Jump to content

Psr Update And Changes - Jun 2020


462 replies to this topic

#461 Mustang_11b35d

    Member

  • Pip
  • The 1 Percent
  • The 1 Percent
  • 16 posts
  • LocationArizona

Posted 14 August 2020 - 10:56 AM

Match Score: A measure of skill? If "skill" means laying back and sucking down late damage, or lobbing LRMs from a safe distance, then then the current system is right on the money! However, if skill means working the field maneuvering, sharing armor, and sometimes giving yourself up for the team, then it pretty much sucks. I'm content being a level 3 player, if becoming a level 2 player requires me to be more selfish.

Can we weight the team-oriented "positive behaviors" (for lack of a better term) such as capping, scouting, effective UAVs, assisting teammates in trouble, etc. more heavily? That might make the match score better at reflecting more than just damage. Damage is important, but it's not everything. It's supposed to be a team game, last I checked.

#462 GARION26

    Member

  • PipPipPip
  • Giant Helper
  • 90 posts

Posted 14 August 2020 - 01:23 PM

View PostMustang_11b35d, on 14 August 2020 - 10:56 AM, said:

Match Score: A measure of skill? If "skill" means laying back and sucking down late damage, or lobbing LRMs from a safe distance, then then the current system is right on the money! However, if skill means working the field maneuvering, sharing armor, and sometimes giving yourself up for the team, then it pretty much sucks. I'm content being a level 3 player, if becoming a level 2 player requires me to be more selfish.

Can we weight the team-oriented "positive behaviors" (for lack of a better term) such as capping, scouting, effective UAVs, assisting teammates in trouble, etc. more heavily? That might make the match score better at reflecting more than just damage. Damage is important, but it's not everything. It's supposed to be a team game, last I checked.


Look I agree with the idea there are a lot of things that go into 'skill' and that match score may not be perfect.

BUT go to this list: https://leaderboard....g/?o=ams&d=DESC
Sort by match score then look at win loss ratio
I just flipped through (again) and looked at the first six pages of top match scores - no one with a WL at or below 1:1 even ignoring the small sample size players you still see the first page of the sort averaging WL of 2-5:1
KD also likely is a similar story when you look at it compared to match score but I agree its possible that some great players might not get a lot of kills.

Match score certainly may have faults - but it sure does seem to associate closely with player 'skill' which IMO is should correlate with the the team you are on winning more often then it loses.

Sometimes an LRM mech is going to have a great matchscore - but it's also going to have some stinkers where a light mech pack eats it 90 seconds into a match. Consistent high effectiveness play is where you get your high average match score (and consistent PSR elevations if you care about that thing)

Edited by GARION26, 14 August 2020 - 01:35 PM.


#463 MasterBurte

    Member

  • PipPipPipPipPip
  • The Privateer
  • The Privateer
  • 124 posts

Posted 15 September 2020 - 03:45 AM

This is a great improvement to the old system. It's important that players are able to stagnate at a tier when they are not good enough. But this system is still not perfect.

Pros:
The W/L ratio is a good indicater for player performance. Maybe even better than match score. Why? W/L ratio is a statistic that is independend from play style. It only matters, if the play style is succesfull in the majority of the matches. It doesn't matter if bad players get carried or good players carry their team when the match very even. After a few hundreds of matches there is no way that the bad player has a better W/L ratio than the good player.

Cons:
The new system is heavily dependend on the matchscore. The matchscore is not a good indicator of the players performance. In fact it is very difficult to measure the individual contribution to the game. Matchscore is mostly dependend on damage output, wich favours heavier mechs. Lights that destract the enemies and backstab them with high precision are not represented equal to their battlefield value.

Solution:
To let the matchscore represent the players performance better it should be dependend on tonnage and heavier influenced from supporting behavier.
1. Damage = (20 x dmg)/(20 + tonnage)
2. Tactic score = Hit and Run, UAV kill, Flanking, Kill, KMDD, Solokill etc (summed up)
3. Matchscore = (multiplicator x Damage score + multiplicator x tactic score)

The optimal weight between the scores can balanced through different values and multiplicators. UAV kill could be 20 points, Hit and Run 10 points, etc..





2 user(s) are reading this topic

0 members, 2 guests, 0 anonymous users