The game of chess as the universe

Every chess player knows the comparisons between the game of chess and war, between the game of chess and ‘life’. But computers show that the game of chess should actually not be compared to real life combat or life itself, but rather to… the universe.

Okay, so Man didn’t make it against Machine. Intuition and creativity fail more often and often, although sometimes it’s still close, against the computer’s brute calculation force. As I argued before, this can be explained by the simple fact that chess, in the end, is a matter of calculation, and computers are simply better at this than humans. (By the way, the last match game Deep Fritz-Kramnik was a very nice demonstration of what I meant could happen if a Sicilian came on the board. Or did you understand anything of Fritz’ tarnishing way of treating the opening - succesfully?) Humans are good at intuitively finding relations, at fast recognition of approximately, but not exactly the same patterns. What does this mean for the future of chess?

The possibilities are fascination, but also a little frightening. By tearing down strategic rules, intuitive pattern recognition, opening theory and knowledge of theoretical endgames, the computer shows us a view into the inmeasureable abbys en complexity of chess.

Jonathan Rowson, in his already classical book Chess for Zebras, speculates about what computers can show us in the long term:

“If computers ever manage to ’solve’ the game, we would know the fact of the matter about which moves tilt the balance between a draw and a win. We might also discover, for instance, that 1.d4 leads to a draw in most variations, while 1.e4 wins in every line except the Najdorf! It is also possible, and personally I think it is more likely, that chess would look radically different, perhaps even unimaginably different. By this I mean that something that seems utterly ridiculous now, like claiming 1.a4? is losing for White while 1.a3!? is a draw and 1.b3 is winning, might turn out to be true, but for reasons that are beyond us now.”

By the way, this ‘uninmaginable difference’ is currently shown mainly by computers that work on calculating theoretical endgames, so called tablebases. How completely counter-intuitive these endgames are, if played perfectly, is described strikingly by chess computer expert Tim Krabbé: “A grandmaster wouldn’t be better at these endgames than someone who had learned chess yesterday. It’s a sort of chess that has nothing to do with chess, a chess that we could never have imagined without computers. The (…) moves are awesome, almost scary, because you know they are the truth, God’s Algorithm - it’s like being revealed the Meaning of Life, but you don’t understand one word.”

Recently, I watched a lecture of the Brittish scientist Richard Dawkins, especially known for his book The Selfish Gene. Dawkins starts his lecture with two quotes. The first is by biologist J.B.S. Haldane: “Now, my own suspicion is that the universe is not only queerer than we suppose, but queerer than we can suppose.” The second quote is by Richard Feynman, and is about quantum mechanics: “If you think you understand quantum theory… you don’t understand quantum theory.”

Striking echos of the quotes by Rowson and Krabbé! Dawkins explains in his lecture that the human brain evolved to handle the reality of every day where people had to survice - just like the bat brain evolved to handle the reality of every day where bats had to survive. Human brains perceive stones as solid things, although they consist of empty space for the most part. Human intuition evolved to deal with ‘middle size’ objects surrounding humans - stones, trees, other animals - but not to handle extremely small things, like separate atoms or ‘strings‘, and not extremely large things, like solar systems or black holes.

I suspect it’s also like that with chess. Human chess intuition works as long as we don’t have to deal with extremely large or extremely small problems. The large problems – the tablebase endings – can only be ‘understood’ by computers, but we humans also have problems handling small problems. Which chess analyst doesn’t know the positions where you just feel that White must have a small advantage, which you’re unable to prove when you start analysing the position extensively (with a computer)? The small advantage must be there, you can smell it, but when you actually start looking, it eludes you time and time again. Our apparent certainties, our beacons, stop functioning. ‘Our’ chess theory (not only openings, but also strategy) is, I think, not about Chess (with a captical C) itself, but about the likelihood with which certain moves are followed by other moves, and ultimately about the behaviour of chess players, rather than the formal properties of the game. We were able to imprison the game for centuries, but the computer forces us to face the fact that the game is starting to elude us. More and more often you see analysis with commentaries like: “the computer says the position is equal here, and perhaps it is right, but from a human perspective, White is better.”

And so perhaps the analogy of Dawkin’s ‘middle size world’ can teach us something about our chess understanding and the way we used to view chess in the past. Chess is not an analogy for life - that’s exactly what it’s not! In life, we are dealing with ‘middle size’ things, and in chess we increasingly have to deal with non-middle size objects like tablebases and stereotyped evaluations that turn out to be wrong. Chess is, rather, an analogy for the universe. We take brave attempts to understand it, and even think we can battle it. But if what chess computers show us now is only the tip of the iceberg, we will not only lose the battle, but also perhaps never understand why we lost it.

This is no reason for pessimism, perhaps only for a different frame of reference.