM decided I scored nearly 40 points less than that person. CARE TO EXPLAIN THAT? Of course not. Players btw should not score more points for doing less damage and no kills. On the same hand MM detests conquest missions on large maps for players that might be doing caps during the mission. MM thinks people who use voice comms as well to coordinate strats are no different from a bozo who does not once work with the team. MM does not penalize farming mechs IE guys who intentional go off suicide into the enemy time after time. MM thinks throwing an elite light against 4 elite assault players is good match making. None of that has been adjusted. Yet resetting PSR will make a bad MM point scoring system go away. Would you like people to post video games where they out performed their team mates but MM says nope they played better and deserve a better MM score?

#295 Eurystheus

    Member

  • PipPipPipPipPip
  • The Grizzly
  • The Grizzly
  • 131 posts

Posted 05 June 2020 - 02:09 PM

I don't have strong feelings about this either way. But I have run into players who will just sit and do nothing, looking like AFK, but actually just sitting there, who won't move until the match is nearly over. I think moving down with a low match score, even if your team wins, would discourage this kind of behavior.

#296 Bowelhacker

    Member

  • PipPipPipPipPipPipPip
  • Hero of Marik
  • Hero of Marik
  • 922 posts
  • LocationKooken's Pleasure Pit

Posted 05 June 2020 - 02:12 PM

The more I think about it the more this stuff should have been looked at after a great big overhaul of how match score is actually calculated.

Edited by Bowelhacker, 05 June 2020 - 05:39 PM.


#297 General Solo

    Member

  • PipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • 3,625 posts

Posted 05 June 2020 - 02:37 PM

If only match maker did a better job, them and us would never have to meet.
O well too late now that the horse has bolted aka low population,...
To fix the leaky bathtub with an even bigger leak haha

#298 Anomalocaris

    Member

  • PipPipPipPipPipPipPip
  • Bad Company
  • 671 posts

Posted 05 June 2020 - 02:42 PM

View PostAndrzej Lechrenski, on 05 June 2020 - 02:15 PM, said:

[Redacted]


I think you're missing the point here. Look past the bitterness (a lot of us are still pissed about the merge) and consider what Brauer has been dishing out. The meta is a strong group of players that drop together in a 4 man, ideally with strong complimentary mechs.

Regardless of your skill level or mech choice, if you drop solo in the merge queue today (or in some cases even if you drop light in a group) you are not bringing your A-game to support your team. It's like taking a weak mech or a bad build. No matter your skill level you would have done better if you brought a good mech/loadout. The same thing applies to dropping solo vs. group now.

Really, really think on that one. Players who drop solo are handicapping themselves and their team - I don't care if its Brauer, Magic Pain Glove or ASH. If they drop solo they are not helping the team as much as if they drop group. The only problem is, the merge queue is designed to force 7-8 solos into a match with groups. Now ask yourself how scoring those players on their skill in that mismatch is going to be useful for building competitive matches. Particularly if you use a primary condition that is dominated by the skill of the group.

Frankly, even using relative positioning within a match gets borked with groups in the queue, but at least if you get a bad group (increased chance of loss) you could play well enough to demonstrate your skill.

#299 Warning incoming Humble Dexterer

    Member

  • PipPipPipPipPipPipPipPip
  • Bridesmaid
  • 1,077 posts

Posted 05 June 2020 - 02:44 PM

View PostKamikaze Viking, on 04 June 2020 - 08:43 PM, said:

Posted Image

This picture has gathered the most likes in the least words.

The player ranking could just be the "average rank within team" really : A number between 1 and 12.

And only using the last ~100 matches to keep it relevant.

Edited by Warning incoming Humble Dexterer, 05 June 2020 - 03:10 PM.


#300 The Teddy Bear

    Member

  • PipPipPip
  • Bad Company
  • Bad Company
  • 87 posts
  • LocationSomewhere cuddling

Posted 05 June 2020 - 03:09 PM

View Post5th Fedcom Rat, on 05 June 2020 - 12:50 PM, said:

I like that "ranking within the team" model best. That's always how I have evaluated my performance in this game, rather than match score.


Lols, you have been playing with teams for a pretty long time back. Now you should be giving solo playing a good chance as well. I can only guess right now what the result would be in the long run?





7 user(s) are reading this topic

0 members, 7 guests, 0 anonymous users