chess.com has a really cool vision game for picking out squares and moves by name: https://www.chess.com/vision

It seems silly but makes reading chess books and analyzing games much easier. The first time I played with the vision tool for 20 minutes I felt way faster at it already.

i assume you have a direct mapping in your mind between squares and colors or actually visualize the board but for those that don't you can convert the column letter to a number then add the two numbers together and depending on whether the number is even or odd the square will be black or white. ie: d1 -> 4 + 1 -> 5 -> white.

there are a bunch of tricks you can do if you use number,number notation but being able to visualize the board is probably better than relying on numeric tricks.

At first I used tricks but the more I practiced the better I got.

Wow, I'd have enough trouble remembering which color I was!

I played once on a board where the pieces were red and blue instead of black and white. I was completely messed up... Funny how the brain works. Or maybe it is just me?

Some people believe that learning with a colored "rainbow chess" set improves retention and pattern recognition. http://home.hu.inter.net/~prochess/rainbow.htm

Sort of like syntax highlighting I suppose.

How long did it take you before you could play a whole game blindfolded?

I created a new account on lichess and started playing blindfold games right away. It depends on how long the games are and how complicated the position gets but anywhere from 13 to 50 moves I'll be okay. I won two games so far.

I wonder if this would work for programming. Closing your eyes, then recording yourself describing an architecture or module.

I noticed improvements in general visualization ability from doing this. I was grinding leetcode during this time and noticed improvements visualizing data structures after I practiced board visualization for a bit.

Measuring skill in programming is much harder than measuring skill in chess. ELO is imperfect but you can be reasonably sure a 2100 will beat a 1200 most of the time. With programming, it's difficult to even make a choice between the person who write a sudoku solver in an afternoon or the person who makes a sturdy website which is secure against all known CVEs (but can't make a sudoku solver, or invert the proverbial binary tree on a whiteboard).