I happen to know a USCF NM and a near-expert who cannot "yet" play blind-fold. I used to think that visualization was a pre-requirement to getting better, though I've been learning that it is merely a means to an end.
Though if you really must attempt to fight genetic factors, here are a few exercises that I had really strong players tell me to do. I'm barely an average player, so I'm going to cite sources for the exercises below.
- * Grab a database and download a bunch of games with forced mate in less than 5 moves. Play it out in your head. Move your way up to 10. => source Master at local club.
- Chunk the board ... i.e. don't see 8x8, see FOUR 4x4 quadrants. It makes it easier for your brain to store patterns...for example the King's Indian fianchetto position ... it is simple and fits into the 4x4 grid. Practice re-constructing a position you just saw (in a book, online) with zero mistakes, gradually giving yourself less time to memorize the position. =>Source : Thought and Choice in Chess : Adrian DeGroot.
- Play REALLY SLOW games and take your maximum alloted time per move. NM Dan Heisman says he suddenly discovered he could play blindfold one day (after he got pretty strong) and attributed it to playing tons of slow games where you let your brain burn in the positions. (Source: Dan Heisman)
- When reading books with a diagram between 5-10 moves, do not fidget around for a chess board to play the moves and variations in between ... see it in your head...no matter how long it takes. (Source: Master at local club)
When I was a youth, I had no trouble visualizing the chessmen moving about the board - for instance, I had no problem calculating all the moves of the famous Lasker vs. Bauer double Bishop sacrifice and firmly retaining the resulting position in my mind (I might have become a strong player if I had stuck to chess instead of partying with my friends, but this is another story).
Now I cannot really even firmly hold the image of a single piece leaving its square and occupying another. Most of my "calculation" is via inference with the aid what hazy visualization I can muster.
This isn't playing chess, this is playing at chess.
Can anyone suggest what really might help?