ESN - Emanuel's Simple Notation - Represent any chess position in only two words!

[Uri Blass]

In the middle of the night I rose, and starting designing my system. The impressive conclusion is that any chess position can be represented using only two regular English words.

The impressive conclusion is wrong. Even with a million words dictionary, with two words you can map one trillion (10^12) positions. If chess was limited to one trillion positions, the game would have been solved in the last century.

Despite its apparent simplicity, the FEN notation can represent any possible board position, even if illegal, with many kings, and so on. Anyway, the possible legal chess positions are a much bigger number than 10^12 (check for example http://talkchess.com/forum3/viewtopic.php?f=7&t=77685 this trade). Let's say 10^40, very lower bound. "Regular" English words are less than 200.000. Do the math: you need at least 8 words. And to keep the encoding/decoding algorithm simple enough, you should divide the board into chunks, like FENs, and the number of words would be higher. And you need a fixed standard dictionary shared between all applications using this system.

Last but not least, you lose any readability:"family pottery" (unreal example since you need many more words) doesn't say anything, "3R1R2/K3k3/1p1nPb2/pN2P2N/nP1ppp2/4P3/6P1/4Qq1r w" is by far more clear, at least for an experienced reader. And FENs are usually read by experienced readers.

Alex

I totally agree that it is impossible to store positions in this way.

I think it may be more interesting to see a program that teach humans how strong humans can memorize chess positions quickly.

Strong humans may look at a position from games for 3 seconds and in most cases memorize all the position when they cannot do it for a random position not from games.
I read that basically they memorize patterns and they do not memorize every piece as a single piece.

For example they may memorize king g1 rook f1 pawns h2 g2 f2 as one pattern so basically I would like to have a program that get a chess position and show me in 99% of the cases no more than 6 patterns that the compose the position so in order to memorize the position it is enough if I learn to memorize all the patterns when there are no more than 10000 patterns to memorize(I think that memorizing millions of patterns is too much for most humans).