62 replies to this topic

[BWS2K]

Greetings, all!

I've been reading up on Elo and I think I understand it. Maybe. A little. But I get how it's mostly a 1v1 system and might not work so well with MWO. It feels like we all know the Tiers/PSR is kind of a mess too and I occasionally come across the terms 'buckets' without fully understanding them either... but is there are reason we don't use raw Win/Loss ratio (maybe with a modifier like the Cadet Bonus to keep the Beginner's Luck crowd out of the Veteran's playground)?

Thanks!

Bonus questions for clarification: PSR doesn't actually apply to Group Drops or Faction Play, does it? It's literally just for QuickPlay?

[Champion of Khorne Lord of Blood]

Because people can be carried or brought down by their team quite easily in a non 1v1 environment causing Elo based matchmaking to be pointless. In a situation where you are only 1/12th of the deciding factor of a win why would if you win or not be what decides your ranking?

[BWS2K]

Champion of Khorne Lord of Blood, on 01 June 2018 - 03:49 PM, said:

Because people can be carried or brought down by their team quite easily in a non 1v1 environment causing Elo based matchmaking to be pointless. In a situation where you are only 1/12th of the deciding factor of a win why would if you win or not be what decides your ranking?

I'm not totally sure I understand what you mean but if a person is always playing to win, whatever the cost (and I realize that's not always the case), then wouldn't W/L represent that better than the PSR? Maybe it doesn't help for MatchMaking though...

[Champion of Khorne Lord of Blood]

BWS2K, on 01 June 2018 - 03:53 PM, said:

I'm not totally sure I understand what you mean but if a person is always playing to win, whatever the cost (and I realize that's not always the case), then wouldn't W/L represent that better than the PSR? Maybe it doesn't help for MatchMaking though...

Generally a person can be playing to win all they want but without the help of the other 11 people on their team all their hard work can be for nothing. You can kill half the enemy team all by yourself but if the other remaining people can't win a 11v6 (which has happened to me many times) then its a loss for you at not fault of your own other than not being a super human who can deathstar every match.

Meanwhile if you get about 4 people on your team who can work together and do very well their coordination can bring a victory to a team that would otherwise not be contributing much, thus they'd move up in rank without any effort put in.

Elo only works for 1v1 situations.

[Ghogiel]

BWS2K, on 01 June 2018 - 03:53 PM, said:

I'm not totally sure I understand what you mean but if a person is always playing to win, whatever the cost (and I realize that's not always the case), then wouldn't W/L represent that better than the PSR? Maybe it doesn't help for MatchMaking though...

W/L does represent ability to affect a matches outcome better than PSR, or any stat for that matter, but PSR isn't designed to do that
It's to sort total potatoes and nubs from being completely farmed in most of their games while at the same time allowing good players, who might have been playing this game for like 6 years, to actually get matches in less than 30min wait times.
It's not rating player skill.

As for why they dont use it, well i was here when we had Elo and it sucked waiting a minimum of like 15mins between matchs to upwards of 3hrs. 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.

Edited by Ghogiel, 01 June 2018 - 04:04 PM.

[BWS2K]

Hmm. These things make sense, lol. Thanks.

[Angel of Annihilation]

BWS2K, on 01 June 2018 - 03:53 PM, said:

I'm not totally sure I understand what you mean but if a person is always playing to win, whatever the cost (and I realize that's not always the case), then wouldn't W/L represent that better than the PSR? Maybe it doesn't help for MatchMaking though...

This is only true if you believe that a person, through their own actions, completely independent of their team, can actually influence a win on a consistent basis. For myself, I have personally been in way to many matches where I have killed 3-4 enemy mechs, have had most of them be solo kills and outputted 800+ damage only to find myself be the last team member alive against 6-8 enemy mechs and got shredded by those 6-8 enemy mechs to believe that I, myself, alone can consistantly insure my team wins the majority of their matches.

Seriously though, I do absolutely believe that there are some players out there that are good enough to actually influence the majority of matches they play in but just like Michael Jordan or insert any superstar, elite athlete out there, they are 1 in 1000 or maybe even 1 in 100,000. The majority of us just aren't that good despite what we might think.

So for those of us who just aren't that good, the 98-99% of us who are playing this game, we have to hope that our teammates know their head from a hole in the ground, in order to secure a win.

[El Bandito]

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

[The Lighthouse]

In team games, ELO and performance are separated matters.

Of course, there are some people who can influence outcome of the game reliability, but such talented people are really tiny, and the whole matchmaking system should be worked for majority of people, not those in tiny percentage.

Using ELO rating to team games only cause grief and toxic behaviour. Look at current status of Dota 2 and LoL community.

[BWS2K]

So it wasn't always the PSR, right? Why... where did the PSR come from then? What problem is it supposed to be solving?

[Champion of Khorne Lord of Blood]

BWS2K, on 01 June 2018 - 07:54 PM, said:

So it wasn't always the PSR, right? Why... where did the PSR come from then? What problem is it supposed to be solving?

Previously we had an Elo system that didn't work out well due to reasons mentioned so we moved over to PSR which takes into account the pilot's performance while still taking into account whether they are winning or losing so that people who carry hard still move up on losses and people who did absolutely nothing on wins don't move up.

[Mister Glitchdragon]

https://mwomercs.com...-losing-streak/

[Tarl Cabot]

https://mwomercs.com...ost__p__1626065

[MrMadguy]

BWS2K, on 01 June 2018 - 03:42 PM, said:

Greetings, all!

I've been reading up on Elo and I think I understand it. Maybe. A little. But I get how it's mostly a 1v1 system and might not work so well with MWO. It feels like we all know the Tiers/PSR is kind of a mess too and I occasionally come across the terms 'buckets' without fully understanding them either... but is there are reason we don't use raw Win/Loss ratio (maybe with a modifier like the Cadet Bonus to keep the Beginner's Luck crowd out of the Veteran's playground)?

Thanks!

Bonus questions for clarification: PSR doesn't actually apply to Group Drops or Faction Play, does it? It's literally just for QuickPlay?

As always, because W/L = 1 in Tier 5 and W/L = 1 in Tier 1 - are two completely different things.

[Chados]

Yeah, ELO is how they used to do it.

It’s the main reason my overall career KDR is 0.75. When I started in July 2015 the learning curve was even more brutal than it is now. It was past vertical and you were climbing an inverted slope, man, I remember it well. No skill tree, progression was chassis wide and linear, and an elite mech was significantly better than a basiced build, which was significantly better than a new mech. You were in QP with the top players in the game from your first drop, the cadet bonus didn’t cover cost of more than one new mech...and you needed three to advance a chassis to elite...and games were battlefield-wide seal-pounders for anyone with less than a couple hundred drops.

My global KDR was something like 0.1 at one point, about a month in. I got tested for recruitment by one of the then-high end FP guilds (we called it “community warfare” then) and sucked so very bad that I didn’t qualify...but I got a few pointers. That helped. But I was about to bail on the game entirely.

Then, I stripped and sold everything in inventory, all of it, including one champion (JM6A before the dual gauss build) and one hero (CDA-X5) I’d bought with MC to see if that would help, including all my armament I’d bought. And invested in a Catapult C1. It cleaned me out entirely. I discovered Noob LRMing. The Cat C1 slowed the game down, kept me in the game, and I began to learn. Joined another guild and began to learn more. Bought the whole Cat mastery pack and had by then enough Cbills to get the C4, and the A1(C) was the first ever mech I mastered out fully. In the process I discovered Point Blank LRMs and that wasn’t enough, so I turned to the Jester and K2 to learn how to direct fire. The change to the PSR system helped even more, as I was so far down at the bottom of T5 that there was no orange line on the scale.

I’m still a spud, but now thanks to the Catapult I’m a McDonald’s French-fry as opposed to a raw potato. I’m still in the game and have elevated my game to T2, where I should not be at (I’m probably more low T3/high T4, skill wise) but I usually don’t embarrass myself in matches anymore. And PGI’s made too much money off me in the process. PSR was a win for players like me, to be frank. For the current noobs, I’d hate to see ELO make a comeback. If you want a taste of that gameplay try Solaris or FP. There’s a reason I avoid both. I’m outclassed, and I know my limitations-I’m not a twitch player and I never will be. My reflexes aren’t that fast, my eyesight isn’t that good, and I’m a casual player. Throwing that kind of player up against teams stocked with twitchkids living on adrenaline and Red Bull while playing twelve hours a day every day is always going to end in a curb stomp for the casual. Always.

[jss78]

Champion of Khorne Lord of Blood, on 01 June 2018 - 03:49 PM, said:

Because people can be carried or brought down by their team quite easily in a non 1v1 environment causing Elo based matchmaking to be pointless. In a situation where you are only 1/12th of the deciding factor of a win why would if you win or not be what decides your ranking?

Because all other elements except you are randomized. That's the usual argument in favour of W/L, and I'd agree actually.

[Asym]

The only true test of W/L is if you are only playing 1x1... Any numbers > than 1 creates so much variation, that if you are looking for any real precision, it evaporates as the numbers of players being considered increases... Compounding errors... As said above, you can have a record single game and still lose.... W/L is a binary calculation: you either win or lose.....

PSR or ELO are far better but, can only limit variables to a point. Low populations increases the potential of negative variation and as populations keeps getting smaller, the negative influence gets exponential... Consider we have a player pool of 100 and El Bandito is the top player.... You have a 1 in 100 chance you'd see him when MM selects a drop matrix.... Now the player pool is 20 at the time you want to drop and El is there.... Your chances of facing him are significantly greater and, after he and his cohorts stomp you, you W/L and PSR are negatively affected. As game populations dwindle, the new players leave first and the last are the veterans and die hards... So, when pop's get smaller and smaller, the skill pool get harder and harder...... And, we wonder why the numbers just don't make any real sense....or, have extreme shifts...

[BWS2K]

Since the game is a shooter and role warfare is... not really done well... would it make sense to just let PSR equal an average of damage done per match or something? T1 would be the top damage-dealers, etc. Limit it to a single factor that aligns with what the game is supposed to be about (not my words - MWO calls itself a tactical shooter)? Maybe that leads to the same problem though. Hmm...

[S O L A I S]

The current system already favors winning over skill. PGI has defended current system and despite naming it pilot skill rating Paula told Phil it is actually an experience bar after all! There has been a lot of complaints and the community has tried to get PGI to adjust or change the system so pilots of similar skill are matched together but it has fallen onto deaf ears. They have pretty much said that we will be stuck with the way it is.

[Nightbird]

S O L A I S, on 02 June 2018 - 01:19 PM, said:

The current system already favors winning over skill.

You mean number of matches played over skill right? Or number of wins if you put it another way, without regard to win rate (percentage).