I assume you used end-game values? I think it's hard to extract anything from that because it's not at the point of winning(the "tipping point" if you will). Once you start winning the game it's only logical to get more kills, farm and objectives than the losing team. Probably less so with farm though.
Anyways, that could make a potential kill lead at the tipping point of 1 to a kill lead of 20 at the end game screen, making your coefficients not very relevant in my opinion.
Thanks for the share, numbers are never bad :)
Anyways, that could make a potential kill lead at the tipping point of 1 to a kill lead of 20 at the end game screen, making your coefficients not very relevant in my opinion.
Thanks for the share, numbers are never bad :)
********'s a pretty good fertilizer
Latest Legend wrote:
I assume you used end-game values? I think it's hard to extract anything from that because it's not at the point of winning(the "tipping point" if you will). Once you start winning the game it's only logical to get more kills, farm and objectives than the losing team. Probably less so with farm though.
Anyways, that could make a potential kill lead at the tipping point of 1 to a kill lead of 20 at the end game screen, making your coefficients not very relevant in my opinion.
Anyways, that could make a potential kill lead at the tipping point of 1 to a kill lead of 20 at the end game screen, making your coefficients not very relevant in my opinion.
The stats are from LoLking, so they are end-of-game values. Good point, but I would say the converse happens as well - dying more when you're behind - and they probably balance each other out. Also there have been some articles on LoL Esports about farm scores when a given team is winning or losing, and some teams actually farm more or less when they're losing or winning, I didn't look at that explicitly.
I'm assuming that the crazy snowballing that occurs end-game biases an even amount in both directions; when you lose hard, there is an instance where you won hard to balance it out. This probably isn't a great assumption, but since my sample size is pretty large and consists of many different champions, it's probably negligible. It's highly dependent on the player as well, so I encourage you (and others) to analyze your games similarly and let me know what you get.
Check out my blog Baron v. Squid.
You need to log in before commenting.
I recently did some statistics with my season 4 games, and correlated Win % to other statistics like CS score, kills, deaths, KDA, and so on.
Read the whole story here: http://intentionalfeeding.blogspot.com/2014/07/crunching-numbers-win-rates.html
What I found was that farm correlates most positively with Win %, and KDA correlates almost as much as farm.
Let me know your opinions, questions, angry comments, etc.