I analyze my tournament games with my chess coach on the computer and in the end I have a pgn file which contains all the variations, mistakes, and tactics i missed. The question is am i suppose to play the game all over again on the board trying all the variations etc, or just studying them on the computer screen enough?
It is always a good habit of moving pieces over the board . The best players of the World practice it and they recommend it also . Sometimes I can understand that in a busy world you may do not have the time & space to occupy yourself and arrange the pieces OTB .
As far as games are concerned it depends whether you are learning a specific opening or a variation then it always useful that you practice your winning games or the games where you have learnt a lot . GM Igor Smirnov says that before you play a tournament you must have a look into your own games where you had a good win . It not only exuberates your confidence but gives you a string determination to win the Game . Kasparov also says that you must memorize some of the games related to a specific opening which you play . It helps you to remember patterns and best moves in a particular opening .
I agree that moving the physical pieces impresses upon the mind more thoroughly than by electronic means. We think and conceptualize in 3 dimensions, so it is only beneficial you should learn and study in the same manner. I know I get much more out of reviewing master games on a physical set than pumping through them on my computer in Chessbase. It takes longer, but it's not a race.
The wrong thing would be to think you learned everything from that one session and file the analysis away. Come back and review it frequently and you will make more discoveries.
Once you have thoroughly analyzed your game, it completely depends on your preferences.
What I usually do is first analyze it on my own on the chess board, trying to come up with the critical positions and evaluating the position on my own. I then use a computer to check the analysis, specifically if the variations I thought as possible replies or alternative options were sound. Finally, I play the game again over the board and share my analysis with my coach.
While the very last option is not indispensable since you have already done the work previously, I found it useful: exposing one's ideas help to clarify them and the critical evaluation of the coach and the rest of the students in the class usually brings up alternative interesting variations.
Number one, you should ask your coach if you haven't :).
Secondly, I think it is a really good idea to go over the pgns if you approach it as if it were an OTB game.
I say load the PGN and pretend you're playing one side as if it were a long OTB game. Then after moves have been made, read the annotations and back it up with engine analysis. I think this is good enough. You don't have to memorize anything unless it is an opening trap or an opening line. Otherwise, if it is deep in the middlegame, then you don't have to memorize unless u want to build a game database and store it in your head, studied, all by memory, but thats debatable and probably not super helpful unless you've reached the position a lot and you need a recipe. Anyway thats a topic for a different thread.
Yeah just read through it, study the chess board, play it like OTB, and most importantly take it slowly.