Regarding the pick rate / sample size of mages bot lane
When discussing mages bot lane and the statistics associated with them, many people point out how mages bot lane have a very low pick rate. In other words, they have a very low sample size, implying that over a larger number of games we would see many of these very high win rates 'return to the mean', or, return to their 'true win rate' as I will call it.
It is absolutely true that a lower sample size means a higher variability. Another way of putting this is that some mages are likely statistically overperforming compared to their 'true win rate' and other mages are likely underperforming compared to their 'true win rate'. So, logically, in addition to the mages with abnormally high win rates, we should also see some mages with low pick rates that have abnormally low win rates.
This, however, is not borne out by the statistics. If you look at the lowest win rate champions in the bot lane, there are zero mages (outside of Mel, which was obviously nerfed into the ground). I think this is very interesting and far more concerning regarding the state of bot lane. It seems quite evident that this is not simply a case of a lack of sample size. If all of these "outliers" are overperforming, they might not really be outliers. Obviously, this is not a full statistical analysis of any sort. I just found it interesting.
TLDR: mages r cringe frfr