277 replies to this topic

[An Atlas]

Chemie, on 25 September 2015 - 03:39 AM, said:

Finer point:
6/8 ways to increase or stay the same
2/8 ways to drop

Hard to not end up T1 regardless of skill.

If you ignore the part where you have to keep getting better at the game... sure "hard not to"

[Widowmaker1981]

The Paulconomy, on 25 September 2015 - 03:28 AM, said:

Oh really?



Translation: Tiers work, b/c people better than me are required to artificially lift me out of my current Tier



Translation: Tier doesn't matter b/c you can just play a lot and get the same result as being good.

Ok, ill try one last time. All i am saying here is that group queue games warp PSR, and that it shouldnt, because it should be an individual skill rating, not a what company you keep rating. Take the following completely fictional example:

My girlfriend starts playing MWO, because she wants to play games with me. Shes obviously not very good, since she has never played a game like this before. She then plays, but plays mainly in group with me and my unit, because that is the experience she is looking for. So shes in the group, and we try to teach her. We are easily capable of carrying one player, so we win all our games as normal. She gets the odd lucky game where she is ignored and gets a good score, though mostly not pulling her weight. Since its impossible to lose PSR in a win, her PSR constantly increases, and eventually reaches T1. She now tries to play in the solo queue, because shes got hooked on the game, and the system thinks she is a T1 player and matches her with the best in the game, so she gets annihilated. Is that a good outcome? Shes not T1 in actual personal skill...

There are a number of possible fixes to this issue, and to my mind the best one is to remove the win/loss component of PSR. That way its possible for a poor player to play with a group of elite player friends without killing their solo queue experience.

Edited by Widowmaker1981, 25 September 2015 - 03:59 AM.

[Gnume]

In my opinion, their biggest flaw in this is tying it to the Win/Loss of the Match. That is NOT an indicator of Pilot Skill. That is an indicator of TEAM PERFORMANCE (or lack thereof). PSR should ONLY take into consideration things an individual pilot do to contribute to the match. Everything you do should count somehow. Doing damage, countering ECM, Disrupting targetting via ECM, UAV's, Shooting Down UAV's, protecting a teammate (Savior Kill, Protected Light, etc.), Spotting Targets, Capturing Points, DRIVING OFF ENEMY MECHS from a Capture point/Base, and so on. The outcome of that Match should NOT determine "how much of a rise or fall" happens with your PSR.

[Random Carnage]

Having this much discussion about it already makes it a flawed system, regardless of any actual mechanic.

[TWIAFU]

Vlad Ward, on 24 September 2015 - 06:38 PM, said:

I'd imagine the "extremely low score + win = no gain" condition is only there to prevent discos and afks from farming free PSR. Outside of this edge case pretty much any win will count as a plus.

You can Win with a match score of 50 and break even.

So minimal amount of effort on a win to not change in PSR.

[Lugh]

Jin Ma, on 24 September 2015 - 05:51 PM, said:

Not a big fan. If all it does is protect new players. Then it should never have been displayed.

Seems like a shotty system, not sure what its goal is. Not sure if it is able to accomplish it.

Seems like a mashem together of

• Leveling System - to protect new players(LoL)
  
• Ladder System - for ranked play (Starcraft2)
  
• ELO - Matches players of equal skill

If you want my honest opinion, i'd describe it visually as some disgusting fused abomination that screams "Kill Me"




It does none of the above things well. If everybody eventually reaches Tier 1/Tier 2 then how does it match players various skill in the game against each other like the old elo system did?

Like the old one, it doesn't.

They don't have the faintest idea how to do that well, because they have trouble with SIMPLE math, let alone anything that requires an algorithm.

Edited by Lugh, 25 September 2015 - 04:30 AM.

[TWIAFU]

Random Carnage, on 25 September 2015 - 02:26 AM, said:

Playing a lot is no guarantee that a person will get better.

233 posts?

Quote

Your unique perspective challanges and refreshes us. If I gave a ****, I may defend myself. Somehow I can't muster the energy to explain things to someone like you.

See, I can pull that BS too.

Nope, they won't get better by doing the same thing over and over every match.

Won't get better by refusing to be a team player in a team/group based game.

Won't get any better by blaming everyone and everything else for one's own failings.

Won't get any better if the pilot does not look in the mirror and find ways to self improve instead of wanting PGI to make the game more derp friendly.

Edited by TWIAFU, 25 September 2015 - 04:27 AM.

[Belazaar]

RussianWolf, on 24 September 2015 - 06:06 PM, said:

And there is the falicy in his thinking.

To him "better" means conforming to the play style that PGI perscribes. Those who do not conform to that playstyle, regardless of skill level, will have lower Tier rankings.

I very rarely conform to their perscribed play style, but would be described by those that I play with as average to better than average in skill. Yet here I am in Tier 5.

Now I'm actually happy about being Tier 5 instead of going ...WTF.... Why, you may ask? because I have never wanted to play this as a soley team based game. I've been an early demander of bringing in Solaris VII and free-for-all matches. I have no desire to conform to their idea of how to play.

And this is why the team focus won't work. Not everyone on the team has the same idea as what team play is or how to do it.

There is no standard way to play. There are different play styles and strategies. So as long as their are random drops the solo que will never work. The only way to do this is by grouping with like minded people.

[PyckenZot]

CaptRosha, on 25 September 2015 - 04:05 AM, said:

In my opinion, their biggest flaw in this is tying it to the Win/Loss of the Match. That is NOT an indicator of Pilot Skill. That is an indicator of TEAM PERFORMANCE (or lack thereof). PSR should ONLY take into consideration things an individual pilot do to contribute to the match. Everything you do should count somehow. Doing damage, countering ECM, Disrupting targetting via ECM, UAV's, Shooting Down UAV's, protecting a teammate (Savior Kill, Protected Light, etc.), Spotting Targets, Capturing Points, DRIVING OFF ENEMY MECHS from a Capture point/Base, and so on. The outcome of that Match should NOT determine "how much of a rise or fall" happens with your PSR.

Widowmaker1981, on 25 September 2015 - 03:50 AM, said:

There are a number of possible fixes to this issue, and to my mind the best one is to remove the win/loss component of PSR. That way its possible for a poor player to play with a group of elite player friends without killing their solo queue experience.

THIS! PSR is in dire need of this specific little fix!

Edited by PyckenZot, 25 September 2015 - 04:29 AM.

[Belazaar]

The Paulconomy, on 25 September 2015 - 02:11 AM, said:

So players that cower instead of working as a team will go down in rank?

SOUNDS LIKE IT'S WORKING PERFECTLY

Or players that keep blocking bottleneck or popping over a hill get killed first and say everyone else on the team is cowered because the took cover and waited for the red team to come out from cover.

Now staying back with the team is advancing is cowering.

[Belazaar]

Chemie, on 24 September 2015 - 05:05 PM, said:

"You are Tier 4/5 because you do not play as a team."
....but...
PSR can not drop if your team wins (regardless of how you played).

?? It is a win/loss plus games played counter but somehow that equates to teamwork?

You can get a zero match score but if your teams carrys and wins, you don't drop PSR. Go teamwork!

So what he is saying is we are not playing people of equal skill, we are also playing solo players who are good at killing mechs, but not will to play as a team.

This should work out... right.

[Doman Hugin]

From many other posts and actually playing the game I know “team work is OP”, and they make win / loss a major factor in PSR. Sound about right to me.

If you can score 298 and go “down” on a loss, the next time you play that well on a win will probably get more points and a larger “up”, overall netting a positive gain.

Conversely scoring badly on a win might be a small “up”, but play that bad on a loss could very easily net an overall loss.

And having a positive win / loss could get you to tier 1, but remember you’ll need to carry that positive win loss through T4, T3, & T2 before you do.

I’m in favour of tweaking the game score values to match the game play, but as current game play is basically CODWarrior Laservomit Online, the current values are about right.

Edited by Doman Hugin, 25 September 2015 - 04:54 AM.

[INKBALL]

It's a shame that the match score isnt class related. It makes it WAY easier for assault to get upper in tiers:

Heavier mech always have more damage potencial then lighter. And in the winning cases, lights have to take risks, to get through enemy lines and backstab (which could result in a terrible game (lost/win and no/little damage), and while the fatass assault just sit and wait and worst case lose with good damage..

And, whats the point of making tiers if eventualy everyone can be t1-2 ? Dont tell me that top lords/emp/sjr/AS (and name it), can play with average players (not even speaking of under average and still t2-t3)...

[Johnny Z]

CaptRosha, on 25 September 2015 - 04:05 AM, said:

In my opinion, their biggest flaw in this is tying it to the Win/Loss of the Match. That is NOT an indicator of Pilot Skill. That is an indicator of TEAM PERFORMANCE (or lack thereof). PSR should ONLY take into consideration things an individual pilot do to contribute to the match. Everything you do should count somehow. Doing damage, countering ECM, Disrupting targetting via ECM, UAV's, Shooting Down UAV's, protecting a teammate (Savior Kill, Protected Light, etc.), Spotting Targets, Capturing Points, DRIVING OFF ENEMY MECHS from a Capture point/Base, and so on. The outcome of that Match should NOT determine "how much of a rise or fall" happens with your PSR.

Ok alot are saying that this tier shouldnt be tied to win/loss. Well your wish is granted lol. Its not tied entirely to win/loss only partly. Saying otherwise isnt right.

Win/loss should have no effect what so ever? Really? With a little thought I think most agree it should have some bearing on a players tier even if not entirely.

I cant see a way for them to improve on what appears to be the best threat system even made lol. Maybe other games have an equally well devised threat system but this one looks good to.

Just repeating myself here but going only on win/loss maybe not be right and completely ignoring win/loss doesnt seem right either so thats what Mechwarrior has.

And again this isnt about "ranking" players or player "skill", its about making good matches and protecting newbies. That seem to be forgot alot on this topic.

For player ranks look to the faction wars and solaris. Two entirely(I assume who knows what Solaris brings) different ways to be ranked among other players. The factin wars will allow units to compete and maybe some Solaris modes will allow teams to compete? If not individual ranks or titles or trophies or something on Solaris.

Edited by Johnny Z, 25 September 2015 - 05:13 AM.

[Veritae]

I made a thread yesterday about these problems before Paul posted, identifying the issues in the current psr system.

The answer is to ignore W/L, assign players to a skill tier based on their average match score. Where players fall on the bell curve determines their tier. Ideally, have a separate value for solo and group queue, or break it down further based on weight class.

If everyone but new players will be tier 1 or 2, then this is just a level up experience system.

How is matchmaker supposed to build good games if it is saying we're all equal special snowflakes?

[Johnny Z]

Veritae, on 25 September 2015 - 05:15 AM, said:

I made a thread yesterday about these problems before Paul posted, identifying the issues in the current psr system.

The answer is to ignore W/L, assign players to a skill tier based on their average match score. Where players fall on the bell curve determines their tier. Ideally, have a separate value for solo and group queue, or break it down further based on weight class.

If everyone but new players will be tier 1 or 2, then this is just a level up experience system.

How is matchmaker supposed to build good games if it is saying we're all equal special snowflakes?

See maybe your a special snowkflake. The reason I say that is because you seem fixated on it. Its not about special snowflakes its about making good game matches.

Also it seems this arguement about player skill as opposed to determining a players threat level is best saved for a topic on Solaris rankings. What ever or how ever that is done. Monthly and yearly trophy? Trophy for FFA or duels? Who knows.

To go over a previous reply, this system is determine players threat level and to keep high threat players from facing low or mid range threat levels. 1 to 3 is highest to high, 4 is mid and 5 is low. Low being new to this kind of game entirely, afk or bad ping, or maybe a newbie that just had a couple bad starting matches.

Edited by Johnny Z, 25 September 2015 - 05:28 AM.

[RussianWolf]

A short message from our Tier 5 sponsors.......

[Gnume]

Johnny Z, on 25 September 2015 - 05:06 AM, said:

Ok alot are saying that this tier shouldnt be tied to win/loss. Well your wish is granted lol. Its not tied entirely to win/loss only partly. Saying otherwise isnt right.

Win/loss should have no effect what so ever? Really? With a little thought I think most agree it should have some bearing on a players tier even if not entirely.

I cant see a way for them to improve on what appears to be the best threat system even made lol. Maybe other games have an equally well devised threat system but this one looks good to.

Just repeating myself here but going only on win/loss maybe not be right and completely ignoring win/loss doesnt seem right either so thats what Mechwarrior has.

And again this isnt about "ranking" players or player "skill", its about making good matches and protecting newbies. That seem to be forgot alot on this topic.

For player ranks look to the faction wars and solaris. Two entirely(I assume who knows what Solaris brings) different ways to be ranked among other players. The factin wars will allow units to compete and maybe some Solaris modes will allow teams to compete? If not individual ranks or titles or trophies or something on Solaris.

It is totally tied to W/L. You don't go down AT ALL if you Win...and you only rise a small amount if you are exceptional in a loss.

I am saying take out the Win/Lose portion of the calculation because that is not a function of Individual Skill. The Win/Loss is a function of Team Performance. Individual performance within the team will contribute to winning or losing, but it doesn't work the other way around.

You will still end up with matches against players within your skill range. I would think moreso without the W/L added because your PSR was totally dependent on YOUR performance, not whether you were carried to a win by a team a lot. People that play with Units that predominantly win their matches can artificially inflate someone's PSR with the current system just because they win. In what way does that reflect upon an individual's skill?

[Bilbo]

Agent 0 Fortune, on 24 September 2015 - 06:14 PM, said:

There are going to be players who after 5000 games (or any number of games) will not be at the same player skill level as other players, unfortunately with the PSR system, everything converges at tier 1.

You could literally have a player who plays nothing but stock Urbanmechs, using a steering wheel as a controller, and playing blindfolded, who will eventually reach tier 1, and play against the best players in the game.

That is the problem with this system.

I am a prime example of one reason this isn't the case. Because of the way I choose to play, when I fail I fail epically. So even with a positive win/loss I probably wont come out of even tier 4 much less surpass tier 3. I'm fine with that and I'm certainly not going to change a playstyle I enjoy simply to move up tiers.

[Veritae]

Johnny Z, on 25 September 2015 - 05:20 AM, said:

See maybe your a special snowkflake. The reason I say that is because you seem fixated on it. Its not about special snowflakes its about making good game matches.

Also it seems this arguement about player skill as opposed to determining a players threat level is best saved for a topic on Solaris rankings. What ever or how ever that is done. Monthly and yearly trophy? Trophy for FFA or duels? Who knows.

To go over a previous reply, this system is determine players threat level and to keep high threat players from facing low or mid range threat levels. 1 to 3 is high to higest, 4 is mid and 5 is low.

How can that be? The system is designed to funnel us all into tier 1. How can the matchmaker make good games if we're all the same tier? It can't, can it?

So, this is a very poor design to build good matches. Getting 100 match score on a win advanced your psr. How is that person demonstrating a "high threat level"? His team just improved his experience bar. That's all.