It seems odd to have used avatars as the sole metric. I know many male individuals that use female avatars and the vast majority of both males and females I kno

It seems odd to have used avatars as the sole metric. I know many male individuals that use female avatars and the vast majority of both males and females I know do not use any representative avatar. Is there a particular reason you chose to use the avatar instead of gender weighting based off of the posts and questions, which tend to provide more accurate results? While we noted when we tried this that most technical discussion on stack overflow leaned towards masculine even when written by a female, it should still be far more accurate than the avatars.