GM Garry Kasparov - Fritz_X3D;  

  I have taken my time annotating (& formatting) this game.   

My goal here was NOT speed, but accuracy. Many of the jobs of annotation contained gross inaccuracies or oversights. I also recorded all of the games coverage on TV, and I wanted to watch those tapes repeatedly in order to do the best possible job. (Although the television coverage of games 1-3 were interrupted ... something I have no control over. Only the television coverage of Game Four was complete.) But I am thankful I was able to record all the available TV coverage that was available. 

(I also wanted to give - at least in a small way - a flavor of what it must have been like to watch these games on TV. Where appropriate, 
 I have quoted the ESPN Team of commentators.  See  game four  for my breakdown of the ESPN Team.) 

Speed was not required, as I posted (at least) the score (and a re-cap) of this game within just a few hours of the completion of the first game. There were also many other annotated versions of this game on the web, and I also provided links to those as well. 

 Tuesday;  June 22nd, 2004. 

This game has taken a lot more time than the other three. However, there were very good reasons. 

Since this game was first played: 

1. I have had several big events in my life. (Some of these are private - I have no intention of discussing them here.) 

2. My wife was diagnosed with cancer - which set back my chess business, (and all of my websites); at least a year. 

3. I have had several computer crashes and at least one hard drive crash. 

4. I had several major virus attacks ... which wiped out literally hundreds of files. 

5. My main computer failed ... and had to be replaced. 

Additionally, this page - along with several of the other pages connected with this match - CONTINUOUSLY kept becoming unstable, unreadable and corrupted. I would have to delete it ... and start from scratch. (I never did figure out what or where this particular bug came from.) HOWEVER, I never gave up!!!!!!!!  I kept returning to it, and eventually got it done! {A.J.G.} 

 GM Garry Kasparov (2830) - FRITZ_X3D (2675) 
 [D45] 
  Human vs. Computer/VR/3D Athletic Club,  
 New York City, NY; (Round # 03) 16.11.2003 

  [A.J. Goldsby I]  

Game Number Three. (# 3.)  {Of a four game match.}  

If Garry was to have a chance ... he must hit back and try to win today. 

It was obvious that Kasparov's emotions were running very high ... and that he game to the board ... "pumped" and ready to go. 

 1.Nf3,  {Diagram?}   
Once again the tricky Reti.  (A highly transpositional opening.)   

******************************************************************************************  

      [ Lately ... against other GM's ... Garry's favorite (first move) weapon has been:  
        1.d4,  {Diagram?}  as  Kasparov  played  against  Chuchelov  in the ECC event.  
        {GM Garry Kasparov ... in a very brilliant and quick win, (a mere 27 moves!!); ... 
          against a FIDE 2600+ player ... as White from a QP opening.}   

      ****************************************************************  

        A more normal move order for this line would be the following continuation:   
        1.d4 d5;   2.c4 c6;  3.Nc3 Nf6;   4.e3 a6;  5.Nf3 e6; "~"  {Diagram?}   
         reaching the same position that we had reached in this game.   

  *******  

         ( A good sample game in this whole variation is the following highly instructive contest:  
            5...b5!?;  6.b3!? Bg4!?;  7.Be2 Nbd7;  8.0-0 e6;  9.h3 Bf5;  10.Bd3 Bb4!?N;   
            (Apparently, this move was new to opening theory at the time this game was played.)  

            11.Bb2 Bxd3;  12.Qxd3 0-0;  13.Rfc1 Bd6;  14.Rc2! Rc8!?;  {Diagram?}  
             This does not appear to be completely accurate here. 

                ( Maybe just:   14...b4!?; "~"  instead? )     

            15.Rac1 dxc4!?;  16.bxc4 b4;  17.Ne4 Nxe4;  18.Qxe4 Qa5;  19.Nd2!? Qh5;  {Diagram?} 
             I am not sure if this is best or not here, maybe ...Qf5;  instead.  

                ( Not   </= 19...Qxa2??;  20.Ra1, "+/-" )     

            20.c5 Bc7;  21.Qd3 a5;  22.Qa6!? f5!; "<=>"  23.Nc4 Nb8;  24.Qb7 f4?;  {Diagram?}  
            This should lose instantly ... Black had to retreat the Queen to the f7-square. (Or even e8.)    

            25.exf4?, (hmmm)  {Diagram?}   I guess he didn't see it!     

*******

                ( White should have played the following continuation:     
                   >/=  25.d5! exd5;  26.Nd6! fxe3; {Diagram?}      
                   {Black cannot capture the Knight on d6 with his Bishop,      
                     as this would allow QxP/g7 mate.}       

                   27.fxe3! Qg5;  28.Nxc8 Qg3;  29.Nd6 Qxe3+;       
                   30.Kh1,  "+/-"  {Diagram?}        
                   The first player should win from here ... without great difficulty. )       

*******

            25...Bxf4!?;  26.Re1 Bc7;  27.d5! exd5!?; (tempo) (Maybe - '?')  {Diagram?}    
            I think that Black had to try the move,  ...Qg6;  here - in this position.  

            28.Nd6 Qg6!?;  29.Re7,  ('!')  {Diagram?}   
            Black Resigns, ...  his  game  is hopeless. ("+/-")   

            GM Boris Gelfand - GM Peter Acs; /  Budapest, HUN;  2003.   [D15]  

             See Informant # 87, game # 337. (Page # 210.) );  ]   

******************************************************************************************  

There was a slight delay here ... as the computer did not respond to 1.Nf3 right away. 
{Either it was a small technical glitch, or they did it on purpose ... in order to give Garry a moments respite. 
  I.e., let the photographer's take their pictures, and then let the room clear before beginning the game in earnest.} 

In the meantime, it was obvious that Garry's eyes were bothering him, he rubbed them almost constantly during the 
first two games. Meanwhile, Garry was trying on several different pairs of 3D glasses while waiting for the computer 
to respond to 1.Nf3.  
 1...Nf6;  2.c4 e6;    
Fritz_X3D had a special opening book in it, which the programmers and the Fritz team had prepared especially for this match.  

They boasted that it included every single game that Kasparov had ever played.  

But here ... that may not be enough. In my opinion, the computer failed miserably - from a strategical point 
of view - in both Game One ... AND also in this game. (Only the computer's tactical ability allowed it to 
save the first game of this match.)  

2...e6;  heads for a type of position in which the computer does not seem to really handle very well.  

     [ I think the computer - and its team of programmers - should have seriously considered playing 
       the move:   2...g6;  {Diagram?}   
       which leads to the kind of position where tactics play a greater role. ]  

Garry deserves cudo's for his move order ... which tempts or forces the box back into the Slav Defense.   
 3.Nc3 d5;   
This leads ... practically by force ...to the Queen's Gambit Declined.  

     [ Maybe the move:   3...Bb4!?;   was worth a try?  ]   

 4.d4,  ('!')   
Probably the best move. At this point, the Fritz team on TV looked to be quite happy.  

     [ Possible was:  4.b3!? ]   

 4...c6!?;   
Black is obviously angling ... (again) ... for the Slav. But is this really the best line for the box?  

     [ After the following continuation:    
       4...Nbd7;   5.Bg5,  {Diagram?}   
      we can get the Classical Variation of the Queen's Gambit Declined.   
      (Rock-solid ...  but slightly more drawish than the text.) ]   

Once again, Garry heads for the Semi-Slav lines that feature an early e3 and Qc2.  
(See Game # 1 for a thorough look into the opening.)  

 5.e3! a6!?;  (A mistake in strategy.)   {See the diagram - just below.}      
With Garry seemingly content to repeat the opening of Game # 1, the team for Fritz is the first to blink.   

This strikes me as slow and relatively risky ... especially as Garry shows an extremely good grasp of all   
of these lines.  

Of course these lines are currently all the rage  ...  (VERY topical!);  there is even a brand-new book about an
early  ...a6 for Black in the Semi-Slav.  (As of this writing, {January, 2004};  I have ordered this book, but I   
have not yet received it in the mail.) 

 ******************** 

 ******************** 

This line or variation is actually very old, with this position having occurred in one of Rubinstein's games. 
(Also M. Euwe defeated A. Alekhine in this line in one of their World Championship Matches.) 

So the wheel - once again - comes full circle ... a line which was popular 50-to-75 years ago in now   
the current height of GM practice. (Black's idea is to play ...b5, and then maybe  ...Ra7;  ---> and move  
the Rook into the center.)   

     [ After the moves:  5...Nbd7;  6.Qc2 Bd6;  {Diag?}  5...Nbd7;  6.Qc2 Bd6;  {Diag?}   
        we have the same position as in Game # 1.  
       (Which was NOT really in the computer's favor!) ]   

Garry's next move is brilliant ... if for no other reason than it probably takes the program  ... "out of book."  
{A programmer who used to work for ChessBase e-mailed me and told me that this move was not in Fritz's  
  book just a year or two ago.}   

At this point - according to the opening "tree" that comes with the program, Fritz 8.0 - the following moves are   
in the computer's book: Qc2, Bd3, b3, and also the move Ne5.  

 6.c5!,  (Q-side, dark squares)    {See the diagram ... just below.}   
Garry closes the position ... the speed with which he played his first few moves would lead me to believe that this is all ... "a prepared line"  ...  by Kasparov. (Garry took about nine minutes to play this move ... during which time he was probably - mostly - just trying to remember all of his analysis.) 

 ******************** 

 ******************** 

This move also is an obvious attempt to exploit the weakened Black square at b6 by White.  

This is also an excellent  "anti-computer line,"  as White closes the position, and also sets up a long pawn chain. 
Garry also has created a position that emphasizes things like positional play, planning, and also strategy ...    
all things any 'box'  ... or program does very poorly at.   

At first, when this move was first played, everyone accused the Fritz team of trying to employ  ...  "Psychological Warfare." 
(Garry had badly lost a game - as Black - in this line, and maybe everyone thought they were simply trying to ... 'psyche 
 him out.') But I think it goes much deeper than this. I am sure that this idea - by Garry - was prepared many months in   
advance. There also were not that many games in the database. So once the opening was chosen, I do not think that the   
Fritz Team could have realistically avoided this line.  

Garry did think for close to five minutes here. But I don't think that Garry was 'thinking' chess ideas as much as he was   
simply reviewing - in his head - the various (and multitudinous) opening variations that Garry had prepared for this match.   
(Indeed ... during much of this period, Garry took off the "3D" glasses ... and even had his eyes closed.)    

     [ Possible was:  6.Qc2!?,  "+/="  {Diagram?}  
       with play similar to  Game One.  

                                              *******   

       Be2 looked playable here:  
       6.Be2!? b5!?;  7.b3!?, "+/="  {Diag?}    
       and White maintains a tiny edge. ]   

The next few moves were played at an almost machine gun pace.   
 6...Nbd7!?;  ('?!')   
It is almost never wrong just to simply develop.  (But here, this looks just a little naive.)   

The computer played  {responded} with this move instantly.  

     [ Possible was:   "="  6...b6!?;  "~"   {Diagram?}     
       to immediately undermine the White Pawn chain.  

                                              *******   

       Upon days of reflection, I think the best move had to be the try:  
       >/=  6...a5!;  {Diagram?}   
       which would have prevented a quick b4 by White. ]   

 7.b4!,  {See the diagram ... just below.}     
An excellent move ... which bolsters the White Pawn on the c5-square.   
(It's main draw-back is that Garry delays developing a piece for one move.)  

 ******************** 

 ******************** 

I have a friend who is also a former Internet student. He has a database which he maintains with an almost religious zeal. I asked him to put this position into the computer and search his vast database of games. He came up with only 57 relevant matches. So this line is an excellent choice by Garry, many of the lines have not been critically examined in over fifty years ... or even more.  

     [ While watching the game on TV and on the Internet, I expected simply development with a   
        move like:  7.Bd3, "+/="  {Diagram?}   and White is probably just  slightly better.  ]   

The computer - also - responded with its next move in a very rapid manner.   
 7...a5!?;  ('?!')   
I thought this might be a bit premature for Black. 
(The Pawn moves twice - in the opening phase - in a very brief period of time. Not good.)   

While criticized by some ... at this point, the computer is only following the games that are in 
its 'opening book' / its data-base of games on this line. 

     [ Playable was:   7...Qc7!?; "~"  (Unclear.)  ]   

Garry now stuffs the Pawn down into the b5-square, a blow from which the machine never   
really recovers from.  
 8.b5!,   
The correct move, gaining some very serious Queen-side space for White in this position. 
(Exchanging presents Black with no real problems from this position.)  

      [ </= 8.bxa5!? Qxa5;  9.Bd2 Qc7; "=" ]   

 8...e5!?;   
 Black breaks in the center.
(Whether or not this is absolutely the best approach for the second player in this position is 
 really anyone's guess.)  

{This move was also played in a very rapid - maybe even in an almost instantaneous - manner.} 

     [ Not to be recommended was:   
       </= 8...cxb5?!;  9.Bxb5 Be7;  10.0-0 0-0;  11.Rb1, "+/="  {Diag?}   
       and White is clearly better. 

 ******* 

       Maybe OK for Black would be the continuation of:   

       = 8...Be7!?;  9.Bd3 0-0;  10.0-0,  "+/="  {Diagram?} 
       but White has a solid edge. ]    

Kasparov paused just slightly, before playing his next very fine move.  
 9.Qa4!,  ("+/=")    
White chooses to  {correctly} increase the pressure. White should also beware any continuation   
that involves breaking up his pawn chain, as this rarely leads to a permanent advantage.  

 ******************** 

 ******************** 

White is already clearly (a little)  better in this position. 

<< Given an exclamation point by Gligoric and Wade in their book, "The World Chess Championship,"  
      (1972). This move was apparently not in X3D Fritz's opening book so it was now on its own. >>   
       - ChessBase.  

Garry has used about 11-12 minutes to play all of the moves thus far.  

     [ After a few moves of:  9.dxe5 Ne4;  10.Nxe4 dxe4;  11.Nd4 Nxe5; "="  {Dm?}   
        Black's position appears to be rock-solid here. ]  

Apparently the computer was not prepared for Garry's ninth move, it was several minutes before   
the computer was able to respond.  
 9...Qc7; ('!?')   
It is not really clear if this Queen move is either good or even accurate.  

But at the same time, it is not really very easy for Black to find a line that will guarantee the second player   
true equality. (Black obviously saw the need to defend the c6-square.)  

     [ </=  9...Nb8??;  10.Nxe5,  "+/-"   

       </=  9...cxb5?;  10.Bxb5,  '±'  

                    *******   

       Black had to try and break out of the bind:   
       >/=  9...e4;  10.bxc6!? exf3;  11.cxd7+ Bxd7;  12.Qd1 fxg2;  "<=>"   
       with at least some active play for Black from here. {A.J.G.} ]   

 10.Ba3!?,  ('!')    
In a sense, this is simply a high-class waiting move ... but one that complements White's strategy    
of trying to dominate the dark squares.  

     [ It was also very playable for White to play b6, and seal the Queen-side.  
        It was also playable for White to play a simple move like:  10.Be2!?, "+/="   
        and White maintains good pressure. ]   

 10...e4!?;  ('?!')    
The computer closes the center ... Garry should say ... "THANKS!!"   

With the center completely blocked, the emphasis is on STRATEGY ... and long-term planning.  
This is also (now) a POSITIONAL struggle rather than primarily a tactical one.  
(I.e.,  the emphasis is now about a very slow struggle for controlling key squares on the chess board.  
In a tactical struggle ... the main part of the struggle focuses on the real interplay between the pieces.  
And of course when the position is wide-open, and the fight is mostly a tactical struggle ...   
the computers have the advantage because tactics is what the boxes do best - at least at this stage  
of the development of computer-chess technology.)  

     [ Maybe:  10...Be7!? ]  

Now Garry will soon close the Queen-side completely ... with the idea of (artificially) isolating ... 
and later annexing ... Black's poor QRP. 
(White's next move is virtually forced ... d2 being the only decent square available to White's attacked   
 steed ... that is currently on f3.)  
 11.Nd2,  {Box?}    
This could be best ... I am sure moves like Ng1!?,  (dubious?);  and Nh4? were not even being seriously   
considered by Garry Kasparov.  

     [ Was  11.Ng1!?,  playable here? ]   

 11...Be7;    
A rather normal developing move. 
{Seirawan predicted this move - indeed he correctly predicted many of the moves played in this opening.}  

     [ Maybe  11...h6!?;   was playable? ]  

 12.b6!,  (TN)   {See the diagram ... just below here.}   
Without delay ... Garry closes the Queen-side. 
(Waiting might give Black a chance to try and {maybe} liquidate {swap-off} the Q-side Pawns.)   

 ******************** 

 ******************** 

And according to several experts ... this is the first original move of this particular game.  

     [ The only other contest to have seen this same position proceeded as follows:  
        ("</=" or "=")  12.Be2 h5!?;  13.b6 Qd8;  14.h3!?, "+/="  (Maybe - "±") {D?}  
        and White is (was) very clearly better ... although Black went on to win this 
        historic contest.  

        GM Samuel Reshevsky (2723) - GM Paul Keres (2738);   
        (FIDE)  World Championship Tourney   
        "The Hague" / "Hall of Columns" / NLD / U.S.S.R;  1948.  

             (Interesting was:  14.Nb3!?, "+/="  {Diag?}   
                with a slight edge for White.)  ]    

 12...Qd8;   
This appears to be the best square for the attacked Queen.  

     [ After the move of:  (</=)  12...Qb8!?;  (?!)  {Diagram?}  
        it is not clear if Black's Queen will ever {again} see the 
        light of day! ]  

 Garry took several minutes before playing his next move.  
 13.h3!,  (Maybe - '!!')   
An extremely sophisticated ... "ANTI-Computer" type of move.
(White has no intention of castling King-side, as this might allow White to generate a meaningful attack.  
 The computer them might try shoving the King-side Pawns ... in an attempt to open as many lines as    
 quickly as possible.)  

GM Yasser Seirawan pointed out something very profound here. The move h3 prevents ...Ng4;  
{and maybe a later ...Bh4}  by Black. And if Garry had played Be2 instead, his g-pawn would   
be vulnerable to the Knight maneuver of  ...Nf8-g6-h4.  

     [ Interesting was the try:   13.Be2!?,  "+/="  {Diagram?}    
        and White still holds an evident advantage in this position. ]    

 13...0-0;     
A simple developing move ... if there is anything truly wrong with this move ... then all the rules and principles   
that I teach to my students can be thrown out the window!  {Editor's Note: A couple of the pundits and the  
experts on the Internet condemned this move as inaccurate.}   

     [ Possibly ... Black could try the play:  13...h5!?;   as was played in in the (historic)   
       Reshevsky - Keres contest?  (This game was held to find a replacement for Alekhine ...   
       who died while holding the title.) ]  

 14.Nb3,  (dbl-attack)   {See the diagram ... just below.}    
Garry obviously is intensely interested in simply winning a pawn in this position.   

In only 14 moves ... Garry is a real burglar ... stealing a Pawn ... in broad daylight!  

 ******************** 

 ******************** 

This position is so amusing, it really deserves a diagram here.  

Another interesting item to note ... in the first two games Garry was almost always behind on the clock. Here the times    
clearly favor Kasparov, as White has only used about 15 minutes thus far. The computer, on the other hand, is falling   
further and further behind Garry on the clock. (The box has already consumed 24 minutes, and seems to be taking more   
and more time for every single move.)  

     [ White could also play:  14.Rb1, ("+/=")  {Diagram?}  
       (with a nice edge)  ... 

    ***************************************************************  

       or simply the try of:  14.Be2, ("+/=")  and maintain a slight edge in either case ...   
       Black's game remains very cramped, and it is not really clear how Black can   
       free his game. ]  

 14...Bd6!?;  (Maybe - '?!/?')     
The computer lays a trap that might snare the average "Class C" player ... but is a joke when played vs. Garry.   

Many of the critics were very harsh and vocal in their criticisms of this move. But Black's position is already    
so rotten that he has few really viable and playable alternatives!   

     [ Black had to try something like:   
        >/=  14...Ne8; ('!')  15.Bb2 f5;  16.g3!? g5!;  "<=>"  {Diagram?}   
        and force open some lines on the King-side ... while Garry   
        is completely pre-occupied and also tied up ... on the Queen-side.   
       (But one expert pointed out that Fritz would NEVER play this way ...   
        the computer is exposing its own King!)  ]  

 15.Rb1, ('!')  (very fine)   
White slowly and very methodically increases the pressure ... and he ignores the very silly and obvious trap   
(that was) set by the program here.  

     [ There is no way that Garry was  going to fall for:  
        </=  15.cxd6?? Nxb6;  "-/+"  {Diagram?}    
        and White will lose the Queen.  

       This might work playing speed chess in a park someplace ...   
       but has little value against a player like Garry!!  

   *****************************************************    

       It is also too early to try and capture the b-pawn as yet for White:  
       </=  15.Nxa5?! Nxb6!;  16.Qb4,  {Diagram?}    
       This is pretty much forced.  

            ( But NOT the moves:    
              </= 16.cxb6?? Bxa3; {Diag?} Black must play this first.    
              17.Qxa3!? Qxb6; "=/+" (Maybe - "/+")  {Diagram?}      
               with a definite edge for Black. )    

       16...Be7;  17.Bc1, "~"  {Diagram?}   
       Once again ... this is probably forced for White in this position.  

            ( Very bad for White is:  </=     
               17.Qxb6?? Qxb6;  18.cxb6 Bxa3;  winning ("-/+") for Black.     

               Terrible is:  </=  17.cxb6?? Bxb4;  "-/+"  {Dm?}      
                and White drops the Queen. )     

       17...Nfd7!;  ('!!')  ("=")  {Diagram?}  
        and any advantage White had ... has flown out the window.  ]  

 15...Be7?! ; (Maybe - '?' or even '??')   [Truly pathetic.]   (zzzz)     
Soon White will be threatening to actually capture this piece ... so the box - like the real coward it is -  
retreats  ...  and admits that the idea behind ...Bd6; was unsound.  
(GM Maurice Ashley called this move things like: "stupid," and "completely ridiculous.")   

  *********************  

Of course this move gives Garry something like two or three tempi ...
after this, the computer is really lost. (Unless - of course - Garry makes another blunder!)  

One GM - on chess.fm - related the witty thought that it is good Garry's opponent is a computer!!   
(It cannot feel the embarrassment ... and also the humiliation that any human would!)   

     [ If Black was serious about the move ...Bd6; then the program had to try   
       something like  ...Ne8;  or even the very committal:  15...Bb8;  {Diag?}    
       But the Queenside could be sealed for all eternity!  ]   

 16.Nxa5,   (gulp - yummy!)    
Garry seemed quite pleased with himself as he took this pawn.  

Garry took about 4.5 minutes on this move, mainly making sure that there were no hidden tricks!  

     [ Interesting was:  16.Be2!? ]  

 16...Nb8?!;     (Maybe - '?/??')        
Distinctly inferior. And a loss of even more time, as this piece must now be "re-developed"  at some later stage.  
At this point, the game has become more of a joke than a real chess contest.  

Commentator Paul Hoffman correctly pointed out that the computer has NO long-term planning ability, and is  
therefore reduced to simply shuffling his pieces - moving the pieces back and forth and await the final blow.  

Now both sides had used about 35 minutes each for their moves. (A rough approximation.)    

     [  </=  16...Nxb6??;  17.Rxb6,  "+/-"  

    ***     ***     ***     ***      ***      ***      ***      ***    

        The computer had to try something like:   
        >/=  16...Ne8!?;  17.Bb4! Bg5;  "<=>"  {Diagram?}   
        preparing the King-side Pawn avalanche of ...f5-f4xe3, etc.   

              ( Not  </= 17...f5?;  18.Nxc6!!, "+/-")  

    ***     ***     ***     ***     ***      ***      ***      ***      ***      ***  

        Maybe the box could try:  16...Nh5!?;   {Diagram?}   
        also with the idea of playing  {eventually} the pawn advance of ...f5-f4, etc.  ]   

Unfortunately here, ESPN dropped their continuous TV coverage for women's basketball. 
And I was forced to scramble on the Internet and find some game commentary there. (ICC and http://chess.fm)  
 17.Bb4,    {See the diagram ... just below here.}    
Garry thought about 10 minutes ... (at least) before playing this move.  

 ******************** 

 ******************** 

The idea is an obvious re-organization of the White army ... ... ... Garry wants to protect his pieces, and not leave  
anything hanging. (As he did in the last game.) He also wants to protect the Knight on a5, and work his way OUT  
of the pin - by the Black Rook on a8 - on the a-file in this position.  

     [ Again ... when will White get around to playing the move:  17.Be2!?,  {Diag?}  in this game? ]  

 17...Qd7!?; ('?!')    
Yet another nearly pointless ... 'computer move.'  That Black is having problems here ... 
 is now obvious to even the most casual of observers here.  

      [ Maybe  17...Na6!?;  instead? ]  

A friend sent me a copy of the ... "side-band"  {ESPN} feed ... this is the AUDIO portion of the TV coverage.  
(On cassette tape.)  This helped greatly to get the feel ... of the ... "blow-by-blow"  for this contest.  

 18.Rb2!?,  (Maybe - '!!')    {See the diagram ... just below.}   
The beginning of a brilliant and extremely deep plan by Garry to completely re-organize his position.  

 ******************** 

 ******************** 

Yasser Seirawan did not like this move, but I think it is excellent. It also covers the f2-square ... 
the one real soft point in all of White's position. 

Meanwhile both Franz Morsch and also Armand Rousso called this move,  " ... a real stroke of genius!"  

"This useless-looking move confused most of the commentators, ..."  - ChessBase  

     [ Maybe  18.Be2!?,  "+/="  {Diagram?}  ...  when, when, when? ]  

 18...Qe6;   
At this point I was nearly clueless ... I had no real idea what  Fritz_X3D  was trying to do in this position.  

     [ Maybe  18...Na6!?  ]  

Now Garry gets out of the pin, by moving his Queen off the a-file.  
 19.Qd1 Nfd7!?;    
A glimmer of hope? Will the machine finally do something constructive like push its own f-pawn,   
 in an attempt to open some lines? (Nope.)  

     [ Maybe just  19...Re8!? ]  

Now Kasparov protects his Bishop on b4, giving himself a completely solid - nearly 'bullet proof' - position.  
 20.a3!? Qh6;    
Black gets the Queen to an interesting square - where it actually threatens something. The only problem is   
that this one little maneuver cost the computer program somewhere in the neighborhood of 5-7 tempi!!!  
(Terrific!? Not!!)  

If you were dedicated enough to keep your TV tuned to the  "ESPN News" channel,  you would know  
that the chess team broke in here to give a brief report on the last few moves of the game.  

Seirawan flatly stated that he did not like White's position, that the computer was whipping up a dangerous   
attack. ... HA!  (Was he told to say something like this ... were the producers afraid that the public had lost   
interest in this particular chess game?)   

White now re-posts the d2-Knight on the b3-square ... where it is much more effective than on d2.   
 21.Nb3 Bh4!?;  ('?!')   {See the diagram ... just below.}   
Another brief report ... Black has actually managed to create a real and genuine threat.  
(The move ...QxP/e3+.) 

But this simple problem is very easily solved by one of the greatest players to ever play the game!  

 ******************** 

 ******************** 

"A pathetic one-move threat that ends up wasting more time."   - ChessBase  

     [ Maybe just  21...Nf6!?;  or even   21...Na6;  instead. ]   

 22.Qd2[],    
This move is practically forced for White, especially in his over-all scheme of things. 

     [ Interesting was:   22.g3!?,  {Diag?}  
        but this might give Black targets for pawn-levers to open the game. ]  

 22...Nf6!?;  ('?!/?')     
More of this pointless and nearly useless ...  and almost perpetual re-shuffling of the pieces by Black.   

     [ Maybe Black should try:  >/=  22...f5!?;  {Diagram?}   
        trying to go for the advance, (the Black Pawn) of ...f5-f4? 

                                           ***    

       Maybe even the move:  22...Na6!?;  {Diagram?}   
       would have been better than the way the game was played. ]   

  ******************************  

Garry now shows what all the maneuvering has been about. While the program may not have any (real) idea    
of what the heck it should be doing here, Garry has come up the idea of simply marching his King away from   
any <possible> threats by Black! It will become obvious that the White King ... surrounded by a host of his    
own pieces  ...  will be perfectly safe on the Queen-side. Incredible!   

{GM Kasparov's plan was so savvy and sophisticated that most GM's did not seem to grasp its finer points,   
 at least not right away.  The comments of many of the on-line guys showed that did not really have a clue ...   
 as to what was actually happening in this battle royale.} 
 23.Kd1!! Be6!?;   24.Kc1! Rd8!?;  ('?!' - dubious?)   
Black continues to move his pieces away from the King-side ... and away from the Pawn Break that was the    
box's only chance.  

     [ Why not just:   24...Nbd7!? ]   

 25.Rc2! Nbd7;   26.Kb2,  {See the diagram given ... just below.}   
 White has castled  ... 'by hand' and now has an extremely good position.  

 ******************** 

 ******************** 

Another brief report from the chess team on ESPN.  Its unanimous. 
All the commentators agree Fritz has no clue here ... and that White's edge is quickly becoming overwhelming.  

     [ Interesting was:  26.f3!?, "+/=" ]   

 26...Nf8;    
"Meanwhile ... back on the farm ... "   

The computer continues its endless, and - by now - nearly nauseating maneuvers ... 
and the continual re-arranging of its pieces.  

 ******* 

Garry now begins to shove his QRP right up the left-hand side of the board. Its march is unstoppable, brilliant ... 
and completely decisive.   
 27.a4! Ng6;  28.a5! Ne7;   
Yet another brief report from the ESPN (TV) team of commentators.  

They re-capped the moves, and by now the normally patient P. Hoffman had thrown his hands in the air in almost  
complete despair. (He was obviously disgusted with the fact that the computer could not come up with a plan and   
was content to simply move its Knight back and forth.)   

By this time, I was even becoming bored with the game. I also regretted my decision to tape this nonsense ...   
and the fact I was missing some pretty darn good football games!  

 29.a6!!,  (pawn-sack)   
Garry does not wait, but immediately shoves this Pawn up. (Black must capture.) Meanwhile about a dozen    
different commentators - on various (chess) Internet sites - were calling this move:  "an outright blunder."   

Black can do nothing but watch. The a-pawn will eventually be won back with a decisive advantage for White.   
(But I guess the humans could not 'see' this far ahead.)   

GM Maurice Ashley called this a truly great and 'inspired' sacrifice.   

"Kasparov gives back the pawn temporarily in order to gain a protected passed b-pawn and squares for his   
 pieces. He will now build up his forces for the final assault."  - ChessBase   

     [ Interesting was:  29.Rg1!?,  {Diagram?}    
        a subtle attempt to try and trap Black's Bishop on the h4-square. ]    

Black has no choice but to take. 
(If I were the computer operator, I would have the machine play ...Rd-c8; and after PxP/b7;   
 which forks the Rooks - I could resign and end the machine's misery.)   
 29...bxa6[];  30.Na5! Rdb8;    
Another brief report ... the guys on ESPN is saying this game is now  ... "all Garry Kasparov."  

 31.g3!,   
White pushes Black's pieces back. 

This is a signal that Garry is almost ready to start final operations on the Queen-side here.  

     [ Very interesting was:  31.Rg1!?,  {Diagram?}   
        with the idea of g3 and then h4, trapping the Black Bishop. ]    

 31...Bg5;  32.Bg2!?,  ('!')    
"Getting out of the way of the rook while threatening to win a piece with h4, trapping the bishop."   - ChessBase  

     [ Possible was:  32.Ka1!? ]  

 32...Qg6;   
Black had to give his Bishop on the a5-square an escape route.  

     [ </=  32...Bc8??;  33.h4, "+/-" ]   

 33.Ka1! Kh8?!;  ('?')   
(This)  Looks like a loss of time to me.  
(I also don't see this move ever really helping Black.)  

     [ Probably better was:   33...Bc8!?;  {Diagram?}    
       saving two tempi over the way the game actually went. ]   

 34.Na2 Bd7;   
The box cannot figure out where to put its pieces. (More time is lost.)   

     [ Or  34...Bh6!?;  35.Rb1,  "+/=" ]   

 35.Bc3! Ne8;  36.Nb4,  ('!')   
Another brief report by the ESPN Team of commentators.   

"Total domination," says Seirawan.  

     [ Playable was:   36.Rb2!?,  "+/="  ]   

 36...Kg8;   (urgh)    
Black  ...  'takes back'  his 33rd move.  

White continues to re-arrange his pieces in a completely positive way, in stark contrast to  
Black's nearly useless piece re-shuffling.  
 37.Rb1 Bc8;  38.Ra2 Bh6;  39.Bf1 Qe6;  40.Qd1!? Nf6;  41.Qa4! Bb7;     
Another brief report by the ESPN Team here ... and a brief re-cap of the last few moves.   
(I won't tell you what they said ... by now you should be able to guess.)  

 42.Nxb7,  (Maybe - '!')   {See the diagram ... just below.}   
White uses a 'good' Knight to capture a truly 'bad' Bishop.   

Either this is very bad ... or very good, and the true ... "beginning of the end."   

 ******************** 

 ******************** 

Once again, the picture here is most amusing ...and therefore deserving of a diagram.  

     [ Maybe   42.g4!? ]  

The final phase sees a breakthrough, the result of Garry's careful and well thought out preparations. 
(And more pointless  ... "back-and-forth" moves by Fritz_X3D.)   
 42...Rxb7;  43.Nxa6 Qd7;  44.Qc2!? Kh8!?;  45.Rb3!, '±'  (Really "+/-")   {Diagram?}    
 Black Resigns.  (1 - 0)     

(The game is truly hopeless for Black and the Fritz team, to play on is an insult to Garry Kasparov.)  

     [ After the moves of:   45.Rb3! Nf5;  {Diagram?}   
        It is not clear that any move here is really any good for Black.  

            ( Or   45...Ne8!?;  46.Rba3!? Nc8;  47.Nb4 Rab8?!;  48.Ra8?! Bg5;     
               49.Rxb8 Rxb8;  50.Ra6 Bd8?!;  51.Qa4 Ne7;  52.Ra8 Rxa8!?;       
               53.Qxa8,  "+/-"  - ChessBase.  )        

       46.Ra5 g6;  47.Rba3 Ne8;  48.Nb4 Rxa5;  49.Rxa5 Kg8;  50.Qa4 Ne7;   
       51.Ra8 Bf8!?;  52.Ba6,  ("+/-")  {Diagram?}   
       and it is deadly obvious that White's position is dominating.   
       {White is winning material as well.} ]   

A truly amazing game, this game was taken by Garry ... and he hardly had to engage in any real ...
"hand-to-hand"  dogfights. (Nearly zero big tactics.)   

This game is also a complete victory - in terms of just preparation - for Garry and his (opening) team.  

The computers have a LONG way to go in terms of strategy. 
(Any human could have come up with a better and more incisive plan than the computer.)  

  Copyright (c) A.J. Goldsby I  
  Copyright (c) A.J. Goldsby, 2003, 2004 and 2005.  All rights reserved.  

The original analysis and also the initial text for this game was generated with the program, ChessBase 8.0