129 replies to this topic

[Mystere]

Solaris mode, if and when it arrives, is where you should all be measuring your e-peens, not via PSR.

[Felicitatem Parco]

PGI said there are already players who have reached the PSR point-cap. They cannot get more points.

PSR is based on accumulated points, and playing at 50/50/W/L causes a net gain over time.

See Clarifcation#2
http://mwomercs.com/...-tiers-and-psr/

Edited by Prosperity Park, 01 October 2015 - 11:28 AM.

[Appogee]

Stefka Kerensky, on 01 October 2015 - 11:10 AM, said:

I still think, as always thought, that the best way to determine is a player is comp/good/average/bad/steeringwheel, is to read his/her stats.
w/l ratio
k/d ratio
dmg/match.

But these are not comparable across players. For example:

• A player who always plays as part of a large group of Tier 1 players will generally have a better w/l ratio than one who plays mostly in the PUG queue, even if they are of equivalent skill.
  
• A player who mostly dual Gauss snipes will have a better k/d ratio than one who doesn't.
  
• Damage per match is not calculated across matches. But even if it were, a player who used a LRM80 Assault would have a better stat - but probably be less of an actual team contributor - than an equivalent player who brawls.

I don't think PSR is the ultimate measure of a Mechwarrior's skill either - it's too influenced by the vagaries of the matchmaker, and the dynamics of Group vs Solo queue.

But hey, it's better than nothing.

[Kira Onime]

Water Bear, on 01 October 2015 - 11:22 AM, said:

Why not? If everyone in the game is picked so that their Elo is within +- 5%, say, of your Elo, then isn't it fair to judge whether your Elo should go up or down based on the outcome of that game?

Starcraft 2 had rankings for it's multiplayer modes (2v2, 3v3, 4v4) and in all of those players tended to settle into consistent ratings just like they do in 1v1. I am under the impression that LOL or similar games also have rating systems that seem pretty stable.

ELO at its core is a flawed system for team based games, even more when you take group queue into consideration.

The current ranking system, like it or not, at least tries to rank people based on their own performance instead of a flat "did you win or lose" scenario.

[sycocys]

Kira Onime, on 01 October 2015 - 10:57 AM, said:

ELO doesn't work that well when you have 11 random elements on your team.

And it doesn't even take into account that on average at least half of them are drunk or stoned - seriously why is that not a part of your PSR rating?

[Water Bear]

Kira Onime, on 01 October 2015 - 11:29 AM, said:

ELO at its core is a flawed system for team based games, even more when you take group queue into consideration.

The current ranking system, like it or not, at least tries to rank people based on their own performance instead of a flat "did you win or lose" scenario.

Please explain specifically why Elo is flawed here.

Also everything we are discussing in this thread ranks people based on how well they did and not just whether they won or lost so I'm not sure what you're talking about.

[STEF_]

Appogee, on 01 October 2015 - 11:26 AM, said:

But these are not comparable across players. For example:

• A player who always plays as part of a large group of Tier 1 players will generally have a better w/l ratio than one who plays mostly in the PUG queue, even if they are of equivalent skill.
  
• A player who mostly dual Gauss snipes will have a better k/d ratio than one who doesn't.
  
• Damage per match is not calculated across matches. But even if it were, a player who used a LRM80 Assault would have a better stat - but probably be less of an actual team contributor - than an equivalent player who brawls.

I don't think PSR is the ultimate measure of a Mechwarrior's skill either - it's too influenced by the vagaries of the matchmaker, and the dynamics of Group vs Solo queue.


But hey, it's better than nothing.

But those stats are meant to be read per chassis. I'm not talking about general stats.

I agree that with PSR, matches are more challening, but also, I think it's true that, playing a lot, every average pilot can be tier 1.

So, I like stats more.

[Moomtazz]

I simply don't like the name. It implies that the Tier is more than just a passing reflection of a player's skill. It is really just a Win/Loss indicator

[Kira Onime]

Water Bear, on 01 October 2015 - 11:32 AM, said:

Please explain specifically why Elo is flawed here.

Again, ELO simply looks at the result of the game.
Did you win? Go up.
Did you lose? Go down.

This works in a 1v1 scenario as the only variable in play for yourself is... yourself.

I'm not sure I can put it more simple that that.

[Mcgral18]

Aresye Kerensky, on 01 October 2015 - 10:38 AM, said:

No, you're not a talentless user. In fact I respected you and your community involvement until this whole Tier thing turned you into a whiny, self-loathing player with a victim complex. You literally threw in the towel and began this whole tirade because of a single person on one thread. You even admitted so in your "I'm done," post.



So we're just supposed to take your word for it? Who were the players involved? What was the topic about? What was the back and forth about? For someone who's been vehemently against the visible Tiers because of these exact types of situations, I find it pretty odd that you would withhold any actual proof that would support your claim. I mean, if I was against the visible Tier thing, I would have snagged some screenshots of people's replies, saved webpages, and did everything in my power to support my opinion if ever questioned in the way that I am questioning you right now. You're either incredibly incompetent at forum warfare, or you're lying.

I'm gonna go ahead and say he's talking about me and my Clam Laser thread.

Someone was calling me a no skill Scrub who needs Crutch nerfs because Clam mechs are fine and overnerfed already, and that I also couldn't torso twist.
They couldn't understand the fact of damage/tick, how all but 2 Lasers are in the Clan favour for damage over the same duration.

I felt it was appropriate. Turns out I wasn't wrong. Tier 3, it turns out.

Edited by Mcgral18, 01 October 2015 - 11:47 AM.

[Water Bear]

Kira Onime, on 01 October 2015 - 11:38 AM, said:

Again, ELO simply looks at the result of the game.
Did you win? Go up.
Did you lose? Go down.

This works in a 1v1 scenario as the only variable in play for yourself is... yourself.

I'm not sure I can put it more simple that that.

Water Bear, on 01 October 2015 - 11:22 AM, said:

Why not? If everyone in the game is picked so that their Elo is within +- 5%, say, of your Elo, then isn't it fair to judge whether your Elo should go up or down based on the outcome of that game?

Starcraft 2 had rankings for it's multiplayer modes (2v2, 3v3, 4v4) and in all of those players tended to settle into consistent ratings just like they do in 1v1. I am under the impression that LOL or similar games also have rating systems that seem pretty stable.

I think we agree to disagree now.

[Apnu]

Prosperity Park, on 01 October 2015 - 09:14 AM, said:

I think that allowing us to view our own PSR tiers was a mistake. It hasn't generated any usefulness, but it HAS generated pointless strife. People bickering about how they are in the wrong Tier, taking "Pride" in their Tier, blaming their Tier for every loss (not for the wins, mind you, just the losses), and all that Mumbo-Jumbo.

Really, making our tiers known hasn't generated any positive results as far as I have seen. I don't give a crap about my Tier, but it seems that others view these tiers as some statistical God,. If the tiers weren't known, then people would not be blaming their Tier for their combat results, but just be blaming the teams at play. This would make more sense because Tiers don't bring weapons into combat; teams do.

You can't blame everything on a Tier, and knowing your Tier has ZERO impact on your ability to change your performance capabilities. Suddenly becoming conscious of your Tier has not made anyone better or worse at the game.

Maybe we should take it away so we can focus on the things that matter, like blowing up stompy Mechs.

EDIT
---------

So, Many people have stated that the Tier1 is a "goal" for them.

If a new player joins MWO and wins their first 50 consecutive matches, then they will not be in Tier 1. Not even Tier 2.

Tier 1 means you have been playing for long time. Nothing else, really. If the law of averages nets you a 50/50 win loss ratio, and you play for a long time, then you are Tier 1. Is that your goal? To get a 50/50 win/loss ratio?

How is that a lofty goal that requires a Tier rating? That's not very impressive.

Getting a 50/50 wn/loss ratio does not require knowledge of our tiers.

Honestly, several different PGI people have said what you said. That PSR (a fancier term for Elo, let's be real), isn't helpful for anybody outside of the MM. And that players would instantly wad their panties over it. And you know what? They were right.

Players are panting over PSR because the game is so scant on content. We got lots of mechs a few maps, three game modes and a less than alpha-quality CW. Boredom is deep and thick here and players are desperate for anything, ANYTHING, to keep them going.

So now they're hung on PSR, because there's nothing else to do here.

[Moomtazz]

I simply don't like the name. It implies that the Tier is more than just a passing reflection of a player's skill. It is really just a Win/Loss indicator

[pwnface]

Water Bear, on 01 October 2015 - 11:46 AM, said:

I think we agree to disagree now.

ELO doesn't work in MWO because of the existence of group queue. If this game were 100% solo queue only, ELO could be an effective way to do matchmaking. The way ELO was supposed to work was to balance both teams so that they have a relatively similar average ELO rating, not every player on each team was similar in rating though. Low ELO players and high ELO players are put on the same team to balance each other out.


On PSR:
I just tell everyone I'm in Tier 5 so they get sad when I spank them in solo queue.

Edited by pwnface, 01 October 2015 - 11:54 AM.

[Kira Onime]

Water Bear, on 01 October 2015 - 11:46 AM, said:

I think we agree to disagree now.

I just think you don't understand how ELO works.

[Felicitatem Parco]

Kira Onime, on 01 October 2015 - 11:58 AM, said:

I just think you don't understand how ELO works.

"Elo" is a gentlemen's last name, and he popularized a ranking system for chess players. ELO is a rock band from the 70's that produced pop hits such as "Don't bring me down" and "Evil Woman"

Elo systems rank you based on how you play against others and weighs your wins/losses against your opponent's calculated skill level.

The PSR system does not use cross-comparisons to your opponents, and rather gives you more points or takes points away from you based on your match outcome (regardless of your opponents' calculated skill levels).

If you play for a long time and keep a 50/50 win/loss ratio, then your PSR rating will climb until you reach the Tier 1 point cap. Once there, you cannot get more points. You cannot go "higher" at that time.

[Kira Onime]

Prosperity Park, on 01 October 2015 - 12:11 PM, said:

"Elo" is a gentlemen's last name, and he popularized a ranking system for chess players. ELO is a rock band from the 70's that produced pop hits such as "Don't bring me down" and "Evil Woman"

Elo systems rank you based on how you play against others and weighs your wins/losses against your opponent's calculated skill level.

The PSR system does not use cross-comparisons to your opponents, and rather gives you more points or takes points away from you based on your match outcome (regardless of your opponents' calculated skill levels).

If you play for a long time and keep a 50/50 win/loss ratio, then your PSR rating will climb until you reach the Tier 1 point cap. Once there, you cannot get more points. You cannot go "higher" at that time.

And one of the biggest thing to keep in mind is that Elo was created for a 1v1 game. Not a 12v12 one.
The system works fine in the context of no outside variable out of your control.

[1453 R]

Mcgral18, on 01 October 2015 - 11:44 AM, said:

I'm gonna go ahead and say he's talking about me and my Clam Laser thread.

Someone was calling me a no skill Scrub who needs Crutch nerfs because Clam mechs are fine and overnerfed already, and that I also couldn't torso twist.
They couldn't understand the fact of damage/tick, how all but 2 Lasers are in the Clan favour for damage over the same duration.

I felt it was appropriate. Turns out I wasn't wrong. Tier 3, it turns out.

Did not actually read much of that thread. Too depressing to see a good debate I couldn't participate in.

Apparently a good thing I didn't get too deep into it, though. I shall remember I'm not supposed to bother you in the future. Handy bit of information to have.

[Mcgral18]

1453 R, on 01 October 2015 - 12:16 PM, said:

Did not actually read much of that thread. Too depressing to see a good debate I couldn't participate in.

Apparently a good thing I didn't get too deep into it, though. I shall remember I'm not supposed to bother you in the future. Handy bit of information to have.

As long as you're not an idiot, you'll be fine.


Those two, however, were unreasonable.

[Wintersdark]

Water Bear, on 01 October 2015 - 11:09 AM, said:

Edit: To make the argument more clear:

The probability that you lose and do very well is (probably) lower than the probability that you win and don't do too well. So if you are matched against even opponents, the distribution of your performance (poor, good, excellent) is not the same conditioned upon whether you won or lost.

Therefore things may not 'balance out.'

Its nearly impossible to lose rating on a win.

You don't need to "do well" to maintain or improve rating on a loss, its easy to get enough damage to on a loss if your equipped for it.

As I said, I gain rating on almost all losses as it stands, and I'm t3 and not amazingly good at the game.