Quite a lot of things actually:
1)Why is there a one game delay in your true CSR?
There’s no reason for it.
2)Have CSR based roughly 75% score and 25% K/D instead, not just 100% score based
Might sound strange, but compare two players, A and B in games of BTB
A is a bum rusher, who simply throws grenades left right and centre, and overall has about the same number of kills to deaths per game. We’ll say they go 25 for 23 in a game
B is a person who sits back a bit (not a camper), doesn’t rush in a lot, yet gets say 20 kills and 8 deaths, a respectable contribution to the team.
Yet because B simply doesn’t get as much score as A, he therefore comes lower placed and is asssumed by CSR to be a worse player.
No, B is the better player overall, because they can get a fair amount of kills, but at a cost of fewer deaths than A.
See the point?
At the moment it is “He who gets the highest score is the best, even if they died 40 times getting it”
Why should players be treated as playing a full game when they JIP say 1/3 way through a game, yet get penalised simply because they weren’t able to get as much score in the shorter period. This ties in with points 1 and 2.
Surely it could be guessed roughly how well a player would have gone if they’d played the full game, by extrapolating how they went in their shorter game?
Say a player goes 12 -6 in a game where they only play 2/3 game due to JIP. Most people would agree that the player, if they’d played the entire game, would have gone roughly 16-8 (adding an extra third onto the K/D)?
In the end CSR is broken at the monent because it doesn’t recognise a players long term skill (K/D is a good estimate in the end), instead only recognising the players long term score, not accounting for the cost at which they may have got this score (Eg: Player A…)