The confirmation bias I most often see on the board is in game threads when posters only look for the good for player's they like and only look for the bad for player's they dislike. In reality all players do a mix of good or bad in every game, but for the player they like they focus on the 40% of the shots they make vs the player they dislike they focus on the 60% of the shots that they miss (as an example).
As far as statistics, I think it's important to differentiate between black box statistics like RAPTOR, LEBRON, etc. and statistics based on known calculations. If the statistics are based on known calculations, it should be possible to understand the context in which they support/refute an argument for or against an opinion about a player. Whether a person likes or dislikes a player is obviously subjective, but there are so many ways to objectively evaluate players with stats based on known calculations, that we don't need to cry confirmation bias every time statistics are used.