PSR is a variable that is changed by adding or substracting a value according to game results - therefore it reflects your total standing. It has no real memory of what was before - just where you are at. And by producing game results you alter it.
It would be much harder to make it so that a number is calculated to reflect only last 100 matches. Such a system may produce even worse, erroneous results.
Most importantly it is a good concept in itself - it's just the implementation that is bad, in my opinion. The system needs to converge to a real value (or rather an interval in which you move up and down, because that is the real result) which is a reflection of your skill. Right now it is a play more-have more variable.
1
Psr Issues And Why It Is Going To Break The Game In The Long Run
Started by MrKvola, Oct 01 2015 08:37 AM
General Gameplay
21 replies to this topic
#22
-
-
- The Widow Maker
- 5,031 posts
Member
- LocationAt the other end of the pretty lights.
Posted 02 October 2015 - 02:11 AM
three things should be changed and PSR would be good.
1) The tier boundaries should be relative (i.e. moving target. same top X% of playerbase stays at T1. NOT entire playerbase converging on T1)
2) Group queue PSR should be seperated from solo queue PSR. Or Group queue win/loss should not affect PSR
3) PSR should be split by chassis type (L/M/H/A) as Elo was. An absurdly good assault specialist might suck in lights, and vice versa.
1) The tier boundaries should be relative (i.e. moving target. same top X% of playerbase stays at T1. NOT entire playerbase converging on T1)
2) Group queue PSR should be seperated from solo queue PSR. Or Group queue win/loss should not affect PSR
3) PSR should be split by chassis type (L/M/H/A) as Elo was. An absurdly good assault specialist might suck in lights, and vice versa.
Edited by Widowmaker1981, 02 October 2015 - 02:11 AM.
3 user(s) are reading this topic
0 members, 3 guests, 0 anonymous users
