- your ability to calculate complicated variations, and
- your knowledge of the endgame.
What about novelties? What do you do if you come across a move that isn't in any of your references and looks like it's never been played before? Be suspicious. There are thousands of excellent players who have studied the same lines that you are learning. Does the move lose by force? Does it lead to a game where you'll have a chronic positional weakness and will suffer a long time to achieve a draw? If, after doing your own analysis, you don't see anything wrong with the move, play it against your computer before you risk it in a serious game.
What do you do after deciding on your move? Write it down. Keep a notebook with the moves you've chosen, the moves you've rejected, and the reasons why you decided on a certain move. Even better, keep track of your repertoire using your computer. Use your favorite chess database software to manage the tangle of variations that will arise from your work. This will let you incorporate any key reference games that you have collected in digital format.
When you reach a position and you don't know what to do next, practice that position against your computer. It won't complain if you always play White or always play Black or always play the same move. If you have a friend who's interested in the same openings that you are, then work with your friend. You won't need a nondisclosure agreement until you win your first title!
