YvngKurry [#40]
but you used acs to justify that his performance wasnt abysmal at cph. why cant i use acs then?
sorry, I wasn't clear about my point so let me explain better
ACS is a simple and useful indicator of performance, even though it's not fully accurate
the problem is not with the variable itself, since every basic vlr stat is incomplete on its own
the problem is in comparing avg ACS (or any indicator for that matter) because:
-
the tournament in which pancada didn't do well, M2, is the tournament that Zeta didn't even qualify, so their insufficient regional performance which demerits the players is being masked by the abscence of their potential M2 stats.
-
you need to add different weights to the ACSs of each tournament depending on their sample size. If you use your lowest granularity as maps for example, you have to add all the maps played from all tournaments and then divide the result by the number of maps. Idk how much ACS pancada and sugarzero would have if you calculated that way, and that's not the point btw, I'm just explaining my point of view of handling inter-tournament stats