Historical riddle: could Kramnik have defeated the computer?

by Karsten Müller
5/28/2020 – In November 2006 World Champion Vladimir Kramnik played a match against the computer program Deep Fritz. The venue was the National Art Gallery in Bonn, Germany, the prize fund one million dollars. It was the last big man-vs-machine match. Kramnik lost 2:4 and did not win a single game. But in the very first game he had a favourable endgame. Could he have won? Until today that remains unclear. Can you help GMs Yasser Seirawan and Karsten Müller resolve this question?

ChessBase 18 - Mega package ChessBase 18 - Mega package

Winning starts with what you know
The new version 18 offers completely new possibilities for chess training and analysis: playing style analysis, search for strategic themes, access to 6 billion Lichess games, player preparation by matching Lichess games, download Chess.com games with built-in API, built-in cloud engine and much more.

More...

The duel between the 30-year old World Champion from Russia and Deep Fritz, was set for six games. Vladimir Kramnik had the chance to win one million US-Dollar. Specifically: he could double his starting fee of US $500,000 by defeating the computer.

Press conference before the start of the match Kramnik vs Deep Fritz

The broadcast room of the match venue

The Deep Fritz hardware, a Siemens Fujitsu 1.6 Mhz. core duo machine, was located in an elevator at the back of the stage, which was closed during the games to prevent disturbance to Kramnik because of the cooling fans.

The start of game one, with the first move executed by German Finance Minister Peer Steinbrück, himself a keen chess player. Steinbrück went on to become the chancellor-candidate of the Social Democratic Party in the 2013 federal election. On the right is FIDE President Kirsan Ilyumzhinov.

The press photographers at the start of the game

And at last game one of Kramnik vs Deep Fritz gets under way.

The audience following the games in the National Art Gallery in Bonn

Could Kramnik have won game one?

Now we come to our question where we seek your analytical help. In the first game Kramnik reached a favorable endgame. I believe that a winning proof can be found, but matters are not completely clear.

The second entry on out replay board is the deep analysis by GM Yasser Seirawan from 2016.

You probably know that in our replay boards there are a large number of functions you can use to really appreciate the games. Recently we published a comprehensive tutorial which tells you about all the powerful features and buttons that make the ChessBase's replay one of the best watching experiences around.

One big advantage is that you can start an engine (fan icon) that will help you to analyse. You can get multiple lines of analysis by clicking the + button to the right of the engine analysis window. The "!" key, incidentally, shows you the threat in any position, which is incredibly useful in the case of unclear moves.

There is one more thing you can do. It is a lot of fun, but also a serious challenge: Click on the rook icon below the notation window. This will allow you the play the above position against Fritz, at your level of playing strength (e.g. "Club Player"), right here on the news page. Note that your analysis, in which you can delete, move or promote lines, is stored in the notation as new variations. In the end you will find the game with your analysis in the cloud. So nothing is ever lost.

Grandmasters Dr Helmut Pfleger and Artur Jussupow analysing the endgame

Final standing

 Player
Rating
1
2
3
4
5
6
 Tot. 
 Vladimir Kramnik
2760
½
0
½
½
½
0
2.0
Deep Fritz 10
-
½
1
½
½
½
1
4.0

Links


Karsten Müller is considered to be one of the greatest endgame experts in the world. His books on the endgame - among them "Fundamentals of Chess Endings", co-authored with Frank Lamprecht, that helped to improve Magnus Carlsen's endgame knowledge - and his endgame columns for the ChessCafe website and the ChessBase Magazine helped to establish and to confirm this reputation. Karsten's Fritztrainer DVDs on the endgame are bestsellers. The mathematician with a PhD lives in Hamburg, and for more than 25 years he has been scoring points for the Hamburger Schachklub (HSK) in the Bundesliga.

Discuss

Rules for reader comments

 
 

Not registered yet? Register

albitex albitex 5/30/2020 03:34
@SixCoutChess > Carlsen said in an interview two years ago: "Today none of us GM can win a game against an engine."
Kasparov has repeatedly stated that: "The man-machine challenge ended with his lost match with Deep Blue. Today it no longer makes sense to consider this kind of challenge, as now any decent software beats a GM even on a tablet or smartphone."
David Clarke David Clarke 5/29/2020 08:41
With help from Stockfish 11: A few knight moves reduce Black to complete passivity, freeing White's king to pressure the immobilized b-pawn. White then locks up the kingside pawns, freeing the knight to win the b-pawn and the game. 29.Nd5 centralizing the knight on a light square, attacking the b-pawn and keeping Black's king away from the center. 29...Bd4 30.a4 Kramnik's move throws away the win
30.e3 is the path to victory. The Black bishop is now quite confined and white's kingside is safe from it. 30...Bc5 31.Kf3 b5 to free the bishop and restrict the a-pawn 32.Ke2 e4 33.Nc3 the second key knight move, forcing the pawn forward which further restricts the bishop 33...b4 34.Nd5 the b pawn is immobilized and the bishop tied to its defense 34...Kg6 35.h3 Kg5 36.Nf4 h6 Black is completely passive 37.Kd2 Kf6 38.Kc1 Be7 39.Kc2 Kg5 40.Kb3 Bd6 41.Kc4 improving the king's position
41.Nd5 looks simple and winning but throws the win away 41...h5 42.Kxb4 h4 43.gxh4+ Kxh4 44.a4 Kxh3 45.a5 f4 and Black can exchange his bishop for the a-pawn while the King takes any remaining kingside pawns
41...h5 42.Kd5 Bf8 43.Ng2 stopping ...h4 and any thoughts of Black getting his bishop to e1 43...Bg7 44.Kd6 threatening to go after f7 44...Kg6 which drives the king back from g5 and that allows white to play h4 in a few moves 45.Kc5 Bf8+ 46.Kc6 Bg7 47.Kb5 Bf8 48.h4 denying Black counterplay on the h-file and freeing the knight to win the b pawn. 48...Bd6 49.Nf4+ Kh6 50.Nd5 the position is hopeless for Black. The b pawn is lost giving white an outside passed a-pawn for which the bishop must be sacrificed leaving white a knight to the good. Black has nothing on the kingside. 50...Be5 51.Nxb4 1-0
30...Bc5 played by Fritz in response to Kramnik's 30.a4, Black's bishop remains active with pressure on the kingside that will make it difficult for White's king to activate ½-½
SixCoutChess SixCoutChess 5/29/2020 02:06
That's a fantastic question, but if we take the timeline nearly 14 years behind and synchronising with the present time. I think the result will be a more potent victory to the computer, but the real question is: Which system and which chess program will be the opponent to Vladimir Kramnik at this time? The development of chess program has incredibly fast broken the human brain, but still missing some kind of tactical, strategical, positional and a few more chess features. When we were looking at the overall content from the moment of 2006 and how chess programs working today, I would like to give Magnus Carlsen the same options to get a match with 6 or 8 games. Garry Kasparov was the first grandmaster to challenge the computer system by Deep - and lost! So what will it be, if the present world champion gets an invitation to test Fritz 17, Alpha Zero or another computer program? Right now chess players from the whole world using the system to preparing and analysing all from openings to games. So Kramnik would like to be at the same level as Carlsen are right now, but the computers will still winning a match.
Karsten Müller Karsten Müller 5/29/2020 08:43
Many thanks Kraftwerk! From Guy Haworth:
"Dear Karsten,

Chessbase were of course v much involved: Matthias Feist operated Deep Fritz 10 and Matthias Wüllenweber was at the event.

https://en.chessbase.com/tagged?tag=Kramnik%20vs%20Deep%20Fritz gives all the Chessbase articles on the event - but I have not seen the spec of the hardware yet.

https://en.wikipedia.org/wiki/Human%E2%80%93computer_chess_matches#Kramnik%E2%80%93Deep_Fritz_(2006) does better ... saying the hardware was ...

a computer containing two Intel Xeon CPUs (a Xeon DC 5160 3 GHz processor with a 1333 MHz FSB and a 4 MB L2 cache) and was able to evaluate eight million positions per second.

This is the spec of the processor: https://ark.intel.com/content/www/us/en/ark/products/27219/intel-xeon-processor-5160-4m-cache-3-00-ghz-1333-mhz-fsb.html ... and there were two of these."
c_webber c_webber 5/28/2020 06:23
Black has one more defensive try:

29. Nd5 Bd4 30. e3 Bc5 31. Kf3 b5 32. Ke2 e4 33. Nc3 b4 34. Nd5 Kg6 35. h3 Kg5 36. Nf4 h5 37. Ng2

The Knight is temporarily stuck on g2, but now White's King heads for the f7-square and the win remains clear as a sunny sky.
Kraftwerk Kraftwerk 5/28/2020 05:20
The hardware seems to be mixed up in the article. To the best of my knowledge it was a Dual Intel Xeon 5160 at 3GHz which powered Deep Fritz 10 in this match.
chessbra chessbra 5/28/2020 02:39
Vladimir Kramnik-Comp Deep Junior 1-0 Dortmund SuperGM 2000
1