An analysis of the sixth game. (# 6.) 

  Deep Junior (C) (2633) - GM Garry Kasparov (2847) 
[B20]
X3D "Man vs. Machine" Match
New York City, NY;  [U.S.A.]  (Rd. # 6),  07.02.2003

[A.J.G.]

This is the last and final game of this match. ESPN (channel # 61) 
is covering this match on TV. (LIVE!)  (I recorded the whole show.) 

GM Yasser Seirawan  and also  GM Maurice Ashley  are commenting 
on this game as it is being played. 

 (Both sides start with 2 hours.) 

(At this point the match is tied, Kasparov needs a win to really
  impress.) 

 << The comments in brackets are from my web page on this game. >> 

1.e4 c5;  2.Nf3 d6;  3.d4 cxd4;  4.Nxd4 Nf6;  5.Nc3 a6;  {Diag?}
 << Although Garry played a Sicilian in all three games where he 
      championed the Black pieces, this is his very first use of the 
      Najdorf Defense. (In this match.) >>

<< All the commentators were predicting a sharp game. >> 

6.Be2,  {See the diagram just below.} 
A perfectly reasonable developing move, but this move disappointed 
a few of the commentators. 

      [  Many watching the game were hoping for the move:  6.Bg5,  but this 
         would have probably fallen into the brunt of Kasparov's plans ... and 
         (into) the bulk of his  preparation. 

         This move could also lead to the ultra-sharp variations of the Najdorf, 
         which - quite possibly - Garry knows better than any player alive, 
         or at least active today. 

         I am quite sure that GM Boris Alterman - who was charged with 
         streamlining Deep Junior's opening repertoire - thought it was better 
         to avoid a theoretical duel with perhaps the world's leading exponent 
         of these lines. ]  

6...e5; ('!?')  {Diagram?}  

The first ever use of the Boleslavsky - type move here by Garry, at least 
according to expert commentator, 4-time U.S. Champion,  ... ... ... 
GM Yasser Seirawan. 

   (I think Yaz was mistaken. A search of my database reveals that Garry   
    played this in a junior tournament in the USSR in like 1973. He also   
    played this line in a tournament in Leningrad, 1975 against no less    
    than Karpov. (!)  He used it against Arnason in the World Junior Champ,   
    in 1980. He has also been on the Black side of this position 2 or 3    
    times against Anand. {Las Palmas} In fact my database reveals Garry    
    has used this around nine times, prior to this event.)    

Was this  (the move  ...e5)  prepared in advance?
(I think it was.)

According to  ChessBase's  on-line  database,  this variation, (with 6...e5); 
has been played over 1000 times at the master level. (!!) 
(A NIC CD-ROM gives almost 7,000 examples of this opening.) 

The theoretical manuals also consider this thrust, (6...e5); to be 
 the main line here, and was worked out in great detail by Najdorf 
 and (of course!) Bobby Fischer. 

*****

     [  Previously, in most of his other matches, Garry had almost always 
        preferred to play the move:  6...e6;   in this position. 
        (Which is actually a transposition to the main lines of the 
         Scheveningen System.) 

         [ A search of my own personal database of all of Garry's games 
           show nearly 60 examples of this line occurring in Garry's games. 
           For example, this position - with Garry playing the Black pieces - 
           was played like nine times in in his first two {WCS} match(s) with
           Karpov. 

           Garry also played the Black side here like 4 or 5 times in his WCS
           match with Viswanathan Anand. ] 

        According to the Informant data-base, this position (after ...e6;) 
        had been reached in Master-level chess nearly 1500 times. 

        (But the bulk of these games were played in the 70's and 80's when
         Be2 was at the height of its popularity.) 

        Now  MCO  gives the continuation:  
        7.0-0,  {Diagram?}  
        The move that is played the most often in this position. 

          (Or White could also play: 7.a4!? Nc6;  8.Be3!? Be7;  9.0-0 0-0;  
           10.f4!? Qc7; (!)  11.Kh1 Re8;  12.Bd3!? Nb4;  {Diagram?}    
            Black has a good position, and went on to win a nice game.  

            GM C. Lutz - GM G. Kasparov;  The FIDE World Chess Olympiad,   
             /Bled, SLO; 2002/TWIC # 418  (0-1,  42) )    

         (Returning to the MCO line.) 
        7...Qc7!?;  8.f4 Be7;  9.Kh1 0-0;  10.Bf3!? Nc6;  11.Nxc6!? bxc6;  
        12.Na4 Bb7;  13.c4 c5;  14.Qc2 Nd7;  {Diagram?}  
        The end of the column. 

        15.Be3 Rad8;  16.b3 Rfe8;  17.Rad1 Bc6;  18.Nc3 Nf8;  
        19.g3, "~"  {Diagram?}  ... "with just a minor edge." ("+/=")  
        (space) - GM Nick de Firmian. 

        Landa - Milov;  Oberwart, 1994.  

        [ See MCO-14; page # 292,  column # 10, and also note # (o.). ]  ]  

*****

Both sides continue to develop. 
7.Nb3 Be7;  8.0-0 0-0;  {Diagram?}   

Probably it is safest for both sides to castle here. 

     [  Black could also play: 
        8...Be6!?;  9.f4 Qc7;  10.Nd5 Bxd5;  11.exd5 Nbd7;  12.c4 0-0; 
        13.Kh1 Rfe8;  14.Be3 exf4;  15.Bxf4 Bf8;  16.Rc1 Qb6;  
        17.Rc2 g6;  18.Bf3 Rac8;  19.Nc1 Ne5;  20.b3 h5!?;  21.h3 Bg7; "~"
        {Diagram?} 

        This position is unclear ... and very close to being dead even. 

        But Black went on to win fairly quickly, but only after Anand 
        made an error. 

        GM V. Anand - GM G. Kasparov;  
        The 'Super GM' Tournament;  Linares, ESP;  2000.  (32) 

        Bibliophiles can check out the excellent annotation of this game by 
        GM L. Shamkovich.  ('Chess Life', July - 2000.)  I am told for a 
        small contribution, The USCF will send you a photo-copy of this game. ]

9.Kh1,  {Diagram?} 
This move is nearly automatic, according to many of my books; 
especially ECO and MCO. 

I guess the main idea is to get the White King safe, and away from any 
potential threats. White is also preparing f2-f4, while keeping the first 
player's position as flexible as possible. (It seems both sides try to 
avoid dedicating themselves to a single course of action too soon.)

     [  White could also play the move:  9.Be3,  {Diagram?};  
        or even  9.Qd3!? "+/="   {Diag?}  with at least a small edge. ]  

9...Bd7;  ('!?')  (TN?)  {See the diagram just below.}  
Kasparov finds an ingenious way to get the computer out of book -
or at least avoid the bulk of the Deep Junior team's preparation. 

A diligent search of several databases reveals that this position probably 
occurred in a game between two lesser players in a USSR (Soviet)
Championship in the 1930's (Probably by transposition.) 

ChessBase's own online database reveals that their first record of this 
position occurred in the encounter between  V. Jansa  and  L. Gutman; 
in a German League game in 95. 
(Only a handful of games with this position.)  

     [  MCO - instead - gives the following continuation: 
        9...b6;  10.Be3 Bb7;  11.f3 b5!;  12.a4 b4;  
        13.Nd5 Nxd5;  {Diagram?} 

        The end of the column. 

        14.exd5 Nd7;  15.c3!? bxc3;  16.bxc3 Bg5;  17.Bg1 Qc7; "=" 
         {Diagram?}  The position is very balanced.  

        GM V. Anand - GM B. Gelfand;  Dos Hermanos, 1997. 
        (A quick draw.)  

        [ See MCO-14;  page # 256,  column # 19, and note # (f.). ]  ]  

10.Be3 Bc6;  {Diagram?} 
Black forces White to do something about his e-pawn. 

     [  Not as accurate was: 10...Nc6!?;  11.f3 Rc8;  12.Qd2, "+/="  {Diag?} 
         and White still has a very small edge in this position. {A.J.G.}  ]  

11.Bf3 Nbd7; {Diagram?} 
The commentators were already saying here that Black was for 
choice  - - -  in this position. (That was a little much.) 

White's play has not given him an edge, he is dangerously close to 
handing the initiative over to Black. 

12.a4,  {Diagram?}  
This prevents Black from playing the pawn advance, ...b5. This would 
gain valuable space for the second player and would possibly threaten 
... b5-b4, undermining the piece (N on c3) which is supporting the 
White e-pawn. 

     [ 12.Qd2 ]  

Now Black continues to consolidate his position, while the box plays 
{apparently} pointless moves. 
12...b6;  {Diagram?}  
This is good because it prevents White from "freezing" the Q-side 
 with moves like a4-a5, and then maybe N/b3-d2-c4-b6. 

13.Qd3 Bb7;  14.h3 Rc8;  15.Rad1 h6;  16.Rfe1 Qc7;  {Diagram?} 
Black has the slightly more flexible game, while White has managed 
to centralize both of his Rooks. (Which is never a terrible idea.) 

17.g3!?,  {Diagram?} 
An interesting move, but it also weakens greatly White's King-side. 
(Time - 1:27)  

     [ 17.Kg1!? ]  

17...Rfd8;  {Diagram?}  
Black centralizes.  

18.Kh2!?,  {Diagram?}  
Ashley says the program appears to be slightly floundering.  

     [ 18.Bg2 ]  

18...Re8!?;  {Diagram?}  
 << It is hard to say exactly what this move accomplishes.
      (Rope-a-dope?) >>

<< (All the commentators liked the idea of ...Qb8 here.) >> 

GM Ashley  says Garry is just waiting around for the computer 
to make a mistake. 

     [  Maybe best was: >/=  18...Qb8!;  19.a5!? b5;  20.Bg2, 
         20...Qc7!; "=/+"  {Diagram?}  and Black is slightly better.  

         Black could also try: 18...Nc5!?;  or  18...Bf8!? ]  

19.Re2,  {Diagram?}  
Is White preparing to double on the d-file in this position? 
(The c-pawn is also guarded, is the program perhaps preparing 
  the move Nd5?)  

     [ 19.Bg2!? ]  

19...Qc4;  {Diagram?}  
 Garry wants to play an ending.
(Time - 1:04.46)  

<< Black exchanges the Queens, the advantage is definitely to the 
     second player in this position. >> 

     [ 19...Nc5;  or  19...Red8!? ]  

20.Qxc4,  {Diagram?}  
The computer decides to swap, but is this wise?

     [ 20.Bg2!?;  or  20.Kg2;  or  20.Qd2!? ]  

20...Rxc4;  {Diagram?}  
This is a forced recapture.  

21.Nd2,  {Diagram?} 
White re-enforces e4 and runs the Rook off of c4.  

      [ 21.Bd2!? ]  

21...Rc7;  {Diagram?}  
The safest square for the Rook. 

     [ 21...Rcc8!? ]  

22.Bg2,  {Diagram?}  
Is White consolidating here? 

     [ 22.Rc1!? ]  

22...Rec8;  {Diagram?}  
Kasparov doubles on the file.  

     [ 22...Rd8 ]  

23.Nb3,  {Diagram?}  
The computer appears to be sort of wandering aimlessly around. 
(Time - 1:03)  

     [ 23.Kg1!?;  or  23.f4!? ]  

23...Rxc3!!;  {See the diagram just below.}  
A nice sacrifice which comes very near to winning the game for Black. 

This is a stock sacrifice for a human, but computers still don't conceptualize 
well with just very broad, vague and very general ideas. 

According to the commentators, Kasparov actually offered a draw 
here.

(Garry thought for a while before playing this move.)  

 Garry just 'sacked' on c3. 

     [ 23...Rc4!? ]  

24.bxc3, {Diagram?}  
Naturally White has to recapture. 

     [ 24.a5?? Rxc2; "-/+" ]  

24...Bxe4;  {Diagram?} 
This move is best, capturing on c3 immediately might give
White time to defend his e-pawn. 

     [ 24...Rxc3!? ]  

25.Bc1, ('!')  {Diagram?} 
White clears the center for his Rooks.  

This move, which was severely criticized by some on-line commentators, 
may actually be the best move here and may even deserve an exclam.  
(The idea is to place the Bishop on a3 to attack Black's one real weakness, 
  the backward pawn on d6.) 

     [ 25.f4!? ]  

25...Bxg2;  {Diagram?}  
Black decides to swap pieces and further go for material. 

     [ 25...Rc4 ]  

26.Kxg2,  {Diagram?}  
This is absolutely forced.  

     [ 26.Bb2?? Bf3; "-/+" ]  

26...Rxc3;  {Diagram?}  
Garry decides to win another Pawn  --->  immediately.  

<< I am not sure about this move, 26...d5; might have been 
      a fuzz better. >>  

     [ Two viable alternatives were: 26...d5!?;  {Diagram?}  

        and  26...Rc4!?;  {Diagram?}  with a playable game 
        in both cases. ]  

27.Ba3,  {Diagram?}  
White ... almost predictably attacks the d-pawn. 

     [ 27.f4 ]  

27...Ne8;  {Diagram?}  
Black defends the d-pawn.
(Time - 0.40:30) 

     [ 27...Nc5!?;  or  27...Rc6!? ]  

28.f4!?,  (Maybe - '!')   {See the diagram just below.}   
White decides to open the game up at this point, but is this wise?
(Time:  0.34:39) 

( Final position: 
   4n1k1/3nbpp1/pp1p3p/4p3/P4P2/BNr3PP/2P1R1K1/3R4, see below. ) 

<< This is very good, probably a lot better than many of the people 
      commenting on this game originally realized. White opens lines 
      for his Rook(s), and also threatens to isolate the Black QP. >>  

<< Here (apparently) the Deep Junior team offered a draw ... 
      which Kasparov immediately accepted. >>  

Now a draw is agreed upon. Apparently Black's edge has been - 
or will be minimized by the possibility of White's exchange on e5 
and the isolation of Black's d-pawn. 

     [  White could (also) play: 28.a5!?; or  28.Bb2!?; 

        It would seem that the position after the continuation: 
        28.f4!? Kf8!?;   ( 28...f6!?)    29.fxe5 Nxe5; "="  {Diagram?} 
        is dead equal, or even slightly favors the first player. ]  

***

Those interested in the analysis and news stories surrounding this match 
 should visit my ... "Computers vs. Humans" web site.
 (https://www.angelfire.com/fl5/ human_fan02/index.html)
This site is completely dedicated to ONLY the chess played between the 
 best human players and the best programs!

***

Copyright (c) A.J. Goldsby I. Copyright (©) A.J.G; 2003. 

  ½ - ½  

   (Code initially)  Generated with  ChessBase 8.0