Searching for a winning chess formula - what do you think?? - Chess Forums

TheAnalyst7

Jun 26, 2016

0

#1

I had this idea for solving chess, however it may work or not work but "you miss 100% of the shots you don't take" so I'm going to show you what I was thinking. In math, there are things called rigid motions, and basically I take the coordinate of a plane, and apply a transformation rule to it in order to rotate the plane x degrees about the origin. Example: I want to rotate 3,4 on a coordinate grid 90 degrees counter clockwise. So I use the correct formula -

(x,y) (-y,x)

Plug it in

(-4,3) and this is my new coordinate.

So basically I was thinking of how this would apply to chess. My idea is that I take a given chess position, and with the help of stockfish, I calculate the #1 best move from the engine's suggestion. Then I take the previous move, and the engine move to analyze how I got from a to b. Still confused? I'll elaborate

Blacks last move was Rb7.

So we can use coordinate form. e7, b7 = 5,7|2,7

And for white here, stockfish 7 suggests Be3. In the game, Bd3* was played. But for Be3, we write

5,7|2,7| 7,5|5,3

Finally we draw comparisons. x-2,y,2!! See it? It's in Be3, we see the x coordinate -2 and the y coordinate -2. With blacks last move, we see x-3,y-0. Now just imagine annotating an entire game like this using stockfish, who knows what similarities you'll find!

That's all I have for you, I know I may be overthinking, but it was just an idea that popped into my head. Anyways, thanks for taking the time to read this, have a good day

beyondthemask

Jun 26, 2016

0

#2

Awesome post! I hope you continue your processing on this! P.S. This is way beyond me

Sqod

Jun 27, 2016

0

#3

I don't see that this is useful. In chess there is an inherent difference between ranks and files in that pawns are forced to move mostly along files, and all units on both sides start aligned along ranks. Rotating the board 180 degrees is basically just using descriptive notation, which is nothing more than a (sometimes useful) representation difference, so we know right away that a 180-degree flip has known, limited utility. I would make a similar statement for a reflection, which is also a geometrical transformation, but one that flips the 2-D object through 3-D space instead of rotating it within the plane: a reflection would lead to what I call S-O notation, which is just a (sometimes useful) representation difference.

However, you're making a 90-degree rotation, which doesn't expose any useful invariants in representation (unless they're extremely limited) because it's mixing ranks with files, which are inherently different lines of direction due to chess rules and unit motions, so I don't see it even has any utility as a method of representation. I believe there are a lot of productive, mathematical ways to approach aspects of chess, and I encourage you to look for those, but I don't believe the above approach is one of them.