Castling early is one I've learned is no good. Having the option to castle - waiting until it really foils the opponent's plan of attack - is much better. My games often feature tactical, late game castling.
The main issue is that most folks steer beginners towards "safe" play, but in chess there really is no such thing. The best defense is a fast and efficient attack. Nothing is worse than waiting for a better, more experienced player to pick apart your position. The only thing you might learn from it is how to pick apart someone else's position later, and if you don't play attacking chess, you'll never have the opportunity to apply that lesson.
Tactics and endgame analysis are the most instructive, but you must study your games as well and learn openings! Maybe even in that order, but do what works for you.