Jump to content

Why Don't We Use W/l Ratio Instead Of Psr...?


62 replies to this topic

#41 Nik Reaper

    Member

  • PipPipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 1,273 posts

Posted 03 June 2018 - 10:52 AM

Well, other then having a lot bigger pool of players to match up, some other games like LOL keep records of each champion separately to make MM predictions, here you'd not only need to do so for every mech but every build, or some kind of dps/burst/weight classification that the MM checks and modifies with your other stats to come to some aggregate that better represents you in that mech with that build ... and that doesn't count in map or team build cohesion... srm brawler on alpine with a few laser boats agains mostly gauss lrm spam... good times....

#42 Asym

    Member

  • PipPipPipPipPipPipPipPipPip
  • Nova Captain
  • 2,186 posts

Posted 03 June 2018 - 02:28 PM

View PostJC Daxion, on 03 June 2018 - 09:23 AM, said:



I beg to differ.. Look at the leader boards top players.. they are all very good players. good players kill, and they win.. Bad get carried along.


Anyone can go a win streak of good teams and get a positive win/loss.. Your not going to be at the top of the leader boards with many matches played if your not good.

I've had tons of matches with 1k+ damage and 5+ KMMD's and 8 kill matches, but i never broke into the top of the boards, those guys are just better. To me it just means the people on the other team were worse than me. I can often tell must by looking at the pre-match scoreboard and pick out names and know which team is going to win. They are always at the top of the boards and great players.

You know, there's a funny story in that thought: pick out the name and....

I do that and then, the entire team kills that person first. Hey, there's Bombast or El Bandito: whacked with 400 LRM's and every other weapon possible.....it's a great sport. I've individually chased "forum pundits and leader boarders" and destroyed them as easily as anyone else? Does that mean I am an expert? No, just means that anyone can beat anyone else on any given day.... That's all...even potatoes.

:I am not saying the Leader board is wrong.... Good heavens no. Good players are good players BUT, they didn't get there alone... Hundreds of other players improved or reduced where they are on the leader board. And, this is a "push PSR" system.... And, the number of matches played and where those matches are played is a huge determinant of positive or negative progress. Also, whom they played is influenced by a really messed up MM..... How many t4-5's did they massacre versus T1-2's they had to work to beat? Of course, anyone can get to T1???? So, really, what's the point of a leader board anyway??? Is in in the ballpark, Oh yes..... Would I use it to bet real money? No freaking way !

Take the top 50 players and let them fight it out in 12x12 groups and then, let them fight it out as individuals, in one division of S7....what do you think you will see....the same names in the same order? I seriously doubt there's a correlation because "groups" influence performance (+ or -) and individual skill isn't nearly the same as team skill.... Which is a greater influence?

I don't know anymore.....but, it is a good discussion.

#43 BWS2K

    Member

  • PipPipPipPipPipPip
  • 215 posts
  • LocationUSA

Posted 03 June 2018 - 02:54 PM

View PostAsym, on 03 June 2018 - 02:28 PM, said:

I don't know anymore.....but, it is a good discussion.

I think this is my new go-to response for any conversation about PSR.

#44 Popcat

    Member

  • PipPipPip
  • The Shogun
  • The Shogun
  • 74 posts

Posted 03 June 2018 - 05:15 PM

So after reading all this I have two questions

1) match score seems to be heavily weighted by damage which to me could be fair enough as destruction of the enemy does bring victory faster than just spotting an annoying. But does that mean heavy and assault mechs who carry the larger alphas are weighed in favor with this system. I get a good light player can out damage a poor assault but on average an assault has more damage potential.

2) I know good players can carry a team while the team is alive. But how often does in comedown to a 4 to one and get carried to a victory? Seems with no armor angling deflection etc. .. there is only armor spreading so I tend to find 4 on one does not go well. So it would seem the only way to carry your team is while they are still absorbing damage. Is that the case?

#45 justcallme A S H

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • CS 2020 Referee
  • CS 2020 Referee
  • 8,987 posts
  • LocationMelbourne, AU

Posted 03 June 2018 - 06:59 PM

View PostGhogiel, on 01 June 2018 - 03:58 PM, said:

Also psr does do what it's supposed to. Anyone who says it doesn't work, well it sort of does, I have topped the leaderboards for matchscore on like 5/22 seasons in alts to prove that very point. Something works as the skill level down in scrub tiers facilitates the highest farm thats possible in the game.


Topping the board in an Alt is a direct result of farming from T5-T3 users and not a lot else given Match Score is, generally, easier to farm in lower tiers given people literally stand still. I've seen many an alt-acc crack consistenly dominating games and get well over 500 AMS easily, myself included.

I mean the PSR bar will bump you to Tier 3 in 3 games if you over perform that hard but it sitll won't flick you into T1 in 50 games. It does take a while.

What you proved is that the 'seeding' matches, those first 25-50 or so that are accelerated in moving you up in PSR work. Not that PSR works.

The main issue with PSR is as I have OUTLINED HERE. You have users wih 170 average match score, which is below the average of the MWO community, making it into Tier 1. That should never be allowed.

PSR doesn't 'sort' of work IMO. It's flat out broken. It should be tightened up as again I outlined above to Paul and hopefully takes note. I'd happily wait a couple mins more for potentially reasonably skilled opponents.

#46 justcallme A S H

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • CS 2020 Referee
  • CS 2020 Referee
  • 8,987 posts
  • LocationMelbourne, AU

Posted 03 June 2018 - 07:08 PM

View PostEl Bandito, on 01 June 2018 - 07:29 PM, said:

GQ stats should really be separate from SQ stats. Silly PGI...


Yep this is another HUGE issue.

I don't know if PSR gets increased from GroupQ, if it does that is... Ugh. Again it's too easy to affect a win in GroupQ. As Wins count for too much in PSR land over player performance... It's not helpful at all.

#47 justcallme A S H

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • CS 2020 Referee
  • CS 2020 Referee
  • 8,987 posts
  • LocationMelbourne, AU

Posted 03 June 2018 - 07:16 PM

View PostJC Daxion, on 03 June 2018 - 09:23 AM, said:

I beg to differ.. Look at the leader boards top players.. they are all very good players. good players kill, and they win.. Bad get carried along.

Anyone can go a win streak of good teams and get a positive win/loss.. Your not going to be at the top of the leader boards with many matches played if your not good.


Exactly.

The amount of games I have carried to a win against, what most would say, were insurmountable odds... Is well and truly up there.

The problem is when the game is already 0-5 and you somehow pull off a win, those players who essentially suicided with sub 150 damage cause they are just not that good (yet are probably in Tier 1 or Tier 2), get a PSR reward because a win still means they at least stay even although with 150dmg i suspect you still get a RISE.

This rise comes from someone like me doing the work, not them. Thier sub-standard performance is rewarded. It quite simply should not. They should get a PSR drop in the event of a win... And that is the glaring issue with PSR as it stands and why PSR tiers do not work properly at all.

Tier 1 should be reserved for the top 1,000 players or so. Not people that are ranked in the 27,000s on Jarls list (out of 35,000). You should be in Tier 4 if you are that low in the list in all honesty.

#48 Asym

    Member

  • PipPipPipPipPipPipPipPipPip
  • Nova Captain
  • 2,186 posts

Posted 03 June 2018 - 07:25 PM

View Postjustcallme A S H, on 03 June 2018 - 07:16 PM, said:


Exactly.

The amount of games I have carried to a win against, what most would say, were insurmountable odds... Is well and truly up there.

The problem is when the game is already 0-5 and you somehow pull off a win, those players who essentially suicided with sub 150 damage cause they are just not that good (yet are probably in Tier 1 or Tier 2), get a PSR reward because a win still means they at least stay even although with 150dmg i suspect you still get a RISE.

This rise comes from someone like me doing the work, not them. Thier sub-standard performance is rewarded. It quite simply should not. They should get a PSR drop in the event of a win... And that is the glaring issue with PSR as it stands and why PSR tiers do not work properly at all.

Tier 1 should be reserved for the top 1,000 players or so. Not people that are ranked in the 27,000s on Jarls list (out of 35,000). You should be in Tier 4 if you are that low in the list in all honesty.

I believe many have asked to reverted to T4 or 5.... Then, lock out the T1's and 2's completely..... I believe, that if we could do this, and let the elite play the elite full-time, the T4 and T5's would have a blast having fun while the ePeen community fights over who's top of the leader board.... A perfect solution !!!! T-3's would be the fillers for both but, there would be a hard wall between upper and lower division no matter what the MM thinks.

Now, how would we get PGI to do............anything.............even participate in this forum on a regular basis is impossible.

I'd gladly hand in my T2 for a T5..... Some of the best and most exciting games I've been involved in were at lower levels....

Works for me !

Edited by Asym, 03 June 2018 - 07:26 PM.


#49 justcallme A S H

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • CS 2020 Referee
  • CS 2020 Referee
  • 8,987 posts
  • LocationMelbourne, AU

Posted 03 June 2018 - 07:39 PM

View PostAsym, on 03 June 2018 - 07:25 PM, said:

I believe many have asked to reverted to T4 or 5.... Then, lock out the T1's and 2's completely..... I believe, that if we could do this, and let the elite play the elite full-time, the T4 and T5's would have a blast having fun while the ePeen community fights over who's top of the leader board.... A perfect solution !!!! T-3's would be the fillers for both but, there would be a hard wall between upper and lower division no matter what the MM thinks.


If it works for you - go like my post Posted Image

But T1 and T2 can't exist alone to be honest. The population won't support that.

You do need T3 being counted in there as well, which PGI/Paul do reaslise and have done with the recent MM change which is a good thing but only half the solution. The other half of the solution will definately annoy some, but a potential for some improved matches, is what is behind my proposal.

Yes there will be stomps still, it is inevitable. But if players where put into PSR tiers based on actual performance rather than some wish-washy metric that rewards you even if you play really bad... Well that says it all.Posted Image

#50 The Lighthouse

    Member

  • PipPipPipPipPipPipPipPip
  • Moderate Giver
  • Moderate Giver
  • 1,140 posts

Posted 04 June 2018 - 12:45 AM

View Postjustcallme A S H, on 03 June 2018 - 06:59 PM, said:


Topping the board in an Alt is a direct result of farming from T5-T3 users and not a lot else given Match Score is, generally, easier to farm in lower tiers given people literally stand still. I've seen many an alt-acc crack consistenly dominating games and get well over 500 AMS easily, myself included.

I mean the PSR bar will bump you to Tier 3 in 3 games if you over perform that hard but it sitll won't flick you into T1 in 50 games. It does take a while.

What you proved is that the 'seeding' matches, those first 25-50 or so that are accelerated in moving you up in PSR work. Not that PSR works.

The main issue with PSR is as I have OUTLINED HERE. You have users wih 170 average match score, which is below the average of the MWO community, making it into Tier 1. That should never be allowed.

PSR doesn't 'sort' of work IMO. It's flat out broken. It should be tightened up as again I outlined above to Paul and hopefully takes note. I'd happily wait a couple mins more for potentially reasonably skilled opponents.



I have a firm belief that, if PGI ever decides to become at least semi-transparent and shows us player population/tier distribution as a graph, I would not be surprised if that graph is a perfect bell curve, or even is shifted to lower tiers instead of higher tiers.

And the reason would be horrid player retention rate.


I mean despite all of indication that current PSR system forces people to move up tiers, yet Paul said this in recent matchmaking update;

Quote

.....that would result in Tier 1 and Tier 2 players sitting in queue for a VERY long time.. think like 45 mins +.....


It only means we simply do not have enough players, even for tier 2. We should have much more players at tier 1 and tier 2 due to the nature of the system, but clearly this is not the case.

I am not sure tightening up would solve matchmaking problems in current population unfortunately. It may go worse if PGI does that, because there aren't enough players to bridge top-playing players and rest of average players.

#51 Vellron2005

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The Blood-Eye
  • The Blood-Eye
  • 5,444 posts
  • LocationIn the mechbay, telling the techs to put extra LRM ammo on.

Posted 04 June 2018 - 01:33 AM

I think ELO is only good for 1 vs 1..

And PSR.. is not good. It doesn't take into account sooo many things..

I think proper matchmaking should take into account each individual mech's performance, and loadout, as well as win/loss and KDR in each individual mech.. not across a whole pilot, but in each individual mech..

Then it would be alot more accurate..

#52 Asym

    Member

  • PipPipPipPipPipPipPipPipPip
  • Nova Captain
  • 2,186 posts

Posted 04 June 2018 - 06:25 AM

View Postjustcallme A S H, on 03 June 2018 - 07:39 PM, said:


But T1 and T2 can't exist alone to be honest. The population won't support that.


Too bad for them.... Ah, those poor leader board heroes are now all alone... I wonder, how that happened?

Low population isn't the game's fault alone. Low population was caused by a cascading toxic environment supported by really bad corporate decisions to save money.....to cut operating costs to improve revenue and cash flows,,,, PGI assumed that a small niche market was self sufficient and self efficient and it is........to a point.

You reap what you sow.... And now, "there are no chickens in the roost;" even with schools being out for summer..... Crap, in other games the summer hiatus is causing the "experts" nightmares because there are thousands of kids playing full-time and that isn't always a "good thing" and some tempers are fraying.....

I hope MWO can pull through this 'dry spell' and MW5 works..... I've been playing with stompy mechs for decades and it'd be shame to see the franchise implode..........

#53 Bud Crue

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Rage
  • Rage
  • 9,942 posts
  • LocationOn the farm in central Minnesota

Posted 04 June 2018 - 06:45 AM

View PostAsym, on 04 June 2018 - 06:25 AM, said:


I hope MWO can pull through this 'dry spell' and MW5 works..... I've been playing with stompy mechs for decades and it'd be shame to see the franchise implode..........


This past year has been a “dry spell” what we are experiencing now is something quite opposite of that.
https://leaderboard.isengrim.org/stats

Population-wise things are wonderful atm, at least from PGI’s perspective, at least for solo QP (I’m not sure I buy this, at least for GQ numbers, as quite frankly, I only see the same folks night after night in GQ with no more moments of “Who is unit XYZ? Never heard of those guys” than usual). I even did a single solo faction play drop over the weekend. My entire team was individual randos with maybe half having unit designations. Gotta be a good sign if even FP is getting new folks playing it (of course half didn’t listen, we got smoked, and thus most of those players will likely never return, but still...hope for the future!).

#54 BWS2K

    Member

  • PipPipPipPipPipPip
  • 215 posts
  • LocationUSA

Posted 04 June 2018 - 07:54 AM

View PostThe Lighthouse, on 04 June 2018 - 12:45 AM, said:

I mean despite all of indication that current PSR system forces people to move up tiers, yet Paul said this in recent matchmaking update;

[The thing Paul said]

It only means we simply do not have enough players, even for tier 2. We should have much more players at tier 1 and tier 2 due to the nature of the system, but clearly this is not the case.

I am not sure tightening up would solve matchmaking problems in current population unfortunately. It may go worse if PGI does that, because there aren't enough players to bridge top-playing players and rest of average players.

Okay, so now I feel like I can mention... I noticed that as well. It's why I started this train rolling, actually. These are the things that boggle me: PSR working as intended, everybody floats to the top, T1's are playing with T5's (until just recently), T1 and T2 together don't have a large enough population between them for reasonable wait times, some T1's show considerably less 'skill' than others, some T5's have been stuck in T5 for a very long time. All of these statements together cannot all be true unless there's another major factor we're simply not told. I don't expect PGI to spill all the data beans but there's a point at which building trust in your playerbase sometimes means showing your hand a bit. If you don't, you encourage an environment of dis-trust which breeds the kinds of players nobody should want.

I say that to say: I honestly don't know why opening up cross-tiers in QuickPlay right now would make a discernible difference. If there aren't a lot of players in T1 and T2 already then it's not going to 'pollute the waters' to a significant degree if they all have an equal chance of saddling up with T5's or each other and it might mean rookies learn from veterans and the community grows. We still have 'stomps' and that won't change and the PSR can keep on doing whatever it does. Everyone would know, up front, what they'd be facing and even though it wouldn't directly affect Faction Play or Group Play, it might indirectly offer a wider population and, if the veterans are mentoring the rookies and nobody's being a jerk in QuickPlay, it might mean more skilled players checking those modes out. It wouldn't just be spikes of inflow of 'terribads' during FP Events. Just some thoughts. This has been a very informative thread and I appreciate all the feedback for sure.

#55 Asym

    Member

  • PipPipPipPipPipPipPipPipPip
  • Nova Captain
  • 2,186 posts

Posted 04 June 2018 - 09:34 AM

When I started MWO in Jan/Feb of 2017, I rarely saw the same names in QP... Like, maybe a name I'd write down as one to remember for later on, and I'd not see that name or names for days. Faction Play had a lot of teams that had multiple drop teams and no wait times and rarely, would you see the same teams in days.... It was exciting !

I never saw the T1's or T2's starting out: never. All of the names I kept were all of the same tier's I belonged to and we talked about that during matches..... Sometimes, a T1 player would drop as an alternative account T4-5 and start mouthing off about skill, build, potatoes, and noob stupidity. Those were the first indications that there were "better players" and many of us were just joining teams...... Then, we got to see comp players and started dropping in faction play: where there is no MM.

Back then, before the new skill tree, we had team Trails, and team practices, and team skirmishes and you had to earn you place on the drop rosters.... Then, the new skill tree and Civil War and the whole process stopped ! And, I mean really and significantly went Mr. Hyde.... What were difficult matches became so toxic and obtuse, the entire team withdrew from MWO.
And, not just one team either. When FP imploded, QP went to crap even faster as the low population, new tech and meta/op exploded into a mess of unknown territory...... It has never recovered. We knew S7 was coming in April of 2017.........and many, not wanting anything to do with S7, just started abandoning the game full-time. Go to Jarl's and look me up and see the chronology.....

Here we are. I only play events and the team I am on average 6 of 52 players a week playing S7.... An of the 6, the top 4 players of the team are the most active but, for a few matches.... a very quick decline of a very good game.

#56 BWS2K

    Member

  • PipPipPipPipPipPip
  • 215 posts
  • LocationUSA

Posted 04 June 2018 - 06:31 PM

lol - I had a drop the other night with bear_claw and I was like "Bearclaw?! What are you doing down here?" We chuckled.

I dunno. It does seem like an unfixable mess but I hate to see something go so misunderstood.

#57 Xetelian

    Member

  • PipPipPipPipPipPipPipPipPip
  • Elite Founder
  • Elite Founder
  • 4,393 posts

Posted 04 June 2018 - 06:38 PM

Just want to chime in and say what other people have already said in my own words.

In a game where there are 24 people in a match, individual W/L is meaningless for anyone at the above average to slightly below average mark. The main average of players are going to have around a ~1.0 with the usual of them being <1.0. Why bother have match making at all if 60% (just picking a number, didn't do any math) are going to have an average W/L while 20% are going to have a higher than 1.0 and 20% are going to have much much lower than 1.0.

So instead of 5 tiers we'd have 3 effective tiers and the system as a whole doesn't change in any meaningful way compared to how it is now.



In chess, you can do a Win/Loss because you have only yourself to blame for your losses and only yourself to congratulate for your wins when equally matched. They use Elo, which takes into account the W/L of both players. How do you take into account the W/L of individuals when due to a freaky amount of luck they always end up with the best players on their team? Or the opposite.




Match making needs to take a LOT of things into account when you have a multiple player PVP game. I have had 10 damage matches and I have had 1400+ damage matches. It can come down to whether or not I'm instantly headshot by a top tier pilot with DHG or if I have that guy on my team.


One thing I liked about the previous Elo match making is that different weight classes had different Elos.

#58 Pardo Kerensky

    Member

  • PipPipPip
  • FP Veteran - Beta 1
  • 66 posts

Posted 04 June 2018 - 06:45 PM

because W/L ratio based ELO rating would actually expose trash players for the actual trash they are, and PGI cant have that since trash players with disposable income are the ones getting milked constantly and bled dry with Mechpacks™ releases

#59 BWS2K

    Member

  • PipPipPipPipPipPip
  • 215 posts
  • LocationUSA

Posted 04 June 2018 - 06:49 PM

So I'm serious then - what would happen if everyone was just allowed in to play with everyone? Tear down the tiers? All players would have an equal chance of getting on a team with T1's or T5's, so everyone has an equal chance of getting stomped or not... right?

#60 A Headless Chicken

    Member

  • PipPipPipPipPipPip
  • The Hungry
  • The Hungry
  • 273 posts
  • LocationImmersed in Stupid.

Posted 04 June 2018 - 06:54 PM

If within Tier 1~3, we already struggle to find a balanced match, expanding it would just make the game more...

"fun".

I mean, I really miss the days where I could kill someone from the back because he doesn't even bothering to turn around.





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users