41 replies to this topic

[Inner Wolf]

We just had a match where we lost; they capped our base and I tried to block the cap. (Three heavies on our base and me in an ACH-Prime = gang ****.) My cheetah ain't gonna walk right for weeks.

I had 1 kill and did 430 damage. The PSR arrow still went down. I really hoped it would break even and not go down. It's confusing to a weak mind like mine.

[Void Angel]

Rogue Jedi, on 18 August 2016 - 08:48 AM, said:

Please stop sugesting that the PSR bar is an experiance bar, it is not
in the last 6 months my bar has hardly moved, it has been between the R and the number of my current tier for more than 6 months.

I play Light Medium and heavy Mechs, I am hopeless in most assualts, I find meta play to be boring so tend to avoid it, I tend to prioritise teamwork over damage output, and the group I usualy play with are mostly higher tier than I.


If I spend a lot of time playing solo my bar does go up but playing with my group I will never go up in tier and I am fine with that

Well... it's a metaphor, and it's almost entirely correct. Because of the way PSR allows more opportunities to advance than to fail, players will tend to move upward - it's far more difficult to get a score high enough to avoid falling on a loss than it is to score low enough to avoid being carried upward by your teammates, for example. Since players are selected at random each match, this means that even abysmal players will trend upward.

Over time.

That's the point at which the metaphor fails, and at which the people who use the metaphor to criticize the system fail to understand either concept. It's like the old chestnut about monkeys in a room, banging on typewriters, eventually producing the complete script to Hamlet. Sure, they absolutely would - but not within the expected lifespan of the universe. PSR is designed to be an "experience bar," but the experience measured is intended to be teamwork, not time or skill.

That's why we have the PSR rating and algorithm^* instead of just using a modified Elo. We already had an Elo system, but it just wasn't working for the very top (or bottom) players. I know this firsthand; one of them used me as an Elo anchor just so he could get into matches in under 5 minutes - if at all. Many others resorted to the same thing after they'd wait five minutes only to find that even with allI the safety valves opened, the matchmaker would fail to find them a match. PGI tried various things, but wasn't able to find a balance between match speed and seal-clubbing prevention. Either the elite players (and crash helmeted special children) would be placed into matches with vast differences in player skill, or they'd have to wait forever and a day between each and every match.

But PGI seems to have noticed something along the way trying to balance Elo - when mismatches occurred, one elite player mixed in with a bunch of newbies seldom fared well against an enemy team comprised of moderately skilled players whose average Elo was equivalent to the mixed team. This even tended to be the case (based on my anecdotal experience and conversations with high-level players) when there were several top-level players on the mixed team, and perhaps one top-level with the homogenous enemy. Players who were experienced enough to work together made better teammates for high-level players than low-skilled cannon fodder.

And so we have PSR. It's designed so that eventually every player is expected to be in Tier 1 or 2 (this is from the official PGI posts,) and that's ok, because it's not a skill ranking system at all - it only tries to measure teamwork. This focus allows top-level players to reliably find matches while still preventing them from fighting absolute newbies; it balances ease of matchmaking against skill matching fidelity by emphasizing teamwork as well as individual results. This makes it a good system for what it's intended to do - and most of those criticizing it simply ignore history when voicing their complaints.

*^{: You're welcome, Mailin.}

[justcallme A S H]

I went from 3/4 the way up T2 and into T1 in I think maybe 50 matches? I've barely played QP in the last 2 months. But then I'm cranking out 600dmg+ in most games, 4-5 KMMD & lots of component destruction.

So if the OP can up his damage from 200-250 and get it to over 500, net a few kills and KMMD - you'll find the climb is markedly quicker.

That said I find it not all that relevant as when I was in T2 in solo queue, I had T5 players in there as well (that's been confirmed as well on 3 occassions myself and by other players). So the Match Maker clearly doesn't take PSR into account when finding matches anyway.

Edited by R31Nismoid, 19 August 2016 - 02:51 PM.

[Inner Wolf]

Could it be...that there aren't enough players available in each tier to pull names for each match equally from the hat?

No sarcasm...serious question.

[justcallme A S H]

Without a doubt that's why
I suspect somewhere along the line the MM was altered in the background to expand after a set amount of time (which, even in a 2-3min search time, isn't long). So yeah, it's not good having such a broad stretch lumped into one. But for the sake of getting a game when there is a low population - what else do you do?

I mean some people I watch (after I die) are wandering around aimlessly, loadouts that have zero synergy, players constantly shutdown for no gain... It's obvious they are T4/5, and nothing against them - I couldn't care less, long as you have fun.

But that's just it - is it a fun experience to be flogged by someone who's a fair notch ahead in game play/aim and can pick you apart in a short space of time?

[Whiskey Dharma]

Necro'ing this thread with an update: I did finally make it to Tier 4!

It took about 260 games.

[Aleksandr Sergeyevich Kerensky]

Whiskey Dharma, on 12 September 2016 - 02:22 AM, said:

Necro'ing this thread with an update: I did finally make it to Tier 4!

It took about 260 games.

Grats bro!

[JC Daxion]

mailin, on 18 August 2016 - 06:48 AM, said:

PSR really is more of an experience bar. Eventually everyone, even really bad players, will wind up in Tier 1. I wish PGI would redo the algaerhythm so that most players would wind up in T3 and only the truly great could be in Tier 1.

This is not true.... I've played thousands of matches, and my PSR has barely budged. I don't think mine has moved more than 1/8 inch since PSR was released, (less than the width of my mouse pointer). in fact it has moved so little, i don't think it would even raise a level after 100k matches.. And I've improved, as my KDR is my main indicator, along with win/loss ratio.

Even my new account, in clans that i started last year, made it to a little past where i am on my IS account, and one i hit that point, it's barely budged since last feb.. I also do better over all in my clan mechs, i guess they just suit my play style a bit more.

So i am not sure how bad players will make it to t1, unless they are playing in groups.. because i'm a decent player, and i don't think i will ever get out of T3.. Unless ED improved my game play, because i don't play Alpha warrior online.

Most players are already T4 and T3 anyway


But seeing T1 is still the smallest % of players, t1's and 2's would still be playing with 3's. I am in matches all the time with T1 players... And i notice who is better than i, fairly easily.. I will see a few T1's then the rest of are the lowly t3 players

justcallme A S H, on 19 August 2016 - 02:50 PM, said:

That said I find it not all that relevant as when I was in T2 in solo queue, I had T5 players in there as well (that's been confirmed as well on 3 occasions myself and by other players). So the Match Maker clearly doesn't take PSR into account when finding matches anyway.

as i said,, Top tiers are the smallest % of players.. when you are getting T5's in your match, it is because there are not enough higher tiers to fill the match.


Personally i'd like them to add a Play with my tier only check box.. My guess is people would complain that it takes to long to get matches with it checked if you are t2 or t1





Also, IMO PSR is a better system than ELO.. though I did much better under ELO.. that system i was typically on the top 3-4 people on the team.. Now if i am in matches with a bunch of T1's it ain't happening, unless i just have a very good match verse them which is rare..

Edited by JC Daxion, 12 September 2016 - 12:03 PM.

[mailin]

JC Daxion, on 12 September 2016 - 11:54 AM, said:

This is not true.... I've played thousands of matches, and my PSR has barely budged. I don't think mine has moved more than 1/8 inch since PSR was released, (less than the width of my mouse pointer). in fact it has moved so little, i don't think it would even raise a level after 100k matches.. And I've improved, as my KDR is my main indicator, along with win/loss ratio.

Even my new account, in clans that i started last year, made it to a little past where i am on my IS account, and one i hit that point, it's barely budged since last feb.. I also do better over all in my clan mechs, i guess they just suit my play style a bit more.

So i am not sure how bad players will make it to t1, unless they are playing in groups.. because i'm a decent player, and i don't think i will ever get out of T3..

I stand by my statement that all players will eventually achieve Tier 1. This is because the amount of psr that you lose with a loss is less than the amount that you can gain with a win. If it were completely balanced most players would eventually settle in Tier 3, but you can gain more with a win than you can lose with a loss, and even the newest players get better with time. So, eventually everyone will end up in Tier 1.


As far as why your psr hasn't changed much it's important to remember that when PGI put psr into place they looked at game stats, if they existed, for up to I believe 6 months prior to the implementation. When I started the psr journey I was Tier 4, and now I am in Tier 2 and on my way to Tier 1 and I have just over 22 thousand drops. The xp bar still moves, it just moves really slowly if you had a lot of drops before the implementation. But before the implemetation I didn't know that certain factors would matter for any kind of rank. For example if I was the last one left alive in a Spider 5D and there were a lot of enemies left, I would consistently switch my ECM to counter charge into the midst of the enemy and stop moving. I could have kept firing and possibly killed a couple more enemies, but I wanted to end the game quickly so that all players could start a new drop. In hindsight I believe that that behavior really negatively effected my psr, so once psr was implemented I stopped doing that.

[TooDumbToQuit]

My bar was moving but it has either slowed down or stopped now. I was playing my Kodiaks and finally getting my damage to 500-900 per game.

Then the light sale came and I got some. And the Mastery Bundles went on sale and I got 3 sets of lights. So I've been doing pilot trees on lights and I'm not really able to play and I've only played 65 matches since season 3 started. And some of those like the 5 assault games were with my Unit where I'm facing 12 low Tier players.

Whatever...I make jokes about it but I really do not care that much. And I've seen some of the same names that I saw 5 months ago so I'm not the only one not roaring up the ladder.

P.S. Russ, funk you for that rude Tier 6 e-mail!

I can do my silly Flamer builds and just have fun. I even have two Griffins set up for LRMs now. And I've been working on their pilot trees, lol.

Edited by LikeUntoGod, 12 September 2016 - 07:07 PM.

[JC Daxion]

mailin, on 12 September 2016 - 06:35 PM, said:

I stand by my statement that all players will eventually achieve Tier 1. This is because the amount of psr that you lose with a loss is less than the amount that you can gain with a win. If it were completely balanced most players would eventually settle in Tier 3, but you can gain more with a win than you can lose with a loss, and even the newest players get better with time. So, eventually everyone will end up in Tier 1.

Or it could more likely be that people with 20-30k drops are actually better than the typical t3 player? Like i said, I can tell when people are in t1 or t2.. Most people are in t3.. I am not sure why people think other wise.

Also, there can still be a large difference in skill in t1, and still be better than a t3 player. There are players with 1k's of matches that can't even get out of t4.. there are Pros in T1, and there are also players that are better than average while still very good players, they are not close to what the top people are. Like everyone in the NFL is a great collage player.. But only a handful of them are in the "Hall of Fame" Which again is the huge disconnect many people have.. Ohh this guy sucks, why is he in T1? well he is better than a typical T3 player.

Though i will say, damage is not the be all and end all.. I know i have had tons of matches that i only get like 300 damage and have 4+ kills, and totally dominated in a match. I often feel like when i have a 500+ damage game i actually played like crap because my aim was off. Maybe a slightly above average player has dropped in a kodiak 5k times.. that could do it.. Then again, they could just be a pretty decent assault pilot anyway.

Many of the worst rolls i have been on the receiving end, or the winning side, only 1 or 2 team members had over 400, the rest were in the 200 range and a few under.

IMO the only way a bad player would ever get to T1, is if they either A group, Or B, just play the most OP mechs in the game, and when that meta changes change with it.. and even then if you can put out 400-500 damage, you have to be semi good anyway.. You don't realize how many people never even get matches that good, or if they do it's like 1 in 100.. and in those cases PSR doesn't budge, or it drops slowly..

[SlightlyMobileTurret]

LikeUntoGod, on 12 September 2016 - 07:06 PM, said:

I even have two Griffins set up for LRMs now. And I've been working on their pilot trees, lol.

LRM griffins...

Edited by Keshav Murali, 12 September 2016 - 11:50 PM.

[Whiskey Dharma]

mailin, on 12 September 2016 - 06:35 PM, said:

I stand by my statement that all players will eventually achieve Tier 1. This is because the amount of psr that you lose with a loss is less than the amount that you can gain with a win. If it were completely balanced most players would eventually settle in Tier 3, but you can gain more with a win than you can lose with a loss, and even the newest players get better with time. So, eventually everyone will end up in Tier 1.

I'm again leaning on my WoT experience here, but I can assure you that

1. many players are not winning half of their games, potentially offsetting the increased gain from wins, and
2. some players DO NOT get better over time.

Exhibit #1: if you filter the season 1 or 2 leaderboards by # of matches, you will see some WR < 1.0 at the top.

Edited by Whiskey Dharma, 13 September 2016 - 02:05 AM.

[Hunka Junk]

I was t5 for about 3 months. Looking back, I feel pretty confident in saying that doing dailies on a whole herd of different mechs will result in slower psr improvement than sticking with one that seems to be working well and getting in a groove.

I was still t5 during the light event, then the assault event rolled in and bam, I was into t4.

Then the weight of the world slowed progress slightly, then my cyclops squad arrived. By the end of the cyclops event, I was a sliver away from t3. Have since broken through, then got put back, then broke through again. So I decided to stick with assaults and heavies for a stint just to get some color on the t3 bar. It's there, but I can still see myself getting bumped back again.

So 3 months to rise one tier playing fruit salad. Then more like 3 weeks to blow through the next tier playing assaults and heavies. The difference is in the damage. I'm way more consistent with dmg in the bigger mechs, and rightly so. A bad game in a locust is gonna result in less damage than a bad game in a kodiak. For the most part.

Honestly, the biggest draw for me has been to reach these fabled levels where the lrm boats disappear. I'm still waiting.

I'd say the comms are more civil now that I'm t3. Fewer dead men talking about how their whole team sucks because they didn't join whatever failed attempt that Mr Deadman thought was a good idea just before he got his buttocks handed to him.

More people lose with dignity than wailing. that's nice, and the competition level overall must've increased or I wouldn't have bobbled back and forth between 3 and 4 so many times.

[Vasili Kerensky]

PSR is a huge load of crap and an excellent example of bad design to be avoided in games. It only cares about
-- Victory or Defeat ==> how many noobs the matchmaker places in your team/how many uber players it places in the enemy team
-- Damage Output ==> Biggest load of crap to hit this game as a metric.
You have to grind 3 different 'mechs of each chassis, whether you like the loadout and you can handle it or not. You get punished for this.
You have to build a drop deck of 4 Mastered 'mechs = 12 'mechs ranging from lights to assaults. You get punished for this.
You may want to play lights or lighter mediums or casual builds or experimental builds or whatever. You get punished for this.
The game rewards you only if your random pug wins and only if you run the class and model of 'mechs with which you are most comfortable with and manage to do ungodly amounts of damage consistently.

PSR = Huge Load Of Crap.

Edited by Vasili Kerensky, 13 September 2016 - 06:32 AM.

[SlightlyMobileTurret]

Honestly, DMG output or KMDD/Kill/Component destruction? these are the only possibilities with which match score can be measured.

At this point, yes, Lights are not as rewarded, but "protected lights" (and protected mediums for mediums) is exclusive and bumps Cbills, XP and matchscore, and the scouting bonuses are so easy to get in lights compared to muh 60 kph assault mech.

Have a better method? Ping Russ on twitter. Don't? Don't call the dmg output as a ******** way to measure performance.

I like salt, but this is just meaningless (and it's on Tuesday rather than monday)

[Aleksandr Sergeyevich Kerensky]

Vasili Kerensky, on 13 September 2016 - 06:31 AM, said:

PSR is a huge load of crap and an excellent example of bad design to be avoided in games. It only cares about
-- Victory or Defeat ==&gt; how many noobs the matchmaker places in your team/how many uber players it places in the enemy team
-- Damage Output ==&gt; Biggest load of crap to hit this game as a metric.
You have to grind 3 different 'mechs of each chassis, whether you like the loadout and you can handle it or not. You get punished for this.
You have to build a drop deck of 4 Mastered 'mechs = 12 'mechs ranging from lights to assaults. You get punished for this.
You may want to play lights or lighter mediums or casual builds or experimental builds or whatever. You get punished for this.
The game rewards you only if your random pug wins and only if you run the class and model of 'mechs with which you are most comfortable with and manage to do ungodly amounts of damage consistently.

PSR = Huge Load Of Crap.

Honestly, when it comes to puglandia, I think the PSR teir system is actually very good (considering we have such a small population of players). The matches are much better now then they were a long time ago, very rarelly do I see actual stomps anymore.

There are many light and medium pilots in T1 and T2. I see them in most every match I play, and they are an absolute terror on the battlefield. They know who to take down, and they do it with extreme prejudice. They get their damage and kills in, and laugh all the way to the bank!

i dont say good things about pgi often, but PSR was definatelly one of the things I am glad they implemented. Its funner for everyone all around, and getting matches is a lot faster then before. Couple years ago, que could take 15+ minutss long, and when the drop happened, it was a12-0 (or 0-12) stomp.

-------------------
-------------------

I highly recommend going into the in-game settings and checking the "Hide Tier" box, and never look back. Enjoy what ever teir you play in, and let the servers worry about who you play with.

[Vasili Kerensky]

You rarely see stomps? Stomping is the middle name of matchmaking on Tier 5 and from what I read from other players, it's the same in Tier 4 and Tier 2.

You want damage as the metric of PSR? Then give me a button to push and tell you that in this match, I want my PSR to be measured. I don't want to be measured when I play casual builds, fun builds, lights, useless 'mechs or 'mechs that I'm forced to grind and reluctant to spend a ton of c-bills on to optimize them. Really, 1 well-built LCT, 2 BLRs and 2 MADs are enough. When I play the other 2 LCTs and the third BLR and MAD I do it for the grind. I'm not going to spend millions of c-bills to give all 3 LCTs double heat-sinks and Endo-Steel. Button ==> The solution you asked for.

Edited by Vasili Kerensky, 13 September 2016 - 08:34 AM.

[Aleksandr Sergeyevich Kerensky]

Vasili Kerensky, on 13 September 2016 - 08:31 AM, said:

You rarely see stomps? Stomping is the middle name of matchmaking on Tier 5 and from what I read from other players, it's the same in Tier 4 and Tier 2.

You want damage as the metric of PSR? Then give me a button to push and tell you that in this match, I want my PSR to be measured. I don't want to be measured when I play casual builds, fun builds, lights, useless 'mechs or 'mechs that I'm forced to grind and reluctant to spend a ton of c-bills on to optimize them. Really, 1 well-built LCT, 2 BLRs and 2 MADs are enough. When I play the other 2 LCTs and the third BLR and MAD I do it for the grind. I'm not going to spend millions of c-bills to give all 3 LCTs double heat-sinks and Endo-Steel. Button ==&gt; The solution you asked for.

Dont you worry, the PSR does exactly what you want... Without you knowing it!

It allows you to play all those "casual, fun" builds you want! Thats what T5 is all about! Its a magical wonderland where you can play what ever you want, with what ever you want! It keeps the big baddies with their fully optimized mechs with all module slots filled from farming you. (Sure they can play smurfs, but at least they are in trial mechs without mech/pilot xp!!) i wont lie, I play a smurf on occasion when I want to just relax and do what ever the fork I want without having to try hard.

The thing you dont understand is, you dont need to raise in tier to enjoy mwo. Just play the game, and have fun in what ever tier your in.

(p.s. Like i stated before, there are many light pilots in T1... Damage doesnt seem to be a problem for them, they consistently get their high match scores).

[S 0 L E N Y A]

You need to keep in mind the fundamental function of the match maker is greatly compromised if the active player base is low.

In other words, if you play off peak hours then all bets are off for match making.