It's a valid point, this is post-hoc data analysis so you need to be careful of the Texas Sharpshooter Fallacy: https://en.wikipedia.org/wiki/Texas_sharpshooter_fallacy