Chess

[darrelljs]

Neat game Dark Saveant .
Coeurva, that game felt very normal to me until 11. Qh7, then I was , followed by  , followed by .  I rewound to 11 and was able to play it out in my head, the amazing thing is black never really had a choice.

Darrell

[darrelljs]

The amazing thing is that wasn't even the game of the year:

http://www.chessgames.com/perl/chessgame?gid=1094915

Darrell

[Dark Savant]

Reminded me of this game: http://www.chessgames.com/perl/chessgame?gid=1259009

The way I learned that game was "oh, that's the double-check game".  It did mention the uncertainty to the last move, though I think the author preferred castling because castling to checkmate is hilarious.

There's the "long castling" "problem" in Wikipedia's joke chess problems article, the problem by Tim Krabbé.

Tim Krabbé's also wrote Chess Curiosities, worth a read if you haven't.  (There's plenty of material to archive binge on there.)

[darrelljs]

According to the article I read the guy considered castling but was an engineer and found it inefficient to move two pieces when one would suffice. If only he'd been an artist!

Darrell

[novice]

I was worried on Kasparov's behalf that he would get crushed and embarassed. He played the first day safe and managed three draws though. A good effort.

Holding that endgame against Nakamura with little time on the clock was impressive, even if Nakamura made a (in hindsight pretty obvious) mistake in trying to queen the h pawn instead of the a pawn.

Also I have to wonder if the organizers lured in Kasparov by picking a subpar set of wild cards for the tournament.

[darrelljs]

Six months ago I had to give my son a Queen, Rook, Bishop and Knight for him to have a shot.  Now its down to just a Queen.  Once he starts thinking more than one move ahead, he might end up decent .

Darrell

[Gustaran]

Google's AlphaZero Destroys Stockfish In 100-Game Match

This might be the start of a new era in (computer) chess: Google's machine learning AI takes on chess:

https://www.chess.com/news/view/google-s...game-match

[MJW (ya that one)]

Google's AlphaZero Destroys Stockfish In 100-Game Match

This might be the start of a new era in (computer) chess: Google's machine learning AI takes on chess:

https://www.chess.com/news/view/google-s...game-match

There's a bit of an * because AlphaZero got to play with opening's book will stockfish didn't and it was hyperbullet. I'm pretty sure they could have bumped it up to 24-hours and that would solve the problem.

[Gustaran]

That does not seem to be correct:

Since AlphaZero did not benefit from any chess knowledge, which means no games or opening theory, it also means it had to discover opening theory on its own

Source: https://en.chessbase.com/post/the-future...arns-chess

Interestingly, the article shows which openings AlphaZero played first and abandoned later in the learning process. As a matter of fact, the chessbase article as well as the academic paper seem to talk about 24 hours of deep learning, not 4.

https://arxiv.org/pdf/1712.01815.pdf

[darrelljs]

Very interesting article...would love to see it take on Deep Blue at some appropriate point in time. And, of course, Magnus .

Darrell