[{quoted}](name=Very Toxic Jinx,realm=NA,application-id=3ErqAdtq,discussion-id=jBWTfAuz,comment-id=0002000000000000,timestamp=2015-09-10T14:39:38.134+0000)
You also don't main ADC and know nothing about their balance.
Have you seen the carry co-effcient chart? This guy on reddit did a statistical analysis of all champions in the game for the 5.16-5.17 patches.
If you take a look at this chart he provided, the first (bot lane) ADC on the list is rated 107th. Out of 126. Statistically speaking, Vayne and Kalista have about the same chance of carrying a game as Rammus and Olaf. Really. That's how bad this game is for ADCs right now.
http://i.imgur.com/Pg0c2s9.png
While I am definitely not the master at statistics, I did notice something that the mathematician might perhaps forgot to include:
The pick rates.
Unless I missed something from my skim reading (which is entirely possible), I don't think pick rates were actually part of the calculation as it seems to be more of a study on averages.
When there are not that many ADCs in the first place, and they have to be in every single game, I am pretty sure a lower carry coefficient is a given and does not necessarily mean they are very underpowered.
Less seen champs are more likely to carry more because they have less games to carry, thus each game weighs more and so they would mostly end up either in the shiet zone or somewhere up there (in terms of carry coefficient ofc). I think it works like that.
If anything I only see supports to be disgustingly weak from that analysis but it's just me.
**EDIT: **Not to mention he seems to have selected his evidences in such a way that all non-meta games were completely disregarded, games were taken from solo queue where ADCs are meant to be less effective and matches were only recorded from NA which is not a good representation of the actual game as there are many other regions too. Let's just say it wasn't a good start. I can say he went in already with a confirmation bias but I guess I wouldn't accuse him of such a thing.