67 replies to this topic

[Lykaon]

Xavori, on 25 January 2016 - 07:51 AM, said:

From what I've learned playing, and what I've seen watching vids by the top players, teamwork, not skill, is what drives winning and losing. (Seriously, teamwork is severely OP and definitely needs nerfed )

So why then does "pilot skill rating" care whether you won or lost a match? I mean, if you yolo your light mech into a lance of assaults and get blown to smithereens, but your team wins, you still suck and should see your rating go down. On the other hand, if you're running around getting shots in when you can on a team that is getting overrun, doesn't that show more skill? (yes, I know if you can push your damage high enough, it can still go up even in a loss)

And while we're on the subject of more skill...

Why is doing lots of damage in an assault or heavy mech, which is pretty easy, more valuable to your pilot rating than tagging things or capping or harassing lurmy boats in a light mech even tho those things are much harder. I know I didn't suddenly become a better pilot when I switched from running almost all lights to heavies and assaults, by my pilot skill rating is finally (slowly) climbing out of the basement. In fact, I'd argue I'm getting to be a lazier pilot because of it.

ANSWER: PGI has tailored nearly all the support mechanics in the game to casual solos. The "average" MWO player apparently does not belong to a unit or play in groups or play a game mode other than skirmish for that matter.

Almost the entire game is being designed to spoon feed "casuals".

"Casuals" often refered to as pugs or puggies in MWo generally play in a rather narrow margin of variables.

They pilot generally the same chassis.ussually heavies.Having more mechs means more mech bays (mechbays cost money and why would they want to spend any on MWo)

They rarely paint their mechs (paint cost moneyand why would they want to spend any on MWo)

They will almost always select skirmish mode over anything else,Apparently secondary objectives are to hard or confusing or distract them from playing skirmish.

They are the ones who flood the forums with complaints about "premades" or "pugstomps" or any number of other topics that are actually about their lack of ability to cooperate and perform team work. Because they choose to play this game in a less efficent manner the players who do play effectivley have to be punished.

Essentially "casual solos" need to have the PSR system in place as a means of funneling them into generally similar pools of players. Since the PSR system doesn't really work well in the group queue it's clearly for them and not the units.

Why does it not work in groups you may ask?

PSR should keep players in a match with other players within 2 steps of their tier. Yet id a player group contains a tier 1 player and a tier 4 player the group will be placed in the same match obviously this supercedes the 2 step rule so PSR is not working as intended.

[Moldur]

You should have some sway on whether or not you win or lose in a game. It just gets obscured in the 12 v 12.

If it were 5v5 or even 8v8, this would not be a question as often.

[VirtualRiot]

Once everyone is super
No one will be

[TercieI]

Moldur, on 25 January 2016 - 06:06 PM, said:

You should have some sway on whether or not you win or lose in a game. It just gets obscured in the 12 v 12.

If it were 5v5 or even 8v8, this would not be a question as often.

So long as you believe that, it'll be true.

[Narcissistic Martyr]

Because fundamentally over a large enough number of games the only constant that decides winning and losing is going to be you.

If a player is winning more games than they're losing they're contributing to that victory perhaps through damage and kills or perhaps through something like using VOIP to relay info or give commands. The latter can't be tracked through other stats, but by making winning and losing part of it you can account for those things.

[Malachy Karrde]

I've deliberately kept my psr at tier 3. I pilot a mech that I'm really good with when I need wins... and mech like the craphammer or something I'm mastering when I need losses. I have no interest in playing with the elitists that run around in tier 1 and 2 all the time, but I'm not keen on playing with the steamers all the time either. Seem to get a good mix at tier 3.

[MrMadguy]

Xavori, on 25 January 2016 - 07:51 AM, said:

From what I've learned playing, and what I've seen watching vids by the top players, teamwork, not skill, is what drives winning and losing. (Seriously, teamwork is severely OP and definitely needs nerfed )

So why then does "pilot skill rating" care whether you won or lost a match? I mean, if you yolo your light mech into a lance of assaults and get blown to smithereens, but your team wins, you still suck and should see your rating go down. On the other hand, if you're running around getting shots in when you can on a team that is getting overrun, doesn't that show more skill? (yes, I know if you can push your damage high enough, it can still go up even in a loss)

And while we're on the subject of more skill...

Why is doing lots of damage in an assault or heavy mech, which is pretty easy, more valuable to your pilot rating than tagging things or capping or harassing lurmy boats in a light mech even tho those things are much harder. I know I didn't suddenly become a better pilot when I switched from running almost all lights to heavies and assaults, by my pilot skill rating is finally (slowly) climbing out of the basement. In fact, I'd argue I'm getting to be a lazier pilot because of it.

This is the curse of all games, that have some stupid objectives - devs want to encourage you to complete this objectives, otherwise they will be completely pointless and game could be turned into deathmatch. This is the reason, why we still have terrible matchmaking - MM is way too biased towards keeping W/L at some level. Why this is so bad? Team is assembled via using your personal rating. But it's team, who wins or loses, so your personal rating is affected by team's performance. And you are just 1/12 of team and just 1 of 24 players on a battlefield. So it's obvious, that the bigger team is - the bigger rating measurement error is.

For 12vs12 it's complete mess. One team performs well due to some random factors - your PSR grows. Other team performs poorly due to some random factors - your PSR drops. And this random factors affect your personal rating, that intended to be treated as measure of your personal performance, which is almost constant value in most cases. Result - incorrectly measured personal skill rating, which causes almost purely random matchmaking, i.e. stomps and other terrible sorts of team imbalance.

And don't listen to people, who say, that if factors, that affect team's performance, are random, then normal distribution of skill, law of large numbers and other statistic crap are applicable. For this to be true, playerbase should be static in a first place. This method is applicable for chess tournaments for example, where number of players is fixed and everybody is able to play against each other. When some players play in the morning, some in the evening, some play for the whole week, some on weekends only, some are in one timezone, some in another - this method is inapplicable.

Also system is only applicable, if all players has very close skill levels. As soon, as you open release valves and mix Tier 1 players with Tier 5 - you completely mess the matchmaking. And it causes dead loop as the result - you mess matchmaking -> you mess PSR values as the result -> you mess matchmaking even more. Most players in MWO will never have right skill ratings - they will always sit in different sorts of PSR "hells" and "paradises".

TL;DR W/L based MM systems are proven to be applicable only for tournaments, with fixed amount of players, teams with fixed membership or small random teams (4vs4 in best case), that has very close skill levels. Chess tournaments, basketball tournaments, football tournaments. In all other cases this systems are inapplicable and provide inconsistent results.

Edited by MrMadguy, 25 January 2016 - 10:20 PM.

[Kilo 40]

because when a lance of assaults, that you are not in, decides to "scout" by themselves and they all die quick horrible deaths with under 100 damage between all 4 of them, causing your team to collapse, that obviously means your "pilot skill" is bad and needs to come down.

seriously, the only people who think w/l should effect PSR are those who play in groups and surround themselves with better players, thereby artificially inflating their PSR, and their ego.

[Tarogato]

Because the entire point of playing should be to win. That should be the driving force behind player behaviour. Not farming muh damages, farming muh KMDD's, farming muh assists, etc etc. Farming *wins* should be rewarded. On top of that, PSR uses your performance in match score as a modifier. So that if you actually do well consistently, you are rewarded for it, but not as well as somebody who does consistently *and* wins consistently.

[MrMadguy]

Died in a matter of one minute (due to enemy performing Atlas rush), win (Atlas rush - suicide rush), 0 kills, ~100 dmg, 120 match score = PSR increase. Great system, lol.

Edited by MrMadguy, 25 January 2016 - 11:00 PM.

[MischiefSC]

Because a lot goes in to winning beyond damage/kill/assist.

It should be the end goal of every match and the driving force behind a players decisions in a match. Making good decisions that drive wins increases your PSR more quickly on average. Yes, there are times when events beyond your control influence a win/loss but if you drive wins on average then you're consistently doing the things that win, on average, more often.

Anyone who doesn't want PSR to include win/loss is probably someone who just farms kills, ignores their team and objectives and overall success in favor of trying to farm a little more points for themselves at the cost of the team then gets pissed at the team when they lose.

PSR is all about win/loss, in the end. A skilled player drives wins. Everything else is just frosting.

[MrMadguy]

MischiefSC, on 25 January 2016 - 11:00 PM, said:

Because a lot goes in to winning beyond damage/kill/assist.

It should be the end goal of every match and the driving force behind a players decisions in a match. Making good decisions that drive wins increases your PSR more quickly on average. Yes, there are times when events beyond your control influence a win/loss but if you drive wins on average then you're consistently doing the things that win, on average, more often.

Anyone who doesn't want PSR to include win/loss is probably someone who just farms kills, ignores their team and objectives and overall success in favor of trying to farm a little more points for themselves at the cost of the team then gets pissed at the team when they lose.

PSR is all about win/loss, in the end. A skilled player drives wins. Everything else is just frosting.

My team, full of laser vomit boats (I spectated this match), won this match - not I. Why should MY rating increase?

Edited by MrMadguy, 25 January 2016 - 11:10 PM.

[MischiefSC]

MrMadguy, on 25 January 2016 - 11:10 PM, said:

My team, full of laser vomit boats (I spectated this match), won this match - not I. Why should MY rating increase?

Maybe you soaked damage, maybe you were a distraction. You're wanting to generate an overall performance metric based off of one-off situations.

Over 100 matches your behavior has a big impact on the average of your wins and losses. Yes, there will be matches you win or lose because of your team but the exact same goes for the other side; on the average those situations come out in the wash. You're about as likely to win because of your team as lose because of your team. Win/Loss as an average over 100, 1,000 matches though is driven by your performance.

PSR also takes facets of your performance into account so that if you're just a bad tactical player with terrible positioning but you know how to run a brawler and tend to go down swinging you're still going to, on average, raise your PSR faster than someone with bad tactical positioning and who is crap at brawling too.

PSR is not your ranking based on the last match you played. It's your ranking relative to all the other players, on average based on your performance. That's hundreds, even thousands of matches. Win/Loss is the biggest single factor in determining how useful you are to your team in....

winning the match. A guy who runs LRMs all the time and farms lots of damage without really contributing to the teams win (which is the big problem with LRMs) is going to struggle to get into T1 compared to the guy who runs PPFLD and trades like a boss so he tends to kill mechs while doing less than 100 damage per kill. Even more to the point he does less than 100 points to *ensure a kill* on the mech by opening up the CT or removing the torso with the best weapons in it so his team can more easily finish the job.

Win/loss history is the best indication of win/loss in the future.

[MrMadguy]

MischiefSC, on 25 January 2016 - 11:19 PM, said:

Maybe you soaked damage, maybe you were a distraction. You're wanting to generate an overall performance metric based off of one-off situations.

Over 100 matches your behavior has a big impact on the average of your wins and losses. Yes, there will be matches you win or lose because of your team but the exact same goes for the other side; on the average those situations come out in the wash. You're about as likely to win because of your team as lose because of your team. Win/Loss as an average over 100, 1,000 matches though is driven by your performance.

PSR also takes facets of your performance into account so that if you're just a bad tactical player with terrible positioning but you know how to run a brawler and tend to go down swinging you're still going to, on average, raise your PSR faster than someone with bad tactical positioning and who is crap at brawling too.

PSR is not your ranking based on the last match you played. It's your ranking relative to all the other players, on average based on your performance. That's hundreds, even thousands of matches. Win/Loss is the biggest single factor in determining how useful you are to your team in....

winning the match. A guy who runs LRMs all the time and farms lots of damage without really contributing to the teams win (which is the big problem with LRMs) is going to struggle to get into T1 compared to the guy who runs PPFLD and trades like a boss so he tends to kill mechs while doing less than 100 damage per kill. Even more to the point he does less than 100 points to *ensure a kill* on the mech by opening up the CT or removing the torso with the best weapons in it so his team can more easily finish the job.

Win/loss history is the best indication of win/loss in the future.

As wining or losing has nothing to do with personal performance - W/L has nothing to do with skill too. All W/L based MM causes - is matches, where your W/L = 1 at average, but where you perform extremely poorly, so such matches are extremely unenjoyable, unfun, annoying and make you want stop playing. Pyrrhic victorys - aren't what I need, sorry.

[MischiefSC]

MrMadguy, on 25 January 2016 - 11:54 PM, said:

As wining or losing has nothing to do with personal performance - W/L has nothing to do with skill too. All W/L based MM causes - is matches, where your W/L = 1 at average, but where you perform extremely poorly, so such matches are extremely unenjoyable, unfun, annoying and make you want stop playing. Pyrrhic victorys - aren't what I need, sorry.

So good players don't win more than bad players?

You're flat out wrong. Not trying to be offensive here but if you don't impact winning/losing in the match then if you just disconnected every single game your team would still have the exact same odds of winning as if you'd been there?

How you play impacts your team winning or losing. Be that by good suggestions in chat, helping ID targets and using R, covering flanks, support teammates, sharing your armor with the team and all the stuff that goes in to playing well. If you play well you'll win more than you lose.

To pretend otherwise is to pretend that how people play has no impact on their team winning or losing. It always does, every match. Every match is won or lost by the teams playing it. You don't just show up, roll dice and see what happens.

[MrMadguy]

MischiefSC, on 26 January 2016 - 12:01 AM, said:

So good players don't win more than bad players?

You're flat out wrong. Not trying to be offensive here but if you don't impact winning/losing in the match then if you just disconnected every single game your team would still have the exact same odds of winning as if you'd been there?

How you play impacts your team winning or losing. Be that by good suggestions in chat, helping ID targets and using R, covering flanks, support teammates, sharing your armor with the team and all the stuff that goes in to playing well. If you play well you'll win more than you lose.

To pretend otherwise is to pretend that how people play has no impact on their team winning or losing. It always does, every match. Every match is won or lost by the teams playing it. You don't just show up, roll dice and see what happens.

Fat people eat more, then lean ones, but it doesn't mean, that amount of food, you eat, can be used as measurement of your weight. People are different. Same here. Better players win more, but it doesn't mean, that W/L can be used as measurement of their skill. It could be measurement of skill only in binary games, where winning - is all, that matters, and win - the only measurement of performance. Chess for example. But it isn't the case for more complex games. For games, that involve rewards - not only physical, but also psychological.

[Xavori]

MischiefSC, on 25 January 2016 - 05:26 PM, said:

Make a smurf account.

Play in T4/T5.

See people with several thousand matches under them there and struggling.

Understand PSR much better.

I'm actually in T5 because I really enjoy playing light mechs doing the harrassing, targeting stuff that helps teams. This also means my kill/death ratio is meh, and because I play mostly pugs, I'm about 50/50 on wins, and that's only because eventually I quit playing lights and go to heavies where I can carry a team (as noted by being in top 20 on shiny new and temporary warhammer leadersboards).

So ya, I do understand PSR. I understand it has nothing to do with rating a pilot's skill.

[Moldur]

Kilo 40, on 25 January 2016 - 09:54 PM, said:

because when a lance of assaults, that you are not in, decides to "scout" by themselves and they all die quick horrible deaths with under 100 damage between all 4 of them, causing your team to collapse, that obviously means your "pilot skill" is bad and needs to come down.

seriously, the only people who think w/l should effect PSR are those who play in groups and surround themselves with better players, thereby artificially inflating their PSR, and their ego.

4 Assaults pilots don't suicide themselves every round. As stated earlier, the only constant in every game you play is you.


It really depends on perception. If I believe I am affecting the outcome of the game to a large degree, then my PSR being affected by a win or loss makes perfect sense. If somebody else does not perceive that their actions have any influence over the outcome, then PSR makes no sense to them.

[MischiefSC]

MrMadguy, on 26 January 2016 - 12:28 AM, said:

Fat people eat more, then lean ones, but it doesn't mean, that amount of food, you eat, can be used as measurement of your weight. People are different. Same here. Better players win more, but it doesn't mean, that W/L can be used as measurement of their skill. It could be measurement of skill only in binary games, where winning - is all, that matters, and win - the only measurement of performance. Chess for example. But it isn't the case for more complex games. For games, that involve rewards - not only physical, but also psychological.

Actually you can estimate someones weight by the total intake of calories they eat pretty effectively. There's a bit of variance but that's one of the best actual indicators you can use.

W/L is a measurement of.... how often you win relative to losing.

Good players win more often than bad players when playing at the same level. Yes? No?

How people play affects the match. Yes? No?

You're trying to argue that you playing doesn't actually affect how likely your team is to win or lose a match. That's not just wrong but demonstratively so. If you play with the 11 best players in the game and I play with the 11 worst you're going to roll me so badly we're all going to feel terrible we had to watch it happen. We have a matchmaker and tier system specifically because better players sway matches to a win when playing against worst players.

This is like trying to argue that the world is flat.

How good or how poorly someone plays affects the odds of their team winning or losing. Averaged over a lot of matches it's pretty easy to identify from that how likely they are to help a team win or lose a match.

Are there more metrics to it? Sure, that's why PSR uses more metrics. End of the day however their only purpose, the only reason that damage/kills/assists/whatever matters, is in how it affects your teams ability to win the match. How much of an impact you have on winning against opponents of a particular skill profile.

Yes. You playing on a team impacts how that team performs and how a team performs drives which team wins and which team loses.

[MrMadguy]

I know, why PGI doesn't want to remove W/L-based rating, as long as implement skill-based one and why MM is so biased towards winning. It's very simple - they won't be able to open release valves and mix Tier 5 and Tier 1 in one team then. Why? When it's 3xTier 1 vs 3xTier 1 - it's ok. But imagine Tier 1+Tier 2+Tier 3 vs Tier 1+Tier 2+Tier 3. What will happen? It will be essentially Tier1 vs Tier 1 - they'll perform great, and Tier 3s will be just a punch bags - they will perform extremely poorly. It's 50/50 situation in terms of W/L - so in case of having W/L-based MM ratings won't change. But if we'll have skill based MM, then Tier 3s will start losing their ratings very rapidly. Do you realize consequences? Deflation of lower tier ratings. And bias towards winning helps to avoid it: you perform extremely poorly, but if you win anyway - your PSR won't drop. So... Lower skill players will always be just a punch bags for higher skill players. I guess, I have to accept, that the only way of dealing with this situation - is to play this game at prime time only or to quit it.

Currently it's impossible to grind enough CBs even to equip DHS on my new 'Mech... I'm not talking about buying new one. Is PGI expecting lower skill players to pay $$$ for everything? It's some sort of hidden P2W. And I don't want to play P2W game, sorry.

Lower skill players really need to have reward compensation for being mixed with higher skill players. Or let's reduce team sizes and return to 8vs8 - it worked much better.

Edited by MrMadguy, 26 January 2016 - 02:22 AM.