53 replies to this topic

[Telmasa]

EDIT: UPDATE! OFFICIAL WORDS ON PSR FOUND HERE: http://mwomercs.com/...-tiers-and-psr/


source (twitch chat just hanging out today): http://pastebin.com/11vj3JFf

Since it seems like the community at large still has lots of questions about PSR, here's some conclusions I believe I can draw from this informal chat that maybe y'all will find useful information:

So what PSR is *not*:

• PSR is not an achievement system
  
• PSR is not a rewards system
  
• PSR is not relative to the performance of other players
  
• PSR is not a fluid ladder ranking system

What PSR is:

• A metric used to place you the player into a better quality match.

I'd ask for you to *please* note that does not mean every game's score is going to end up 12v10! The game score doesn't linearly correlate with the quality of the match - it doesn't have to be a 'close game' to be a 'GG'




Other points about PSR:

• PSR uses win/loss to determine where you move, and uses match score to determine how far you move; only in more extreme cases will you move up on a loss
  
• By using match score, it's not just about damage - lots of spotting, lots of assists, and anything else that can boost match score matters
  
• The better you perform, the less games it will take for you to move up in tier - as Paul said, the "average baseline" here is that at a 50/50 win/loss rate it'll probably take around 2500 games to move up, assuming you perform decently that entire time, and would probably take around 6-7 months of regular gameplay (which I would assume is based upon average server statistics)
  
• While *technically* possible to perform so well that you could climb in PSR despite losing more often than you win, this is not very likely, especially since consisently performing very well will tend to help your team win more often than you would lose otherwise.
  
• If for whatever reasons your win rate starts climbing abnormally high, you'll be placed against other players accordingly - increasing the difficulty - so in that sense the PSR system naturally 'balances' players

TL/DR: Look at PSR as the means of seeing where in the MW:O player ecosystem you fit most comfortably and enjoyably, and not as a e-peen measuring stick.



edit: P.S. I don't reddit, if someone wants to reach out to that community with a re-post/copy-paste, go right ahead.

Edited by Telmasa, 24 September 2015 - 09:25 PM.

[Deathlike]

PSR is actually just a grind indicator (based on the # of matches you play, assuming you win more than you use).

So, it doesn't really work as intended.

Source:
https://www.reddit.c..._to_settle_out/

Edited by Deathlike, 23 September 2015 - 02:03 PM.

[Dolph Hoskins]

I got your PSR right here:


Don't lie. This is what it's all about.

[Telmasa]

@ Deathlike: I can't totally agree with that assessment. Yes, the number of games matter, and statistics being statistics, things will average out better & more accurately the larger the sample size in question - which here is the number of games played.

But that doesn't really mean PSR is just a measure of grind.

Edited by Telmasa, 23 September 2015 - 02:13 PM.

[AEgg]

On average 2500 games to move up?

Well that explains my tier. I've been playing at least weekly since closed beta and I only have 1810 matches played since the stat reset (1100 on record in archived stats).

[Deathlike]

Telmasa, on 23 September 2015 - 02:13 PM, said:

@ Deathlike: I can't totally agree with that assessment. Yes, the number of games matter, and statistics being statistics, things will average out better & more accurately the larger the sample size in question - which here is the number of games played.

But that doesn't really mean PSR is just a measure of grind.

I should be more clear... it's for "successful grinding". If you're hovering @ 50-50 on the W-L ratio, you're probably not going to make actual progress.

[Greyhart]

I can confirm that if you win but score nothing i.e. no damage no spotting nothing at all. your PSR will remain the same.

So barring team damage it should be impossible to go down on a winning team

[Telmasa]

Deathlike, on 23 September 2015 - 02:30 PM, said:

I should be more clear... it's for "successful grinding". If you're hovering @ 50-50 on the W-L ratio, you're probably not going to make actual progress.

Well doesn't this leave "players who succeed for a larger number of games" > "players who succeed but for a lesser number of games"? Like, genuinely working as intended lol

Calling it a measurement of success is more accurate. Success ought to be about both quantity & quality together, not just the latter.

Which isn't to say that "being a lesser tier means you suck"...well, maybe it kinda does...but like I pointed out, that just means you'll fit more happily and comfortably in that tier that PSR says your in, and you'll have a better time along with the folks you'll be playing with.
Everybody wins!

Honestly, the only problem I'm seeing is those players who want a ladder ranking/rewards system that we'd probably see if/when Solaris mode comes around.

I'd say CW too could somewhat fit that need, just with the possession of planets taking over the presence of a ranking system; which would also mean that planet ownership has to have a reward.

But that's leaning over into a "CW discussion" kind of thread, of which there's been plenty, so I'll cut my own derailment off here.

Edited by Telmasa, 23 September 2015 - 02:42 PM.

[sycocys]

It actually has quite a lot to do with your team unless you are pulling 8-12 kills every match.

If you are anyone else though and your team under-performs the level you are, then you simply get a lower score. There's no way to dispute that - when you miss out on 8-12 assists because 6 people on your team put up a combined 250 damage, then you get boned because they couldn't carry themselves.

[Alistair Winter]

sycocys, on 23 September 2015 - 02:41 PM, said:

It actually has quite a lot to do with your team unless you are pulling 8-12 kills every match.

If you are anyone else though and your team under-performs the level you are, then you simply get a lower score. There's no way to dispute that - when you miss out on 8-12 assists because 6 people on your team put up a combined 250 damage, then you get boned because they couldn't carry themselves.

This is the flaw of the PSR, I think. Also, light mechs do not get the same match scores as heavies and assaults on average. Someone who plays exclusively light mechs is going to have a lower PSR than someone who plays exclusively assault mechs, assuming they are roughly equally skilled. Getting 1000 dmg with an assault mech is something a lot of people can do. Getting 1000 dmg with a light mech is something far fewer people can do.

[Telmasa]

sycocys, on 23 September 2015 - 02:41 PM, said:

It actually has quite a lot to do with your team unless you are pulling 8-12 kills every match.

If you are anyone else though and your team under-performs the level you are, then you simply get a lower score. There's no way to dispute that - when you miss out on 8-12 assists because 6 people on your team put up a combined 250 damage, then you get boned because they couldn't carry themselves.

Alistair Winter, on 23 September 2015 - 02:43 PM, said:

This is the flaw of the PSR, I think. Also, light mechs do not get the same match scores as heavies and assaults on average. Someone who plays exclusively light mechs is going to have a lower PSR than someone who plays exclusively assault mechs, assuming they are roughly equally skilled. Getting 1000 dmg with an assault mech is something a lot of people can do. Getting 1000 dmg with a light mech is something far fewer people can do.

Except that it affects your enemy team as well. It's easier for you to cope with when the enemy performs at about the same level.

Also, 1000 damage is steep no matter what mech you're in - I don't agree that's something a "lot" of people can do, especially not at lower tiers. I would call 600-800 a better number, and that that number is achievable with any kind of 'Mech.

Plus, Assaults are FAR more vulnerable to your team's play than light mechs are, just by virtue of raw speed. Both types of mechs are min/maxed out more in terms of armor/firepower vs size/speed, and thus are more difficult to play compared to mediums/heavies. And Light 'Mechs, for the same reason of speed, have a much easier time flanking, scouting, and doing other similar actions - as I pointed out in my main post, match score, not just damage, is what matters, and match score takes those actions into account.

So if you follow through with thinking about it, like I said it genuinely works as intended.

Edited by Telmasa, 23 September 2015 - 02:56 PM.

[Hydrocarbon]

Telmasa, on 23 September 2015 - 02:13 PM, said:

But that doesn't really mean PSR is just a measure of grind.

Exactly. Grinding a PSR is possible if you pick up skill as you go, since MWO isn't exactly tic-tac-toe. If you're like a block of wood when it comes to learning, grinding into Tier 1 may be next to impossible.

[sycocys]

My experience has been that it definitely does not seem to affect both teams the grand majority of the time.

You'd think it would, but it simply does not - more often than not one side or the other gets a half a team of people that can't put up 100 damage and the other side gets a full team of people that put up a minimum of 250 each.

And either way you look at it, my PSR should be based upon MY performance alone, it shouldn't rely on the performance of 11 other people. If I can consistently put up the numbers that I do, while half the players MM puts on my team consistently put far worse numbers then I'm in the wrong bracket and should be working my way out instead of working my way to Tier 5 because people can't manage to get 100 damage in their 60+ alpha assault mechs.

Having other people so heavily influence my personal score is quite bothersome because all it ends up doing is keeping me away from the tier which would provide better competition.

[Jin Ma]

i've moved up in a losing match with 300 points (600 dmg match?) i think

Deathlike, on 23 September 2015 - 02:30 PM, said:

I should be more clear... it's for "successful grinding". If you're hovering @ 50-50 on the W-L ratio, you're probably not going to make actual progress.

In a real matchmaking system that would be the case.
Except paul confirmed that even with a 50/50 you can make progress.

Iv been playing seriously for the past 20+ matches. and got 19 green uparrows, and 1 equal sign. Losing 2/20 games

still kind of a grind though. I don't mind tryharding, but its kind of tedious

Edited by Jin Ma, 23 September 2015 - 03:08 PM.

[Zensei]

I cant help but to wonder how much spending has to do with this, its not a secret that some new mechs were very OP, and then nerfed, so if you stayed with the 'package' pace, you would gather a higher teir score in the end. And what about the 'cheaters' where they had a most certain advantage, are those games part of our histories now? I'm here since beta, I didnt buy a Founders and I did screw around a lot testing and just having fun, I dont see any of that being figured in. Pugs arent really getting any love here for a lott of reasons, I'm not on a team comm, so I dont move with tthe unit, it goes on and on.

It's not a ltp issue, its not a get a team issue, not after 7000 games part time. Its just hard seeing that 4 after playing this much. Just making a statement not looking for advice.

[Zainadin]

Telmasa, on 23 September 2015 - 02:49 PM, said:

And Light 'Mechs, for the same reason of speed, have a much easier time flanking, scouting, and doing other similar actions - as I pointed out in my main post, match score, not just damage, is what matters, and match score takes those actions into account.

WRONG Damage is king in Match Score! and a light mech (except for the AH cheaters) can't do the damage to get 600+ damage a match.

[MischiefSC]

I took Feb to august off, have mostly only played CW since then and almost exclusively before. I am still currently at about 80% up tier 3. Every single match I've played since it came out, win or lose, has shown going up. Sometimes visibly so.

I'm not sure what the criteria is but at the rate it's moving you can grind mid third to tier 2 in a day or two of matches. 2500 matches to change tiers doesn't make sense to me. I'll probably have at most 500 pug/group queue matches since January and most of those were grinding basics on Urbies and other bad chassis.

I think his estimate off.

[Torgun]

I have to agree Paul's estimation seems off. I've been playing quite infrequently the entire year and might not have logged more than 200 matches with about 1.2 W/L ratio. I'm one third into 3rd tier.

[beerandasmoke]

MischiefSC, on 23 September 2015 - 03:32 PM, said:

I took Feb to august off, have mostly only played CW since then and almost exclusively before. I am still currently at about 80% up tier 3. Every single match I've played since it came out, win or lose, has shown going up. Sometimes visibly so.

I'm not sure what the criteria is but at the rate it's moving you can grind mid third to tier 2 in a day or two of matches. 2500 matches to change tiers doesn't make sense to me. I'll probably have at most 500 pug/group queue matches since January and most of those were grinding basics on Urbies and other bad chassis.

I think his estimate off.

Wins count a for a massive amount Mischief. Somebody had a vid of Sean Lang playing and he got a match score of 106 and still got a greenarrow on a win. He is a tier 1 player. It seems to be wins followed by damage dealt. If you get a loss then the average seems to be around at least 500+dmg dealt not too get the red arrow.

This really screws soloplayers who are at the mercy of their team while at the same time rewards people who run in large 8+groups. It looks as though a subpar player could be carried to tier1 just by being in a decent group of players who drop together. PSR is not a very good indicator of individual skill in my opinion.

Edited by beerandasmoke, 23 September 2015 - 04:04 PM.

[Valdherre Tor]

I'm have not played since the PSR was shown to us. I plan to stay out of the solo ques for at least one week to see if my skill rating goes up or down.