The Shannon number, 10¹²⁰, is an estimated lower bound on the game-tree complexity of chess. Claude Shannon, in 1950, worked this out as, to all intents and purposes, the total number of possible variations from a starting position to checkmate/resignation in a typical game of chess.

I’d like to draw a parallel between chess variations and biological evolution.

Those of you who have read Richard Dawkins will be familiar with ‘The Museum of All Possible Animals’. In this museum, each possible animal is placed in a box, surrounded by boxes of its close cousins. A line drawn through a series of cubes will show a progressive evolution of a specific characteristic, say, along the ‘beak curl’ diagonal (hardy-har), a series of birds appears in each box, each identical to the last save a progressively changing angle of beak curl. The dimensionally savvy amongst you will already note that the number of dimensions through which you can move in the museum, i.e. the number of variable characteristics from a specific species is infinite (here we go again!).

I believe there is a parallel ‘Museum of All Possible Chess Variations’, with precisely the same rules as the animal museum above (with the exception that there are only about 30 possible legal moves in any given position; thus the museum would be ~30-dimensional). In the museum reside all the games that it would ever be possible to play.

I propose that both systems are subject to Natural Selection. Imagine the boxes in the museum are transparent. Each of the fittest animals, and each of the strongest variations, resides in a box that is bright yellow, with progressively less bright boxes until no colour at all is in those boxes which have no hope of survival. Thus, looking through the museum from a distance reveals a glowing spider-web tracing through the boxes in a complex but orderly pattern (assuming one can see anything in 30/Infinite dimensions). In a sufficiently large span of time (the handful of hundreds of years chess has been around are woefully insufficient) the game of chess would satisfy natural selection:

1. They reproduce. This is seen in the repeating of play in opening theory, where the strongest moves survive in multiple games.
2. They are subject to variation upon reproduction. Thus a strong novelty found by a GM is copied in many further games.
3. They are subject to Natural Selection. Thus variations that have been discredited cease to be played.

Eventually, at a far, far point in the future, a (albeit very, very large) set of the strongest games will exist.

Is chess subject to Evolution by Natural Selection?

ED.

Interesting hypothesis. I look forward to people's responses.

I find something different. According to Darwin, changes happen by chance and those of them which may give any kind of an adventage to living beings tend to perpetuate. In chess, masters look for the best moves.

@ ED, interesting thought as usual.

My 2 cents... Given that Chess is convergent in a Game theoritical sense and the number of possible chess games being finite ~10^120 (choosing the best of moves for a given position will lead to capturing and hence eventually reducing the number of pieces to zero barring the kings which proves the number of possible chess games cant be infinite), Yes at some distant point in time there definitely will exist a set of strongest games containing the best sequence of moves made by both sides for a given position!

But it is a question if that stage will be reached with the help of Supercomputers and Quantum Mechanics alone given that not even the total number of electrons in the whole universe is close to 10^120...

Cheers,

Over the millenia we have swicthed quite a few times between hairy elephants (like woolly mammoths), and more bald elephants (like nowadays). There are many examples like this in biological evolution.

'survival of the fittest' refers to the creatures most fit for the particular environment they are stuck with at any particular time; no-one (normal!) suggests that one style of elephant is somehow 'better' than the other style, it's more like a case of having your coat on or having your coat off, depending on the weather.

Just thought I'd note this nuance in biological evolution which is absent in the constant-slope-of-improvement model used in the chess analogy.

I don't want to start an arguement, just blabbering on.