Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
(wL) RWS related to other stats
#2
I thought about making my analysis on the same post .... but nah its way too damn long already.

Are most of these stats linked to RWS ?

Yes, but not at face value, ill explain.

At first view, only ADR and FPR have an almost perfect track record (zero high positive and negative ratio relative to RWS). FPR is obviously less linked to RWS, since rws doesnt consider kills but damage like adr does (but in the end, a kill does damage).

KDR is heavily influenced by what gun the player is using, but in general it has a pretty decent link to RWS (far from perfect). If controled for that it could be a valuable statistic, maybe on equal footing to FPR

ACC% and HSP% are mostly linked to what guns said players are using, like KDR, but it doesn't consider if you die or kill people (unlike KDR), so at face value it isn't as clearly related to "skillz". But if we were to control for what % of guns people are using. We would be left with the ACC% and HSP% of everyone for every gun, meaning we could potentially rank accurately whos best with what, but also calculate the average acc and hsp% for every guns without considering gun usage %. In other words, both of these could have a strong link to RWS, just not how it is currently displayed (even more if we consider KDR - per gun).

And now .... what are we left with ?

Win% (W-T-L, W/L they are roughly the same)

and

RPA (Rounds Played Average)

They aren't strongly related .... until you start digging, RPA's job is to confirm that all the other stats hold value, in other words, if you leave like systematicly (RQ, skipping bad games, joins already begun pug all the time to stack) it will point it out clear as day. It shouldnt be used to be compared 1 to 1, between specific players, but ... when the gap is big enough it can be used that way.

So if good, decent, average, below average and bad players, leaves a lot, it will have an impact on the W%.

For example :

Juniorc4 18.58 (4/4) RWS / 131.6 (3/3)  ADR / 2.06 (2/3) KDR / 1.310 (2/2) FPR / 50.8496  (5/26) W% / 18.088 (5/33) ACC% / 19.631 (2/5) HSP% / 18.9962 (5/66) RPA

Out of 70 players he has the 66th RPA, which might put into question the validity of his other stats, but hes a good player enough that despite him leaving that much (missing 5 rounds on average per game), he stills holds a 50% Win% somehow. Still, his other stats like adr, fpr and kdr is higher then ghoul's and loopz. So why isnt he top2 RWS, id say its because RWS is influenced by his Win% and RPA, in the long run, because if you leave when doing badly or because its boring, yes on the short term you will save/bump your stats, but it will slighltly affect your rws on the long term, because by leaving you aren't doing the 10 other rounds were you would of done decent or maybe even start fragging, or gain RWS back (because RWS isnt only linked to fragging). Also if you have a bumped RWS, you will get more and more unfair teams, so if you leave to skip bad games, well these bad games will get more and more frequent, also somewhat affecting RWS (it will still be bumped, but not enough to be shown as better then better players, only similar level players).

An other example (to the other extreme) :

Ice            05.61 (2/67) RWS / 37.9 (4/69)  ADR / 0.38 (4/69) KDR / 0.287 (4/69) FPR / 56.9620 (1/8) W% / 10.085 (4/68) ACC% / 7.157 (3/68) HSP% / 19.3181 (5/64) RPA

This player is so ..... bad .... euh ..... i meant .... low in rws that because he leaves ALL the time, well who is replacing him is way better then him (literally anyone replacing him), meaning every time he leaves his team has way betters odds of winning. Which by accident, through no effort of his own, makes him have a good WIN% (you can reverse that logic for juniorc4 too).

In other words, RPA, Win% and RWS have also a strong relationship between each other.
[-] The following 2 users Like Ben-'s post:
   QueenCrazy, rookoo

Messages In This Thread
(wL) RWS related to other stats - by Ben- - Sep 15 2018, 03:54 AM
RE: (wL) RWS related to other stats - by Ben- - Sep 15 2018, 04:23 PM
RE: (wL) RWS related to other stats - by woKe - Sep 15 2018, 06:52 PM
RE: (wL) RWS related to other stats - by Ben- - Sep 16 2018, 01:40 AM
RE: (wL) RWS related to other stats - by rookoo - Sep 16 2018, 01:52 AM
RE: (wL) RWS related to other stats - by woKe - Sep 16 2018, 02:56 PM

Users browsing this thread: 1 Guest(s)