Yes pick rates are definitely important to consider when trying to interpret win rates. The win rates I’m looking at are from un-mirrored competitive ladder stats (specifically for PC players on patch 1.43 at the GM+ MMR range), meaning games where the hero is not being actively played on both teams.
We don’t balance solely off of stats anyway but they help inform us when something could be really out of whack or can give us a slightly more unbiased lens when trying to identify the root causes of gameplay issues.
Stats don’t do a great job of telling us subjective things like how fun something is, player feedback does. We can try to infer a little bit of that from pick rate but there are too many competing factors.
The most common feedback I’ve been reading recently is that Baptiste and Mei are very overpowered (there are some upcoming changes to these two as well) and Genji/Soldier are clearly the weakest heroes.
When I looked into the current patch stats, Mei and Genji are both very middle of the pack for GM where I would normally expect players to adhere more closely to the meta.
There’s only about 8 heroes with higher pick rates than Mei/Genji. Soldier pick rate is low, coming in at slightly less than half of Genji’s and is on par with Ashe for both pick/win rate.
One of the perceived strongest and one of the perceived weakest heroes have nearly identical pick rate and win rate. I found this personally very interesting and wanted to share.
My post above was intending to get people to consider that this data, while not absolute or conclusive of anything on its own, may suggest off-meta heroes are much more viable on the competitive ladder than players tend to expect, even at the highest levels. Professional play is likely another story.