(1) Kasparov,G (2785) - Comp Deep Blue [A07]
New York man vs machine New York (1), 03.05.1997
[Nunn]



1.Nf3 d5 2.g3 Bg4 3.b3 Kasparov adopts a very quiet opening called the Barcza Opening - not a bad policy against a computer.

3...Nd7 4.Bb2 e6 5.Bg2 Ngf6 6.0-0 c6 7.d3 Bd6 8.Nbd2 0-0 9.h3 Bh5
By a slightly riundabout route we have reached a position which has occurred many times in master and grandmaster chess. White's basic plan is to gain space in the ecntre with e4. Black will sooner or later have to deal with the threat of e5, either by moving his minor pieces out of the way, or by playing ...e5 himself. In the latter case, there will be a slight weakness on f5 which might be able to exploit by Nh4-f5. At the moment this last manoeuvre is impossible, because the f3-knight is pinned, but this explains why one often sees the odd-looking move Qe1 in this system: White wants to support the push of his e-pawn, while at the same time freeing his knight from the pin.

10.e3
A very unusual move, indeed I can find no examples of it from practical play. The normal moves here are 10 e4 and 10 Qe1. Of the two, I would judge 10 e4 to be the more flexible. In this way White can reserve the option of either Qe1 or Qe2, depending on Black's reply. The merits of 10 e3 are rather hard to find; Kasparov soon adopts the Qe1 and e4 plan, but having lost a tempo in the process. Perhaps his idea was simply to take DB out of its opening book.

10...h6
Kasparov's deviation at once earns its reward. Instead of putting the extra tempo to use by developing queenside counterplay, for example by 10...a5, the computer makes a pawn move which is worse than useless, because it weakens the kingisde slightly. In a few moves we will see the importance of this. Now that the extra tempo has been handed back, the balance again slightly favours White and now the computer is on its own.

11.Qe1 Qa5
The play of DB in the phase of the game leaves much to be desired. It quite reasonably plans ...Ba3 to exchange off the active bishop on b2, but this plan is easily countered and then the queen is doing little on a5.

12.a3 Bc7?!
And this is very odd. There are two possible reasons behind this move. First of all, it may simply be anticipating e4-e5 by White; secondly, DB may want to have its queen defended in case of a line such as [12...Rad8 13.Bxf6 Nxf6 14.Ne4 Qxe1 15.Nxf6+ gxf6 16.Rfxe1 although this is nothing to fear as Black's active bishops easily compensate for the insignificant weakeneing of the kingside pawn structure.; I would prefer 12...Bg6 which prepares to answer e4 by ...e5. The prophylactic ...Bg6 is necessary in this case because otherwise ...e5 can be met by Nh4. This is particularly strong because of the poor move ...h6, which means that after a possible exchnage of knight for bishop on g6, Black cannot make the natural recapture with the h-pawn. Thus Black has to play ...Bg6 in advance, ready to meet Nh4 by ...Bh7. In this phase of the game DB just responds to each momentary problem on a move-by-move, without creating any coherent plan for developing counterplay. In this respect Kasparov's choice of opening appears very intelligent. ]

13.Nh4
A rather awkward move to meet. The threat is 14 g4 Bg6 15 Nxg6, and Black has to make the ugly capture ...fxg6. Nevertheless, I think that this would have been less serious than the further kingisde weakening resulting from Black's next move.

13...g5? 14.Nhf3 e5
All this looks absolutely horrible to the human eye. Having pushed both e- and g-pawns, the f5-square has become a serious weakness. If White could plant a knight there, then the game would be decided. However, DB isn't as stupid as that; there is no straightforward route by which a knight can reach the key square. Nevertheless, this long-term weakness remains a lasting burden for Black.

15.e4
Kasparov at once takes steps to fix the weakness on f5.

15...Rfe8 16.Nh2
Kasparov proceeds cautiously. His plan is Qc1, Re1 and then Nf1-e3-f5. This is rather slow, but as White's position is solid if will be hard for Black to develop ny real counterplay. [So far as I can see, he could have started this plan immediately with 16.Qc1 , since 16...g4 is not a worry because of 17.Nh4 and the knight takes a short cut to f5.]

16...Qb6 17.Qc1 a5
At last DB hits upon a plan for developing counterplay, although in this prticular position it is not very effective. The natural follow-up is ...a4 to induce b4, and then ...c5. However, ...c5 is hard to arrange because of the pressure on d5, and if Black exchanges first on e4, then the route Nc4-e3-f5 is opened up for the knight on d2. We can see how, time and time again, Black's naturalplans are frstrated because they would seriously expose the weakness on f5.

18.Re1 Bd6 19.Ndf1 dxe4
Now that the knight has moved away from d2, the possibility of Nc4 disappears and so this exchnage becomes feasible. However, the kniught is already on its way to f5 by a different route.

20.dxe4 Bc5 21.Ne3 Rad8 22.Nhf1
Black has hit upon a good method of keeping the knight out of f5 , temporarily at any rate. The f2-pawn is quite hard to defend, and until White nullfifies the pressure against it along the b6-f2 diagonal, the knight cannot hop into f5. Kasparov's idea is Nd2-c4; then, if Black wants to mainatin the pressure against f2 he must play ...Qa7, but White continues with b4, exploiting the line-up of a1-rook with the enemy queen, and Black is in trouble.

22...g4
The computer desperately seeks to disturb White's plan. Although this move creates further kingisde weaknesses, it enable Black to develop some piece activity. This is the critical phase. Everybody who has played a computer knows the scenario: you get a strategically winning position, the computer makes some desperate tactical lunge, you make a couple of inaccuracies and suddenly the machine is all over you.

23.hxg4 Nxg4 24.f3?
In this desire to achieve his straegic ends without making the slightest concession, Kasparov commits an inaccuracy which allows the machine to develop significant counterplay. [The most obvious plan is 24.Nxg4 Bxg4 25.Ne3 . Kasparov probably realised that after 25...Be6 , it is not so easy to make progress, because Black has manoeuvred his light-squared bishop to a reasonable active square. For example 26 Qd1, heading for h5, is met by 26...Bd4. However, I think that this was his best option; after 26.Re2 Kh7 27.Qe1 White's remaining pieces come into play, and Black's kingisde weaknesses are not going to run away. Here, for example 27...Bd4 runs into 28.Bxd4 exd4 29.Nd5! cxd5 30.exd5 with a large advantage for White.]

24...Nxe3 25.Nxe3 Be7!
An excellent defensive move. A human would find it hard to abandon the pressure on the b6-g1 diagonal, but the computer ispots that the bishop has an even better square on g5, both shielding the kingside weaknesses and pinning the e3-knight from a different sirection.

26.Kh1 Bg5 27.Re2
Kasparov recovers well. He intends Qg1 followed by Nf5, and the damage is repaired, but the computer strikes first.

27...a4
A good introductory move. As we shall see, having c4 avilable helps Black.

28.b4 f5
Correct. Black must make use of his temporary piece activity to make some inroads. Passive play would allow White to get back on track exploiting his strategic advantages.

29.exf5!?
[If White tries 29.Qe1 fxe4 30.Nc4 Qa6 Now we see why one reason why Black prepared this line with ...a4; the attack against the c4-knight is awkward. 31.Rxe4 Nf6 and White must jettison some material. He will always get some compensation, for example here 32 Nxe5 is unclear, but at least DB has randomised the game. The move Kasparov plays alsoinvolves an exchnage sacrifice.]

29...e4 30.f4 Bxe2
[30...Bxf4 31.gxf4 Bxe2 32.Qd2 Bh5 Forced, or else White also has the possibility of Ng4 33.Qc3 Re7 34.Nc4 Qc7 35.Qh8+ Kf7 36.Qxh6 and wins.]

31.fxg5 Ne5
Black must block the long diagonal. [31...hxg5 32.Nd5 wins at once.]

32.g6
[32.gxh6 Kh7 is a more reasonable reply, when Black has much better chances than in the game. (32...Rd6? The Chess Monthly web site comments that 32 gxh6 Rd6 gives Black a decisive attack, although White can win in one move by 33 Nc4!. 33.Nc4! ) ]

32...Bf3 33.Bc3
A good prophylactic move. In many lines the queens are exchanged or White would like to send his queen away to the kingside. In these cases it is important not to allow ...Rd2.

33...Qb5?
A serious error. DB sees that this will probably lead to the exchange of queens, and has no objection in view of Black's material advantage. However, in the ending White's advanced pawns and general grip on the poisition count for more than the small material plus of rook for bishop and pawn. Instead, Black should have kept the queens on the board. [33...h5 is one possibility. In many line Black gains a tempo because White acnnot now play his queen to f1 in one move. One line runs 34.Qe1 Qb5 35.Qf1 Ng4 36.Nxg4 (36.Qxb5 cxb5 37.f6 Nxe3 38.f7+ Kf8 39.fxe8Q+ Rxe8 40.Bxf3 exf3 41.Kg1 is probably a draw) 36...hxg4 37.Qxb5 cxb5 38.f6 Re6 and Black can defend.; Another idea is 33...Qc7 intending to move the queen over to the kingside. Either of these options would have left the position unclear, whereas now Deep Blue slides downhill.]

34.Qf1 Qxf1+
[Now 34...Ng4 is impossible because of 35.Qxb5 cxb5 36.Nxg4 Bxg4 37.f6 ]

35.Rxf1 h5 36.Kg1!
At first sight this position shouldn't be too bad for Black, since the e5-knight's attack on the g6-pawn means that the pawns cannot advance for the moment. However, Kasparov's move makes it all clear; he can afford to take his time, because Black has no constructive moves. The f3-bishop cannot move because of f6 and f7+, the knight must stay on e5 to cover g6, and this ties down the e8-rook too. The rook on d8 cannot achieve anything by itself, since the d-file penetration squares are under control, which leaves only Black's king.

36...Kf8 37.Bh3 b5 38.Kf2 Kg7?!
Making life easy for White by allowing g4 under favourable circumstances.

39.g4 Kh6 40.Rg1 hxg4 41.Bxg4 Bxg4 42.Nxg4+ Nxg4+ 43.Rxg4 Rd5 44.f6 Rd1
See Frederic's piece for an analysis of 44...Rf5+. [44...Rf5+ 45.Kg3 (45.Ke2 Rg8 46.g7 Kh5 47.Rg2 Rf3 48.Bd4 Kh6 49.c3 Kh5 50.Rg1 Kh6 51.Rg4 Kh5 52.Rxe4 Rf5 53.Re6 Kg6 54.Rxc6 Re8+ 55.Kd2 Rh5 56.Re6 Rh2+ 57.Kd3 Rh3+ 58.Be3 Rd8+ 59.Ke4 Kf7 60.Rc6 ; 45.Ke3! Rf3+ 46.Ke2 Rxc3 47.f7 Rd8 48.g7 Rxc2+ 49.Ke1 Rc1+ 50.Kf2 Rc2+ (50...e3+ 51.Kg2 e2 52.g8Q Rxg8 53.fxg8Q Rg1+ 54.Kf3 Rxg4 55.Qh8+ Kg6 56.Qe8+ Kf5 57.Qf7+ Ke5 58.Kxg4 ) 51.Kg3 Rc3+ 52.Kh4 Rc1 (52...Rd1 53.g8N+ ) 53.g8Q Rh1+ 54.Kg3 Rg1+ 55.Kf4 Rf1+ 56.Ke5 Rd5+ 57.Ke6 Rf6+ 58.Kxf6 Rd6+ ) 45...Rf3+ 46.Kh4 Rd8 47.f7 Rd5 ]

45.g7
A well-played game by Kasparov, from both the chess and the psychological point of view.

1-0