cartixuzi [#34]
This vid feels flawed. Impact of kills a player has isn’t an individual metric but more of a team value thing, and because of that I feel like this vid and its analysis leans towards on which players carried their teams the hardest rather than finding who’re the best players overall. Like when a player gets an impact kill, that kill itself is only impactful for their team, hence why it’s just a metric on who’s putting their team on their back.
A better way to explain what I’m saying is cned’s impact stat. The guy was carrying his team this year, but if he had someone like Less who’s also got great impact to rely on then his impact kills would obviously go down as a result, but that doesn’t mean he’s suddenly a worse player right? This shouldn’t prove anything about one’s individual raw performance when being compared to other great players in OTHER teams (the comment below me does an even better job at explaining the comparison between player impact across diff teams).
What he should’ve done to complete the actual question is to find the most impactful player of each top team from every year (or just the year you want) and then compare their raw individual stats together like ACS, KDA, clutch rate, etc and whatnot to get a closer image of who’s really the GOAT (in terms of peak of course).
Like if Jinggg turns out to be the best player this year over Alfa, Demon1, and aspas, then something must’ve been overlooked lol. Also this guy reallllly loves yay.
you are correct that, due to the nature of all stat based analysis, the results can be misleading
this analysis was very kill and trade focused so it favored duelists
and ofc it cant properly represent the complexity of the game; even when ignoring your point about the team not properly being considered (since its just gonna be impossible to separate a player from their team) it has huge issues
economy impact of kills, how much support util you were receiving during the kill (how good was it?), how many people were on the site, etc would all shift the weighting significantly - it just leads to an endless game of trying to get more aspects of the game into your stats... and in the end theyd probably still be wrong due to the whole team thing
your proposed solution wouldnt lead to a good result either, it would heavily favor players like ZmjjKK who have a risky and inconsistent playstyle, while a more consistent player like for example demon1 would look pretty mediocre
It also has the whole acs & kda already being linked problem that vlrs rating tries to tackle
At the end of the day the stats are great because he presents all his thoughts and reasoning for the stats, and its easy to figure out what aspects he overlooked and then decide how much you value them - no stats will ever beat the "eye test"