London Chess Classic 2017 Underway
More than half the tournament passed in London and there is an outcry in the public on the number of draws. Only two decisive results from 25 games, incidentally, both these games were won by Caruana.
People are complaining, the talks of the “drawing death of chess” is immediately back, the usual suspects are pushing their ideas of abandoning classical chess and moving onto rapid and blitz.
There are 23 drawn games out of 25. And my question is: so what?
I am not even going into the arguments that chess is basically a drawish game, that much we all know. What I would like to point out is that another tournament with the same people at another point in time may as well have more than 50% decisive games. These people are trying the best they can at the given circumstances and sometimes it just doesn’t work out.
They are the best players in the world, they all want to beat each other, they try their best, but more often than not they fail because the other player is doing the same! There is an infinite number of factors that influence these things, current form, physical condition, opening preparation, state of mind and also plain luck. The bottom line is that simply there are tournaments like this and we have to accept that fact. If the games are well-fought and you can see the players trying hard, there is nothing more we can ask of them.
Speaking of the death of chess, a ground-breaking Alpha Zero program crushed Stockfish 8 in a 100-game match, winning 28 games and drawing the rest. There are certain moot points here, like the strength of the hardware the engines were using (incomparably stronger for Alpha Zero), the time control of 1 minute per move and the openings used by Stockfish (and no opening book for it), but that is all beside the point. The main point is that Alpha Zero was only taught the rules of the game and then was left alone to learn the game by itself. It did it by using the Monte Carlo simulation, i.e. playing a mind-bogglingly huge number of games with itself and learning along the way. In a matter of hours (some say 4h, some say 24h, any way equally impressive) it reached a level good enough to annihilate one of the best engines in the world.
This is an actual Artificial Intelligence, capable of learning by itself and dominating such a complex game like chess. The fact that it managed to do it in such a short amount of time makes it even more incredible.
To wrap this up, I offer two excerpts. One of the best humans playing chess and the other of the best computers doing the same. Judge them yourself.
For the computer game I will only comment with exclamation marks to show my amazement at the moves. Enjoy and learn if you can.