|
|
|
The
Kramnik vs. Deep Fritz Match
|
|
June 2003: Just a quick note, I have received dozens of e-mails about this. Many of the links - as they relate to this match - ARE NO LONGER VALID!!! I have left them here for a great number of reasons, the main one being this site is a RECORD OF HISTORY ... I went to great pains to record the events of this epic match ... AS THEY HAPPENED! I am NOT going to remove or update these links ... they are a part of the history record of this match. (I WISH people would maintain their links, but I don't control this!) Sorry. |
Kramnik at the chess board.
For
the many of you who get the magazine,
'CHESS LIFE',
(The
official publication of the U.S.
Chess Federation.);
in the February, 2003 issue - there is a nice article about the match.
(Starting on page # 23.)
(Click on the link(s) below, to be taken to the page for that game.)
Game # 1 Game
# 2 Game # 3
Game # 4
Game # 5
Game
# 6 Game # 7
Game # 8
Game
# 1: October 4th, 2002. Fritz is White. Kramnik trots out his
favorite
Berlin Defense to the Ruy Lopez. Fritz attacks, but makes zero progress.
In
the end, White will be a pawn up, (on the K-side); but it is a totally
useless
advantage. The King-plus-Pawn ending is completely drawn.
Score: One draw each. (1/2 - 1/2)
Game
# 2: October 6th, 2002. GM Kramnik is White and he plays his
favorite
move, 1.d4. The computer responds with a Q.G.A. They trade Queens. It is
starting
to look like a dead-drawn match - I am falling asleep here. Then Kramnik makes
an
excellent combination. He plays very strongly and appears all set to win material.
There seems to be no defense, Kramnik's pieces completely dominate the board.
I
am happily predicting an easy win for White. But the computer - blast its chips -
finds a defense, and we head into another seemingly drawn Rook-plus-Pawn(s)
ending. The computer plays the defense poorly, Kramnik plays his usual excellent
technique, and White notches the victory. The human draws first blood in a VERY
complex
game.
Score: A win and a draw for Kramnik. (1.5 - 0.5)
Game
# 3: October, 8th, 2002. Fritz is White and plays a Scotch Game.
The
computer plays this line as poorly as possible. The machine is in a horrible
bind
by move 20 - something that should NOT happen to any program ...
ESPECIALLY
WITH THE WHITE PIECES!!! Then - a miracle. The computer, for all intents
and
purposes, blunders. It plays Nf6, leaving a pawn that Kramnik scoops up
and
captures in just a few moves. After winning this vital pawn, Kramnik is a
juggernaut.
He shows us why he is one of the best players in the whole world, and
plays
flawless chess (technique) in the remainder of the game to garner the full point.
Score: Two wins to none, with 1 draw. (2.5 - 0.5)
Game
# 4: October, 10th, 2002. Kramnik is White, and again plays
1.d4. The
machine trots out, "The Tarrasch Defense," a line that is considered
by some to
be overly risky. But the team has prepared the box well. Despite incredibly
accurate
and sharp play by Kramnik, the machine defends. But the box seemingly is
only
hanging on by the skin of its proverbial teeth. Kramnik wins a pawn, and I
(of course!) predict yet another victory for the humans. But the wily opponent is
not
so easily put down, the program has landed a Rook on the seventh rank.
After
White's King becomes permanently shut off from escape from his first row,
a draw is
assured.
Score: Two wins to none, with two draws. (3 -1)
***
Halfway
Mark: At this point, I considered the match all but over. I was
telling
everyone on close to a dozen Internet Chess Servers that Kramnik would win
maybe one more, as the programmers would adjust the box so that it would
not
be shellacked. Since Kramnik was never in any real danger, I felt certain
this
same scenario would continue. But a rude shock was in store for our glorious
leader of the human side of this grand effort.
***
Game
# 5: October 13th, 2002. After a two-day break, the fighting
resumes.
Fritz is White, and opens with the Queen's Pawn. Play heads into the very
classical
"Queen's Gambit Declined." (Q.G.D.) Kramnik is one of the best players
in the
world on either side of this variation. White plays soundly, Kramnik offers
the
Lasker's Defence which features an early ...Ne4 by Black. Black appears to
be
in some trouble, yet by move twenty-one, (21.Rxc3 '='); I would have to say
that
Kramnik has equalized completely. But somehow Kramnik allows White to
creep
back in and gain a slight advantage. Now Black is defending in a very
complex
ending with the Queen's still on the board. To add to the human's troubles, he
is
growing short of thinking time as well. On his 34th move, if Kramnik plays
...Nc4;
he may well be on his way to drawing the game. Instead he plays a blunder.
The
machine is of course merciless, immediately seizing upon the tactic.
Kramnik
resigns without bothering to go any further, the game is hopeless.
Score: Two win to one, with 2 draws. (3 - 2)
Game
# 6: October 15th, 2002. First, I want the whole world to
know that I have
declared the human side to be the winner, because the dirty scum that
programs
the box ... CHEATED. It's true! A violation of the LAWS of
chess! (Click here.)
(CNN also discussed this matter in a side-bar news story.)
Aside
from the hugely ILLEGAL tactics employed by the Fritz team, the game
proceeded as follows: Kramnik was White. Black essays the Queen's Indian
Defense, and then follows up with Nimzovich's own move, 4...Ba6. (I
remember
studying these lines several times during the 1980's.) This move by Black
leads
to a hugely complicated game, probably just the sort that the program
needed.
White
developed VERY strongly, perhaps even reaching a won game before
move 20. But on his 19th move, perhaps distracted by the machine's
constant
banter, Kramnik decides to sacrifice a Knight. (The move is not sound.) If
he
had won, "This would have been the most beautiful game of my career,"
said
the Kramnik - who may be the best human chess player in the world. But the
machine - of course - found the best continuation, defending perhaps
better
and more precisely than Petrosian or Korchnoi ever were capable of.
In
the end, Kramnik may have even missed the best way to continue the attack.
But I do have to hand it to Kramnik in several areas: # 1.) It took a
tremendous
amount of courage to play this sacrifice; # 2.) He certainly was not able
to
calculate the sacrifice to a logical conclusion; # 3.) If he wins, he is the
hero of
the entire human race; # 4.) He could count on Fritz to find the most
deeply
hidden and difficult of tactical possibilities.
In
the end, Kramnik - seeing no way to prolong the fight - resigned. And the
real kicker? The final position may have been drawn. (Click HERE.)
Score: Two wins apiece, and 2 draws as well. (3 - 3)
Game
# 7: October 17th, 2002. The speakers around Kramnik's chair
have
been quietly removed. Fritz is White, once again we see a Q-side opening.
Kramnik employs a bit of strategy if his own. He uses the same opening as
the machine used against him, "The Queen's Indian Defense." To me,
the
machine's method of handling the opening is not all that convincing, it
allows
Kramnik to attack the center, virtually forcing the center to be closed.
Now
the team of programmers are genuinely concerned, they are afraid that Black
-
given enough time - will slowly maneuver and launch a decisive
counter-attack.
But just when it looks like Black has something cooking, the box finds f4! After
this smart move, Black has no inroads into White's position. So the game
is
drawn after just a little more shadow-boxing.
Score: Two wins apiece, and 3 draws as well. (3.5 - 3.5)
Game
# 8: October 19th, 2002. Once again, Kramnik relies on his old
stand-by,
by pushing his QP forward two squares on the first move. And after a couple
of
feints, we again find ourselves in a classical game - "The Queen's Gambit
Declined."
This time Fritz uses the freeing maneuver pioneered by Capablanca, ...Nd5. It
is
not clear if Kramnik failed to find the best line, or if the computer simply
found the
most accurate defense. In the end, White had a tiny, tiny advantage; but
probably
not the kind that you could do anything with. Kramnik offers a draw ... that
is
accepted without hesitation by the Fritz team.
Score: Two wins apiece, and 4 draws as well. (4 - 4)
(Below are some comments and things I recorded, after the match had already begun.)
This
program, Deep Fritz was/is generally conceded to be much better than the
program, (Deep Blue) which defeated Garry Kasparov in the late 1990's.
(An earlier
version of Deep Fritz defeated Deep Blue.)
While not running
on as a near intelligent or powerful computer, the program is
simply far more
sophisticated than the simple brute-force program that ran on
an IBM super-computer. Running
on a Compaq machine that features 8 Xenon
processors, early tests had shown this
program to be virtually unbeatable in blitz,
and also a real GM-stumper. But none of
their tests were against a player of
GM Vladimir Kramnik's caliber!!
Thus far, (Oct
12th 2002); the computer has played terrible.
Kramnik has won two games
and
the computer has played somewhat feebly in the other two games. (Both
draws.)
Super-K
leads the
match, (3-1) and looks set to win this match in a cake-walk. If
play continues in this fashion, the match will not even be interesting!!
HOLD THE PHONE!! Kramnik loses????
SHOCKER!
Iceman Kramnik loses game 5
13.10.2002 "In
its best game of the match Fritz had world champion Vladimir Kramnik
under
considerable pressure. With 15 minutes on his clock to make six more moves,
Kramnik tried to
avoid a long, tedious and possibly lost endgame – and played the
worst blunder
of his career to
lose a piece and resigned almost immediately."
More
Click HERE to see the web page for game # 5. (Includes a game score.)
Ok,
the box has actually won a game. Big deal. Kramnik's opening was too passive
and wussified. Can Deep Fritz make a match out of this thing? Or will Kramnik get
back on track - and wipe this stupid program off the face of hyper-space?
Time will
tell.
15.10.2002
Well, crud. The stupid box manages to win again. ChessBase has
not written its report, I am writing this less than an hour after the conclusion
of the
game. (I had 2 Internet chess lessons this morning, and lost track of the time.)
By
the time I logged on to Chess.net
to view the game, (they are relaying them, as
many servers are); it was practically over. The box was playing super-tactical
chess -
just what the managers of the program wanted. Kramnik resigned shortly
thereafter.
To go my page for Game # 6, click here.
Game seven of the Man vs. Machine event in Bahrain ended in a draw after Vladimir Kramnik was unable to capitalize on a typical anti-computer position. Fritz in fact threatened to turn tables on the world champion, who ended up defending for the draw. The score is now 3.5:3.5. A full report and pictures will follow soon.
(From the ChessBase web site.)
Fritz
Defends to Draw Game 8 and the Match!
Final Score:
4 - 4
The Brains in Bahrain Man-Machine match is over and has
finished in a 4-4 draw, with two wins each and four draws! Game 8 was the
shortest game of the match, a 21-move draw. The World Champion was unable to
make any progress against Fritz's solid defense. It was a tremendous result for
Fritz, particularly after starting out with only half a point from the first
three games. Both sides said they would be happy to play a rematch. Read
more.
(From the ChessBase web
site.)
Well,
the match is over. I personally am a little bit disappointed by the
result.
(I wanted to see the stupid thing squashed like a bug!) But it could have
been worse - Fritz could have won!
(Shudder, perish the thought.)
In
the end, this match was good for chess. This match was a top news story
virtually
every day that it was
played, the parent web site took close to 75,000 hits a
day. In the end, both sides should be happy. Kramnik
did not lose, and Fritz {and the parent company} is certain
to sell many more programs as a result of this match.
Thursday; November 21st, 2002.
As you can plainly see, I have FINALLY finished annotating all
of these games.
(From my ChessBase database, "Games for My Web-sites.")
(Now I can get down to brass tacks, and start cranking out all the pages.)
***
I have finally finished.
(03 Dec. 2002)
All the games are annotated for your enjoyment!
I need your help. Click here
to find out how you can keep this
a web site that everyone can enjoy.
Click HERE to see game # 1 of this historic match.
This page was last updated on 06/11/07 .
Copyright (c) A.J. Goldsby I
Copyright (©) A.J. Goldsby, 2002 - 2005 & 2006.
Copyright (©) A.J. Goldsby, 2007. All rights reserved.
