[computer-go] New version of Crazy Stone

[Chrilly]

>
> By the way,  what do you consider a "blunder"?
The "Hydra twin towers" show up on the screen. To be more precise, the eval 
jumps at least by 1 pawn. There can be of course blunders which are to 
complicated to be detected by Hydra. And there can be brilliant moves which 
make the Hydra eval jump for the wrong reason. But so far the 1 pawn jump 
was sufficient to win the game.
Blunders are highly correlated. The mean-distance to the next blunder is 2 
moves, the maximum is 5 moves. Shortly after the first jump (and the 
corresponding bar on the ChessBase-GUI) the next higher bar is showing up. 
The Hydra teams calls this the " Hydra twin-towers".

I have a regular column in the German chess magazine "Kaissiber". One column 
deals how God would play chess. The thesis is: He plays much better - harder 
for the human - than a 32-stone database. He would set up traps. I concluded 
this from the story of Adam and Eve. The tree in the middle of paradise with 
the forbidden apple was clearly a trap. If Eva had the knowledge of God, she 
would have never eaten the apple and hence God would have never tested her 
in this way.
There is the story, that Steinitz wanted to play a pawn-handicap match 
against God (with Steinitz having one pawn less!). I think God accepted this 
offer, but due to his bad experience at his last visit he insisted on a 
home-match. Steinitz soon afterwards started the journey to Gods tournament 
hall.
According to the Vienneses chess-historican Michael Ehn there is no 
historical evidence for the Steinitz proposal. Its one of the many myths in 
chess-history.

Chrilly