LCC Home
TWIC Home
Chess Shop
Chess Express
Kingpin
Book archive
Book Reviews
Dvoretsky Sale
£5/$7.50 Sale Chessbase8
Downloads
Bridge
Go
Backgammon
LCC Links
Fantasy Chess
Events

EMAIL TWIC
EMAIL LCC



Brains in Bahrain Game 2


The million dollar Brains in Bahrain match between Vladimir Kramnik and Deep Fritz takes place 2nd-22nd October 2002 (Opening Ceremony 2nd October. Playing days 4, 6, 8, 10, 13, 15, 17 & 19 October, 2002).

Report Seven: Game 2 Report
by Malcolm Pein

Security

There was a real air of expectancy at the Bahrain Mindsports Centre as Kramnik arrived with his entourage ten minutes before the start of play. I was doing the rounds upstairs and checking the rest rooms and toilets for anything that remotely looked like it should not be there like a Pocket Fritz for example. Security is always an important consideration when you have a big money match. The night before the arbiter watched as the Fritz team tinkered with the opening book as described in the rules.

Aziz. Photo Mig Greengard

For his personal security Vladimir has brought his own Arabic-speaking bodyguard with him. ‘Aziz’ is from Beirut and he certainly looks like he can handle himself.

The one problem I found was that Frederic Friedel of ChessBase had grabbed the big wooden boards we had erected around the computer to stop anyone from tampering with it. Frederic had another use in mind for it. He used them to board up to of the windows in the web room to prevent Mig and the Fritz team getting too hot from the searing Bahrain sun. The room is where the live moves come from and the commentary which is produced in German Spanish, by the journalistic legend Leonxto Garcia and English by Mig and the other commentators.

Blinded by the light

Perhaps it is just as well I am leaving. This Match Director role might be going to my head. After Kasparov’s team negotiated so badly with IBM I saw the result first hand and I am trying to make sure that the match is balanced. Before the game Vlad kindly choreographs the sitting down at the board for the TV cameras. After that the photographers have their time.

In game one we let them shoot before play and for the first few minutes after it started. This time Vlad arrived very early and I thought that six minutes of flash photography was enough. I sent the photographers out after I saw Vlad wince – lets face it, usually its both players who are distracted by the photography but not here.

Carsten Hensel. Photo Mig Greengard

There was grumblings of discontent so I consulted Carsten Hensel, Kramnik’s unflappable manager and we agreed that once play started, a couple of minutes of photographs with no flash were, in his words ‘ no problem’.

Do not adjust your sets

‘‘A bad day’ was how Fritz programmer Matthias Wuellenweber described it and in truth this was not a great advertisement for computer chess although you had to admire how ‘Big Vlad’ kept his composure after the machine came up with some incredible tricks.

There were several phases to the game as Deep Fritz played alternately like a beginner and then like a genius. Kramnik was white and play again followed one of the opening variations seen when Kramnik defeated Kasparov to capture the world title two years ago. The Fritz team assures me Kasparov has not been helping them !

Deep Fritz accepted the Queen’s Gambit, which is a typical computer ploy to start opening lines as soon as possible. Kramnik immediately exchanged queens to steer the game into the kind of quiet positions in which he could use his superior strategic knowledge. The other choice was to play a middle game with an IQP but this line with 7.dxc5 is perfect against a computer.

Then came another very cunning anti computer move 9.Kf1! its worth mentioning that with hindsight, 9.h3 would have been even better. The computer was out of book and pushed …b5 a shade too quickly for my liking. However this was nothing compared with Deep Fritz’s twelfth move which was so weird we all wondered if there had been some kind of operator error. By the way the rules state that Kramnik gets ten extra minutes if an incorrect move is corrected. If it is not spotted within a minute it stands.

Kramnik played the slightly odd 12.Nb3 and Fritz calmly played Bf8. It was an “ incredibly bad move” said Nigel Short and the phrase ‘do not adjust your sets’ came to mind.

No repetition

I was watching Fritz’s thoughts emerging on the screen in the computer room and I was amazed to see that it had erroneously decided that 12.Nb3 was a very bad move. In fact it was a perfectly sensible choice attacking a bishop. The point was that Fritz was sure Kramnik had to play b3 and Bb2 to develop the bishop and could see no further than 12…Bf8 13.Nbd2 !!

It thought that Kramnik had no better plan than to move his knight back to the square it had just come from. Fritz actually anticipated a repetition of moves but this was obviously nonsense and Kramnik promptly attacked the queenside. With two extra tempi he quickly built up a clear advantage.

But how to put the ‘little devil’ away?

Deep Fritz then played incredibly well to stave off immediate defeat. Kramnik called its resources ‘amazing’ and added ‘ no human could see this’. In Nigel Short’s words “ It’s a tricky little devil’ – if he had not been talking to members of the Bahraini Royal Family, ambassadors and government ministers at the time he might have used another expression.

For two years Kramnik must have been wondering whether he could overcome his human fallibility against a monstrous calculating machine and now here it came. Bc4+ and Nd3 was incredible. At one point we thought Fritz was going to equalise but Kramnik maintained his composure and skilfully steered the game into a rook and pawn endgame where the computer had no good moves.

In fact I could see the winning plan even before the time control and it was interesting to watch Fritz gradually work it out. In the rook and pawn endgame the winning plan is to tie black down to a5 and then open up a second front with pawn to f4.

It is not easy for a computer to defend passively and with impeccable technique Kramnik kept the position totally under control until Fritz’s pawns started to fall. On move fifty seven he checked the variations three times and then forced a won king and pawn endgame. The operator Matthias Feist noted that the Fritz evaluation indicated that the position was hopeless and he resigned on the machine’s behalf.

The match is being organised by the General Organisation of Youth and Sport in Bahrain (GOYS) and is held under the patronage of the King of Bahrain Shaikh Hamad bin Isa Al Khalifa. GOYS have put up a prize fund of one million dollars and rights holders Einstein Group PLC are televising the matches across the Arab world. Kramnik will win one million dollars if he defeats the computer, 700,000 if the match ends 4-4 and a mere 500,000 dollars if computers triumph again.

Kramnik,V - Deep Fritz [D27]
Bahrain (2)
[Malcolm Pein]

1.d4 d5 2.c4 dxc4 Computers like to open files but this is an opening that Vlad really has had great success with 3.Nf3 Nf6 4.e3 e6 5.Bxc4 c5 6.0-0 a6 7.dxc5 As played with success against Kasparov 7...Qxd1 8.Rxd1 Bxc5 9.Kf1!

What a clever idea to get the computer out of book and it works. It later turns out that the king is slightly vulnerable on f1 and h2 is undefended but this was naturally quite impossible to foresee. [ 9.Nbd2 Nbd7 10.Be2 b6 Kramnik v Kasparov game 4 London 2000] 9...b5 I don't like this because there is no need for it now. 9...Nbd7 is sounder and less weakening 10.Be2 Bb7 11.Nbd2 Nbd7 12.Nb3 A bit unusual. 12.b3 was standard, so standard Fritz short circuits. I wonder if Kramnik had this position on the board in advance. After the game he denied this. 12...Bf8??

Fritz just thought Kramnik had to play 13.Nbd2 now when it would merrily play 13...Bc5 Of course 13...Be7 was more sensible 13.a4 With the Black rook on h8 a long way from the action its time to attack the weak queenside pawns 13...b4 [ 13...bxa4 14.Rxa4 Nc5 15.Nxc5 Bxc5 16.Bd2 0-0 17.Rda1 Rfd8 18.Be1 and the pawn on a6 is going] 14.Nfd2 The standard move to head for the newly available c4 square 14...Bd5 15.f3 To play e4 and kick the bishop 15...Bd6 16.g3 e5! Its vital to keep the bishop shadowing the Nb3 and limit those white knights which are poised to invade 17.e4 Be6 18.Nc4 Bc7 [ 18...Be7 19.Nba5 0-0 20.Nc6 Rfe8 21.Bd2 and the queenside pawns are very weak again] 19.Be3 a5 [ 19...0-0 20.Nc5 Nxc5 21.Bxc5 Rfb8 22.Bd6 Bxd6 23.Rxd6+/-] 20.Nc5 Now Fritz plays quite brilliantly to find a defence 20...Nxc5 21.Bxc5 Nd7 Only move 22.Nd6+ Kf8! [ 22...Kd8 23.Rac1 ( 23.Nxf7+ Bxf7 24.Bb5 Be6 25.Rxd7+ Bxd7 26.Rd1 Kc8 27.Rxd7 Is exactly how not to play against Fritz even if its quite strong !) 23...Nxc5+/- 24.Nxf7+ Ke7 25.Rxc5 Bb6 26.Rc6+/-] 23.Be3 At this stage it was hard to imagine the computer surviving [ 23.Nb5+ At first sight Nigel Short thought this was winning. I couldn't see why not so in the end we switched on Franz Morsch's Fritz to find out !! 23...Nxc5 24.Nxc7 Rc8 25.Rac1 Looks absolutely crushing but... 25...Bh3+! 26.Ke1 Rxc7 27.Rd8+ Ke7 28.Rxh8 Nd3+ Survives] 23...Bxd6 24.Rxd6 Ke7 25.Rad1 Rhc8 26.Bb5 and now we were expecting 26...Rc7 and Fritz grovels till it dies but... 26...Nc5!! [ 26...Rc7 27.R1d2 Rac8 28.Ra6+/- Nc5 29.Bxc5+ Rxc5 30.Rxa5] 27.Bc6 They think its all over... 27...Bc4+! An amazing shot 28.Ke1 Nd3+ 29.R1xd3! Bxd3 30.Bc5 Bc4 [ 30...Bc2 31.Kd2 b3 32.Rd5+ Kf6 33.Bxa8 Rxa8 34.Rd6+ Kg5 35.h4+ Kh5 36.Bb6 And a5 falls and White should win] 31.Rd4+ Kramnik forces a rook and pawn endgame 31...Kf6 32.Rxc4 Rxc6 33.Be7+ Kxe7 34.Rxc6 Kd7 35.Rc5 f6 36.Kd2 After this forcing sequence its hard to see a good plan for Fritz even though material is level. Nearly all strong human players would try to give up a pawn to become active and set the opponent practical problems but not Fritz, he wants to hang on to his pawns and so Kramnik just expertly improves his position 36...Kd6 37.Rd5+ Kc6 38.Kd3 g6 [ 38...Ra7 39.Kc4 Ra8 40.h4 Ra7 41.h5 h6 42.f4 Opening a second front and straining the defence to the limit 42...exf4 43.gxf4 Re7 44.e5 fxe5 45.fxe5 Kb6 46.Rd6+ Kc7 47.Kd5 Wins ] 39.Kc4 g5

Fritz had now seen that f4 would hurt but this does not stop it. If anything it might encourage the sequence f4 gxf4 gxf4 exf4 Rf5 but Kramnik just prefers to let the machine suffer a bit more 40.h3 h6 41.h4! gxh4 42.gxh4 Ra7 43.h5 Now everything is perfect. Fritz can not afford the waiting move Ra7 because Rd8-h8 wins so Kramnik loses a move to set up Zugszwang 43...Ra8 44.Rc5+ Kb6 45.Rb5+ Kc6 46.Rd5! Your move my friend 46...Kc7 [ 46...Ra7 47.Rd8 Rg7 ( 47...Rf7 48.Rc8+ Kb6 49.Kd5 Rd7+ 50.Ke6 Rd2 51.Kxf6 Rxb2 52.Kxe5 b3 53.f4+-) 48.Rf8+-] 47.Kb5 b3 [ 47...Ra7 48.Rc5+ Kd7 49.Kb6+-] 48.Rd3 Ra7 49.Rxb3 Rb7+ 50.Kc4 Ra7 51.Rb5 Ra8 52.Kd5 Ra6 53.Rc5+ Kd7 54.b3 Rd6+ 55.Kc4 Rd4+ 56.Kc3 Rd1 57.Rd5+

57...Rxd5 58.exd5 Kd6 59.b4 Kxd5 60.bxa5 f5 61.a6 Kc6 62.Kc4 Kb6 63.Kd5 wins 57...Rxd5 58.exd5 Kd6 59.b4 axb4 60.Kxb4 Kxd5 61.Kb5 f5 62.a5 e4 63.a6 queens with check 1-0

 
---------------------------------------------------------------------------

Deep Fritz7/ Fritz7/Hiarcs8

  


October Issue
Out Now

  

Books 2002

  


New Software



Chess Express
  


Chess Assistant

  


Giant Chess Sets

  


Books 2000/1