Knights or Bishops


Depends on the position or the prospects for the piece. The question is whether the position allows you to make one piece superior to the other. For example, in a closed or semi-open position with a good outpost in the opponents half, it might be good to keep the knight if it can be made superior to the bishop. On the other hand, in an open position a bishop might be better. Endgames with pawns on both sides favour bishops, pawns only on one side favour knights. Sometimes you can trade pieces to suit the position, other times you must shape the position to suite your pieces. These are rules of thumb with exceptions but you get the idea. One piece is not objectively better than the other, it's about making the most of what you have.
It also depends on what each person is better with. If you are weak against knights, they can really eat your lunch.
As people are saying, depends on situation. You can see on https://en.wikipedia.org/wiki/Chess_piece_relative_value their relative values.
Obviously depends on the position. In general, bishops seem to be preferable. But in closed positions, or ones where many pawns are on the same color as the bishop, knights are obviously preferable. On a personal note, I find a dominating knight on an outpost one of my favorite pieces. Strong bishops, I love a long-range pair, but it takes a lot to set up and is kind of rare.
What about in beginning game? What is the tactical advantage of taking a bishop on F3 VS a Knight on F3? Always wondered why people take the Knight with their Bishops and vice versa so early on.

The Pawn formation in the center usually dictates whether Knights or Bishops will have the edge, as well as which Bishop (of the pair) will be stronger. To grossly over-simplify the situation, Knights like Rams (locked Pawns), Bishops don't.
There can be a number of reasons for pinning and/or capturing a knight on f3 or c3. The knights control central squares. The player may be trying to remove an attacker or defender of a central pawn or square, maybe to support or prevent the next move. Removing a defender can be just as effective as adding an attacker. So one reason is that it is part of fighting for control of the central squares and to support/prevent centralisation. It also develops a piece (bishop) and at the same time puts it to useful work.

It boils down to pawn structure. At least if I'm going into an endgame without rooks and having to choose between trading off knights, bishops, or some combination of those 4 pieces, it always boils down to how I can take advantage of the pawn structure.

generally prefer the knight cause its somewhat more flexible I guess but it really depends on the situation

Knights seems better in endgames... it's annoying when I get stuck in a chains of attacks in a endgame and I lose like 2 or 3 pawns. And knights can attack both colored and uncolored squares, but bishops are scary for me in open space, and it's scary when someone has the bishop pair (especially open) and I like to get to trade one of their bishops. Plus the frustrating knight forks are harder to find in a fast game, while a bishop may not seem as threatening of forks and chains.
FUN FACT : Knights can only go to the opposite color it is standing on, so knights alternate between dark and light squares.