490 replies to this topic

[crazytimes]

Revener, on 05 December 2020 - 07:40 AM, said:

oh I see , well that was one crap formula that basically only let me play assaulty mechs (since it is easy to do damage in them = easy match score) if I want my PSR to go up, that is boring.

My road to tier 1 was paved almost entirely by mediums and heavies in the 45-65 ton range, with the exception of a fair few in an urbie which I suppose counts as an assault.

If you find yourself struggling to grind tier... maybe that is an indicator that tier is neither something that should be ground nor an achievement in itself. It just means better skilled opponents. If you try and game the system to end up there, you're going to suffer when you do.

Focus perhaps on getting better, and let nature unfold from there.

[Revener]

crazytimes, on 06 December 2020 - 10:53 PM, said:

My road to tier 1 was paved almost entirely by mediums and heavies in the 45-65 ton range, with the exception of a fair few in an urbie which I suppose counts as an assault.

If you find yourself struggling to grind tier... maybe that is an indicator that tier is neither something that should be ground nor an achievement in itself. It just means better skilled opponents. If you try and game the system to end up there, you're going to suffer when you do.

Focus perhaps on getting better, and let nature unfold from there.

Well my point was mostly that it is easier the more damage you do = easier to do it with assaults.
Although playing AMS umbrella with a nova or kit fox seemed to work too.

[crazytimes]

Revener, on 16 December 2020 - 07:45 PM, said:

Well my point was mostly that it is easier the more damage you do = easier to do it with assaults.
Although playing AMS umbrella with a nova or kit fox seemed to work too.

Personally, I do more damage in mediums and heavies than assaults, because I'm not a particularly good positional player and being able to reposition better keeps me alive long enough to do something useful. I understand that issue and will work on it, but that's a me problem.

Most of people I've seen saying "play assaults to grind tier" are 1) somewhere between bad and average 2) particularly bad in assaults. The difference is I know I am bad in assaults and don't pretend they're an easy way to appear good. Someone who actually is great in assaults is not going to making up excuses about tier mechanics.

As for AMS, it's like LRM Fafnirs. Yeah you might have a really high MS on Polar once, that's offset by the next 5 games on Mining collective and Solaris City.

[VonBruinwald]

nuttyrat, on 04 June 2020 - 10:46 AM, said:

Glad to hear that this is coming! Taking to Kurlon he brought up a good question .. how big is the PSR bar in each Tier?

Paul has never gone into detail on this.

From this post we glimpsed the following:

Paul Inouye, on 04 May 2020 - 12:52 PM, said:

Average Tier separation between teams: ~300 (tightened from ~1800)

If I was to speculate I would say the PSR bar is goes from 0-2500 in 500 increments. Possibly 0-2000 in 400 increments.

If you look at the images in that post when PSR was tightened you got spreads below 200 which I would guess is single tier games.
When it was loosened off you got a spread of 1800 which could conceivably cover all 5 tiers.

[crazytimes]

VonBruinwald, on 21 January 2021 - 05:43 PM, said:

If I was to speculate I would say the PSR bar is goes from 0-2500 in 500 increments. Possibly 0-2000 in 400 increments.

If you look at the images in that post when PSR was tightened you got spreads below 200 which I would guess is single tier games.
When it was loosened off you got a spread of 1800 which could conceivably cover all 5 tiers.

0 to 5000 in single point increments. 5000 is tier 1 maxed. Each game can shift you between about -24 to +34.

1800 spread is horrendous for team difference. It means 1 team can average 4500- mid tier 1- and the other 2700, all well down tier 3.

[Hier koennte Ihre Werbung stehen]

Revener, on 16 December 2020 - 07:45 PM, said:

Well my point was mostly that it is easier the more damage you do = easier to do it with assaults.
Although playing AMS umbrella with a nova or kit fox seemed to work too.

AMS is a real MatchScore Booster.

[Cichol Balor]

This change does not make team work critical... it makes scoring high critical if I can always get 400+ ill keep going up but if I get under 100 the best I can hope for from a win is = but risk dropping 5.


So rather than risking myself to benefit the team I'm better off just taking builds and playing in a way that ensures a high score my team be damned.


If score was actually a reflection on team contributing rather than being so inherently tied to damage this might work but as it stands anyone who cares about PSR is pushed hard into running heavier mechs rather than playing what they enjoy/are good at.

This is already putting people who play heavier mechs into tiers they are not having fun in because their skill isn't actually there and holding higher skill players out lower tiers using lighter mechs.

PSR should be tied more inherently to your win/loss ratio. If you are winning significantly more often than you are losing you should go up as odds are you are carrying your teams. if you are losing more often than you are winning you need to be dropped back down.

teams are random so the more games that are played the more the ratio of being on a good team vs a bad team will equal out so the only constant variable is you tipping the odds for how often you are wining.

So if you are playing in a way that results in you having a high score but ultimately isn't benefiting your team your not going to keep going up. And if you are playin in ways that help the team out but don't score well your not going to be penalized in PSR.


Phytologically this would be detrimentally as any streak of being on bad teams would leave a player as feeling helpless even if they were no more common than being on a streak of good teams.

To combat this it would be better to hide the PSR and then add in a seasonal ranking that was displayed to the player and went up or down based more on the system the current PSR uses.

Edited by Cichol Balor, 22 January 2021 - 03:02 AM.

[Anomalocaris]

crazytimes, on 21 January 2021 - 07:49 PM, said:

0 to 5000 in single point increments. 5000 is tier 1 maxed. Each game can shift you between about -24 to +34.

1800 spread is horrendous for team difference. It means 1 team can average 4500- mid tier 1- and the other 2700, all well down tier 3.

And that's why secondary balancing after the 24 players are selected is needed. You'd never have a very big spread because math

[Pappadakka]

Player LOSES:
Match Score: 0-100 goes down in PSR by -3
Match Score: 101-250 goes down in PSR by -1
Match Score: 251-400 =
Match Score: 401+ +1

Player WINS:
Match Score: 0-150 psr -1
Match Score: 151-250 goes up in PSR by +0
Match Score: 251-350 goes up in PSR by +1
Match Score: 351-450 goes up in PSR by +3
450+ +5

This seems more accurate currently because i know i have seen my PSR go up on a lost

[Aidan Crenshaw]

Wrong, Pappadakka. This was the old system, the new one is much more sophisticated. Check Brios' excellent YT vid about it!
https://youtu.be/JFqlZXOVtdw

[Dread Priestess Shakira]

I know this thread is rather old, but so far I have never seen anybody mention if Matchscore is calculated differently each match based upon win conditions, and/or mech weight class or loadout and skill build. All of these factors could have significant impact. A light mech that does not fire a single shot during a conquest match, but individually secures 375 resources from base occupations should most certainly be rewarded for playing to the match objective and contributing so much to their team's win. perhaps more so if they are skilled specifically into capture assist, as it indicates they are playing their intended role during the match, while contributing so significantly to their team's win. it is also unfortunately that there is no way to accurately verify a player's contributions through verbal communication with their team. balancing this function certainly proves to be no simple task. thank you to all the developers committed to attempting to resolve this to the best of the game's capacity given the cross-section of active players we are working with at any given time. i can easily understand how your could offer several different methods of calculation for matchmaking and not be satisfied with a single one for some reason or another, as each one has its own unique strengths and issues.