GM Vassily Ivanchuk - GM Artur Yusupov 

  (FIDE) Candidates Match, (Tie-breaker, Game # 9.)  
  Brussels, Belgium;  1991.   


This is a game I saw and studied right after it was played. (At that time, I got between five and ten chess magazines every month.) Probably the best initial analysis was done in a newspaper column (by Kavalek?) and the "Inside Chess" magazine. (Now defunct.) 

For a long time, (it appeared that) no one seemed interested in taking a serious look at this game. Before Ken Smith died, a friend of mine was going to publish a book through "Chess Digest" out of Dallas, Texas. It was understood that I was to do at least half of the annotations of the games for this book. This was one of the games that I chose for my collection in the half of the book I was to do - the proposed title was something like: 
 "Titanic Chess Struggles." (All this was nixed after Ken Smith passed away.) 

Since I started my many web pages, I have received an almost countless number of requests to annotate this game. (I got over 30 e-mails about this game in one month alone!) Naturally, I wanted to do this game. But for a long time I was unable to come to a conclusion about the worth of certain lines - especially those published by other chess authors. What helped me was the recent giant strides that the computer programs have made. This enabled me - at least, with an investment of a GREAT DEAL of time - to be able to come to relatively positive conclusions over at least a few of the specific sub-variations. But at the same time - this analysis could NOT be conclusive!!  In many cases, the work only raised as many questions as it answered!! 

   Chess is too complicated to be so easily solved!      

I have been working on this game ...  on-and-off  ...  ever since it was played.

This is - without question - one of the greatest and most exciting chess games ever played.

Please!  Respect my copyright, but also ... enjoy this game!


  If you do not understand the many symbols I utilize when annotating a game, HERE  is a fairly detailed explanation.    

  (If you do not have a chess board handy, I have provided a  RE-PLAY page  for your convenience.)     


 

 

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

GM V. Ivanchuk (2680) - GM A. Yusupov (2585) 
[E67]
 (FIDE) Candidates Match 
 Bruxelles, BEL;  (Rd. # 9),  1991. 

[A.J. Goldsby I]

   The CB medal for this great contest. (gcg_iva-yus_c9-br91medal.gif, 02 KB)

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

One of the most brilliant games of modern times. The attack Yusupov mounts has been studied 
and admired by countless chess players. 

A side note is that there was a virtual revolution going on back home, (In old "Mother Russia.");  
for both parties. {While this game was being played.} So it was amazing that these players could 
even concentrate on the game. 

The final product is one of the most amazing chess games since Anderssen played his "Immortal" 
and "Ever-Green" games. (GM Y. Seirawan, who was present as a commentator, confessed to 
often feeling like a confused amateur during the creation of this super-brilliancy.) 

This was the ninth game of the match, and a tie-breaker as well ... and was played at the time limit 
of: The first 45 moves / in ONE (1) hour!! (Then 15 minutes for every 20 moves thereafter.) 

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

The ratings are very close to exact and come from a CB chart showing the actual FIDE ratings. 

Jeff Sonas gives the following ratings for these two players: GM Vassily Ivanchuk - 2765 
(Informant - 2735)  GM Artur Yusupov - (No rating.)  (Informant - 2625)   

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

1.c4,  {Diagram?}   
The English Opening is certainly OK for White, and probably leads to a solid edge for the 
first player.

     [ Ivanchuk is an extremely versatile player, who could open with any opening. But the 
        database shows that he currently favors:  1.d4, {Diag?} as his main opening move. ]  

 

Black responds with ...e5. Note how both players consistently develop their pieces 
during the opening. 
1...e5!?;  {Diagram?}    
This is one of Karpov's favorite lines. 

The only problem with this move is that White can often play a Sicilian ... 
but with an extra tempo because the colors are reversed. 

     [ Black could play the move:  1...Nf6!?{Diagram?} 
       which is also a popular response. 
       (With this move the second player is offering to play one 
         of the many different lines of various "Indian" openings.)  ]   

 

2.g3!?,  {Diagram?}    
While this is a little unusual, it is not bad. Yusupov has a reputation as being very 
well-versed in opening theory, and Ivanchuk was probably trying to get him out 
of his normal lines. 

  Also - in several modern lines of the English - Black plays a very early ...Bb4; and usually   
   gets a good game. Ivanchuk obviously wishes to avoid this approach by Yusupov.     

*******

     [ The main lines probably are:  2.Nc3 Nf63.Nf3 Nc64.g3, "+/="  {Diagram?}  
        with a good game for White.  

        [ See MCO-14;  page # 666. ] ]    

 

2...d6;   3.Bg2 g6!?;   {See the diagram given - just below.}     
It could be a trifle early for this fianchetto. 

 

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

   The position after Black's third move. {...g6}  (gcg_iva-yus_br91-pos1.gif, 45)

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

 

Some authors and writers have criticized this move for Black. But since it does not lose 
by force  ...  and really just amounts to a simple transposition  ...  I don't see what 
the "hub-bub" is all about. 

*******

     [ Another possible line is:  3...Nc64.Nc3 g65.e3 Bg76.Nge2!?, "+/="  {Diag?}  
        when White is only a tiny bit better in this position.  

        According to ChessBase's on-line database, there are over 1000 examples of this position! 
        The first example occurred in the 1930's. (There is also a game of Botvinnik's from that 
        period as well.)  

        The most meaningful (relatively current) example I could find was the encounter:  
        GM M. Krasenkow - GM M. Sadler;  Malmo, (Sweden); 1995.  
        (The game was drawn in under 30 moves - I don't think either player really 
         wanted a fight that day.) ]    

 

Both sides continue to develop sensibly. 
4.d4! Nd7;  5.Nc3 Bg7;  6.Nf3 Ngf6;  7.0-0 0-0;    {See the diagram, below.}      
Both sides have developed fairly consistently and also in a pretty straight-forward manner. 
(White is probably a tad better in this position.)

 

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

   Black just castled ... the opening is almost over.  (gcg_iva-yus_br91-pos2.gif, 45)

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

 

To be honest, I think we have transposed out of the English and into the Fianchetto Variation 
of "The King's Indian Defense." (See MCO-14, page # 608.) 

The Mammoth Book says this was (probably) a surprise to Ivanchuk, as Yusupov hardly ever 
uses this particular opening. (Although I found quite a few examples in the database, especially 
from Yusupov's early years.) 

 

8.Qc2!?,  {Diagram?}    
Certainly respectable, but not the main line anymore. 
(GM A. Soltis recommends that Black now exchange on d4 and he says that the second player 
 has a satisfactory game.)  

Being that this is a "quick" game, Ivanchuk (understandably) wants to get Yusupov out of book 
as quickly possible!  

*******

     [ The main line is:  (>/=)   8.e4 c69.h3 exd4!?{Diagram?}  
        Black can play nearly six different moves here, according to theory.  
         (...Re8; ...Qb6; ...Qa5; ...a6;  etc.)   

       And now - according to MCO - the most often played moves are:  
       10.Nxd4 Re811.Rb1 a512.Re1 Nc513.b3 Nh5!?14.Be3 Qe7; 
       15.Qd2 Qf8!?16.Rbd1 Nf6{Diagram?}  
       The end of the column.

       17.Qc2 Nfd718.f4, "+/="  {Diagram?}  
       White seems to be a fuzz better in this position. 

       "As often happens in this variation, White has a space advantage." 
        - GM Nick de Firmian. (MCO) 

       GM. A. Wojtkiewicz - BjarnssonNew York Open, 1994. 

       [ See MCO-14, page # 608; column # 61, and note # (b.). ]  ]   

*******

 

8...Re8!?;  {See the diagram, just below.}    
This is one of the main lines here.  (The other main line here is ...c6.)  

 

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

   Black just placed a Rook on a center file.  (gcg_iva-yus_br91-pos3.gif, 45)

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

 

Several authors - most notably GM Andrew Soltis - have criticized this move and even awarded 
dubious or question marks. But since this is a theoretical  MAIN LINE ... and is analyzed in about 
a dozen books that I own on the "King's Indian Defence," {opening}  ... it is VERY hard to agree 
with Soltis ... or even understand why he is so critical.  

 

********

     [ Soltis says that Yusupov should play the move:   8...exd4!; "~"  {Diagram?}  
       and Black has already equalized. 
        (BUT ... neither theory, nor computer analysis supports this assessment!)     

       NOTE:  GM Yasser Seirawan also states that  ...exd4!  is the best move 
       in this position for Black. (But statistically speaking, Black's recent results 
       after the exchange in the center, have been dismal. See the note after Black's 
       next move, ...c6.);   

***

       Black could also play ...P-QB3 (...c6); in this position: 
       8...c69.e4!? Qe710.Rd1!? exd4!?11.Nxd4 Re8!?12.b3 Nc5;  
       13.f3 Nfd714.Rb1 Ne515.Nce2!? a5;  "~"  {Diagram?}  
       Black is probably close to equality in this position. 
       (Play has transposed to a well-known line.) 

       See the game:  
       GM V. Salov - GM G. Kasparov;  Super-GM Tournament / 
       Linares, (ESP);  1991.  (The game eventually ended in a draw 
       in around 35 moves.) ]   

*******

{Editorial note: To understand why Rd1 is so good ... you would really have to play 
  through some of the lines and variations in a few of the sources listed in the bibliography. 
  Or ... you can simply take my word for it.} 

 

9.Rd1! c6;  ('!?')  {Diagram?}    
A common idea in these lines, Black wants to play ...Qe7; but does not want to worry 
about White playing Nd5. The pawn on c6 keeps White's pieces off d5 and also b5. And 
now Black has the option of possibly playing ...Qb6 or ...Qa5. 

GM Yasser Seirawan gives Black's ninth move an exclam here. 

    '!' - GM Yasser Seirawan.    

 

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

     [ More often played is:  
        9...e4!?10.Ng5 e3!11.f4!?{Diagram?}  
       Given as FORCED by many books, but ... is it? Really? 
       (MCO does not give any comment at this point.)  

***

           ( After the moves:  (>/=)  = 11.Bxe3!? Rxe3!;  12.fxe3 Ng4!;  13.Nf3!?,    
             13...Nxe3;  14.Qd3 Nxd1;  15.Rxd1 c6;  "="  {Diagram?} 
      
              the game appears to be very finely balanced. )      

***

       11...h612.Nge4 Nxe413.Bxe4!? Nf614.Bg2 Ng415.d5 Nf2;  
       16.Rf1 Bf5{Diagram?}  
       The end of the column.

       17.Qb3 Nh3+18.Kh1 Nf2+19.Kg1 Qc820.Nd1 Nxd1;  
       21.Qxd1 Be4;  "=/+"  {Diagram?}      
        ... "and White found himself seriously cramped by the Black Pawn on e3."  
         -  GM Nick de Firmian.

       Maiwald - KetevanGermany, 1994.     

       [ See MCO-14;  page # 608,  col. # 66, and also note # (n.). ];   

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

       Here ... now that White's Rook is on the d-file, I do not think the exchange in the 
       center is wise:  9...exd4!?; ('?!')  10.Nxd4 a6!?11.b3 Rb812.Bb2, "+/="  {Diag?}  
       White has a very clear advantage in this position, and according to the database - 
       (recently)  the first player has scored close to 75% ...  from this position!!  

       (A nice game is:)  
       S. Mamedyarov (2585) - Zia Rahman (2521)ADCF Masters Tourney/2003.   
       {White won a fairly long game, but there he used 12.h3 instead of Bb2.} ]   

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

 

10.b3!?,    {See the diagram, just below.}     
A modern move. More and more,  players are questioning the older and accepted ways of playing 
the position or opening, and looking to strike out on completely new and different paths.

 

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

   White opts for a fianchetto ... instead of e4.  (gcg_iva-yus_br91-pos4.gif, 45)

************************* 

 

Seirawan like this so much, he awards this move both praise - and an exclam.  

   '!' - GM Yasser Seirawan.    

 

     [ More often played is:  10.e4 Qe711.b3 exd412.Nxd4 Nc513.f3 a5; "~"  {D?}    
       and this position is either just a tiny bit better for White ... or Black has come very 
       close to equalizing. 

       See the encounter:   GM Josif Dorfman (2590) - GM Joe Gallagher (2470)   
       (City)  Championship Tournament/Paris, (FRA); 1990.    
       (White won a very long game, but I am sure the opening was not at fault!) ]   

 

10...Qe7;  11.Ba3!?,  {See the diagram, just below.}    
Ivanchuk does not wish to play the main line ... and is obviously trying to draw Black into 
a position where the Bishop on a3 is a significant menace to Black.  

 

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

   With the Bishop on a3, White sets up a nasty pin. But will this piece be well-posted ... or simply out of the game?  (gcg_iva-yus_br91-pos5.gif, 45)

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

 

(Was this a prepared weapon?) 

  '?!' - GM Yasser Seirawan.   (I don't agree with this.)  

 

*******

     [ Seirawan says that it is better for White to play:  (>/=)  
       11.e4! exd412.Nxd4,  "+/="  {Diagram?}  
        and White maintains a clear edge. 
        (This is a transposition back to one of the "book" lines ...  
         see the note after Black's eighth move.);  

***

         ECO  recommends:   11.e3!?,  "="  {Diagram?}    
         which leads to equality, and poses no real problems for Black. ]   

*******

 

Now Yusupov realizes that with White's Bishop on a3 and the White Rook on d1, 
he cannot allow the center to be opened. (Play a few lines with your computer, 
and this will quickly become obvious.) 

11...e4!;  {Diagram?}  
"A double-edged decision."  - Iakov Damsky.  

   '!' - GM Yasser Seirawan.   

 

     [ Black can also exchange here:  11...exd4!?12.Rxd4!? Nc5 
     
  13.Rad1 Bf514.Qc1 Nfe4;  "="  {Diagram?}     
       One author claims that Black is (much) better in this position ...  
        it looks very close to level to me.  {A.J.G.} ]   

 

12.Ng5,   {See the diagram ... just below.}    
This could be forced for White.  

 

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

   White plays Knight to g5, placing a tremendous pressure on e4.  (gcg_iva-yus_br91-pos6.gif, 45)

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

 

The position is very rapidly building to a crescendo.  

     [ </=  12.Nd2!? e3; "<=>" ]  

 

12...e3;  {Box.}  {Diagram?}    
Otherwise it appears that Black will lose this pawn.  

     [ Of course NOT:   </=  12...d5; (????) 13.Bxe7,  ("+/-")  {Diagram?} 
        and Black has dropped his Queen. ]  

 

13.f4!?,   {See the diagram just below.}   
Because of the tactics, this is nearly forced for White.  

 

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

   White just played f4 - so what now?  (gcg_iva-yus_br91-pos7.gif, 45)

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

 

Opening theory {also} recommends f3 here, but I am sure no good GM has ever examined 
these lines critically. (The games and analysis are absolutely putrid.)  

 

     [ GM Yasser Seirawan  says White MUST play:  13.f3!?,  "~"  {Diagram?}   
        in this position. I am not arguing with him, or saying that he is right or wrong. 
        BUT f3 LOOKS VERY UGLY TO ME!! {A.J.G.} 

        (Yusupov's analysis seems to indicate that Black will {eventually} get an advantage. 
          Damsky also gives a nice line here for Black.)  

       Now one writer gives the following line:     
       13...Nf814.Nge4!? Bf5!?; (?!)  15.Qc1 Bxe4!?16.Nxe4 Nxe4;    
       17.fxe4 c518.e5 Ne6; ('!?')  19.exd6!? Qxd6{Diagram?}     
       While this is forced, White's next move looks to be very doubtful.   

       20.Bxb7?! Nxd4; "=/+"  {Diagram?}   
        ... "with sufficient counterplay for Black."   -  Iakov Damsky.  
       ( This is Damsky's analysis  ...  I will simply state that  almost ANYONE  with a    
         strong computer program could probably improve upon this line!! {A.J.G.} );     

********

       Of course not:  13.fxe3?? Qxe3+14.Kh1 Qxg5{Diagram?}        
        and Black wins a piece. ("-/+")  ]     

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

 

Now ... as if by magic - Black's pieces begin to spring to life. 
13...Nf8;  14.b4!? Bf5;  15.Qb3 h6;  16.Nf3 Ng4;  (Maybe - '!')  {Diagram?}     
This is obviously nice for Black, but he had other very playable options here as well.  

   '!' - GM Yasser Seirawan.   

*******

Just a note for those players who do not have a lot of experience with this line, or have never 
studied the games of this particular opening. (ANY "King's Indian Defence" line!) In this opening, 
White almost always has a pawn on d5 and plays on the Queen-side. And while the structure here 
is a little different, the same principles still apply. (White looks to one side of the board to create 
play, Black will almost always seek his play on the King-side.)  

*******

     [ Maybe (possibly >/=)   16...a6!?{Diagram?}   first? 
       (To open the a-file for Black, especially if White plays the way 
        he does in the game.) ]   

*******

 

GM Yasser Seirawan  considers both White's and Black's next plays to be 
worthy of an exclamation mark.  
17.b5! g5!;  {Diagram?}     
Black obviously intends to open lines that will eventually lead to a violent attack on 
White's King. (Or at least this is what A. Yusupov is hoping for!) 

GM Larry Christiansen - in his very good book on tactics and attacking the King - lets the 
moves up to this point pass without comment. He then (now) gives a VERY long discussion 
of the 'landscape'  and the general ideas here. Basically he says that White might be better, 
but Black MUST seek his chances on the King-side. 

    '!' - GM Andy Soltis.  '!' - FM Graham Burgess.      

 

*****

     [ Interesting was:  17...Nf2!?; "=/+"  {Diagram?}  
       I think Black has the slightly better chances in this position, but I would      
       also be the first to admit that I could be mistaken! ]      

*****

 

18.bxc6 bxc6;  19.Ne5!?,  (Maybe - '!')   {See the diagram below.}   
It takes nerves of steel to play a move like this one, at least according 
to GM Yasser Seirawan.  

  "There is hardly any other reasonable move." (here) - GM Larry Christiansen.     
    (This is illuminating, because a few annotators criticized this try for White!)   

 

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

   White just launched a Knight - spectacularly - into the e5-square ...  but why?  (gcg_iva-yus_br91-pos8.gif, 45)

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

 

  This move, (19.Ne5); is also the FIRST choice of virtually EVERY SINGLE STRONG    
   PROGRAM ...  that I used to test and analyze this game!!!    

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

     [ Also possible was:  19.Qa4!?,  "+/="  {Diagram?}  
        and maybe White is a little better. 
        (A big emphasis on the word, "maybe" here!);  

                                          *******   

        White could also try:  (>/=)  19.fxg5 hxg520.Rf1, "="  {Diag?}    
         and the position appears pretty close to being level. 
         (Yusupov felt this might have been an improvement over the game.) 

        ( The well-known writer and trainer  (M. Dvoretsky)  has suggested that      
          White instead play Ne5 here.  But this looks like it has not been critically examined:     
           </=  20.Ne5!? Nxe5!;  21.dxe5 Rab8!;  22.Qa4 Bxe5!;  23.Rac1 Bxc3!;     
          24.Rxc3 c5; "~"  ('!')  {Diagram?} and many hours of computer-assisted analysis      
          strongly indicates Black is OK ...  maybe even better. )  ]  

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

 

Black continues trying to make inroads into White's King-side. 
19...gxf4;  (Maybe - '!')   {See the diagram - just below.}      
This is probably best for Black. 

 

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

   Black just captured on f4  ... ... ...  he is trying to expose the White King.  (gcg_iva-yus_br91-pos9.gif, 45 KB)

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

 

Further eroding White's King-side.  
   '!' - GM Larry Christiansen.  ('!' - GM Andrew Soltis.)     

 

*******

     [ Certainly NOT:  </=  19...Nxe5?20.fxe5!, '±'  {Diagram?}    
        and Black will lose his d-pawn.    

***

        Also inadvisable for Black was:  </=  19...Rac8?!;   20.fxg5 Qxg5;      
        21.Nxc6, '±'  (Maybe "+/-")  {Diagram?}  and White has a HUGE edge.  ]    

*******

 

20.Nxc6,  {Diagram?}     
"White collects the harvest."  - GM Yasser Seirawan.  

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

     [ After the moves:  
        20.Bxc6!? fxg3!21.Bxa8? gxh2+!22.Kg2 Nf2!; "/+"  {Diagram?}  
        Black is clearly better. 
        (Black may have a winning attack against the badly exposed White King!);  

*******

       Seirawan points out the line: (</=)  20.gxf4?! Bxe5!21.fxe5 
        21...Qh4
;  "/+"  {Diagram?}  and Black is close to winning. ]       

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

 

20...Qg5!;   {See the diagram ... immediately below.}     
Black abandons material concerns ...  and heads for fame and glory.  

 

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

   Black's plays  ...Qg5.  I think he has ambitions on the King-side.  (gcg_iva-yus_br91-pos10.gif, 45 KB)

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

 

   I remember when I first went over this heroic struggle ...    
   I had the very clear impression that White was winning here.(!)      

     [ </=  20...Qe6!?21.Nd5, '±'  ]     

 

GM Yasser Seirawan says both of the upcoming moves by BOTH of the players 
merit praise.  ('!' - GM Yasser Seirawan.)  
 21.Bxd6! Ng6!;  {Diagram?}    
This is best says Seirawan. 

   '!' - GM Yasser Seirawan.      

"Yusupov piles up the pressure ..."  (on the King-side) 
   - GM Larry Christiansen.  (Christiansen also awards an exclam to this move.)  

     [ Interesting was:  21...Nxh2!?; "~"  - GM A. Yusupov    
       (The 'Mammoth Book' strongly questions this and gives a line      
        that appears to be winning for White.) ]     

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

 

22.Nd5!, '±'  ("+/")  {Diagram?}       
White looks to be clearly on top - in this particular position.  

"The only chance. The centralized Knight both attacks and defends."  
  - GM Larry Christiansen.   

(Several different authors award this move an exclam here.)

 

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

     [ Also interesting was:  22.c5!?,  "+/="  {Diagram?}   
        and White is still better.  

*******

       Almost no one - that I am aware of - has pointed out the try:     
        22.Nb8!?,  "+/="  (Maybe - '±')  {Diagram?}    
        when White certainly looks to be better. (Now please understand,   
        I am NOT saying this is better;  just a very interesting alternative that   
        no commentator has mentioned.)    

        {My original version of this game gives an extremely detailed analysis  
          line here ...  which ends in a win for White.} ]      

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

 

22...Qh5;  ('!?')   {See the diagram, just below.}       
The natural move here for Black.  

 

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

   Black just played ...Qh5; - - - with just a little threat to White's h2-square.  (gcg_iva-yus_br91-pos11.gif, 45 KB)

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

 

Seirawan harshly criticizes this, but I am not sure his comments are completely justified. 
(It is always easy to come up with improvements when you are not under tension, a time 
 constraint, and can relax and move the pieces in your study. GM Seirawan even concedes 
 this point in his excellent book!)  

Both GM's Yusupov and Christiansen seem to feel that  ...Nxh2;  is much better than ...Qh5
But I have found some (big) holes in both of these GM's analysis. So I would have to say that  
their work does NOT support their conclusions!! But I will also freely admit that in a game as 
complicated as this one, the final word on this decision has yet to be reached!  

   '!?' - GM John Nunn.  (BCM)   

 

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

    GM Yasser Seirawan  says Black should instead play the line:  (>/=)   
        22...fxg3!23.Bxg3 h5!24.Kh1!? h4!25.Be1[] Nf4!;  "<=>"  {Diag?}       
        and Black has the initiative and appears to making progress on the left-hand 
        side of the board. 
        (Yusupov rejected this because the move ... fxg3;  appears to allow his opponent's 
          Bishop to get back to the King-side and help defend the White King!) 

*******

       BTW - I am not at all 100% convinced this line is best for Black. 
       Days of computer-assisted analysis has yet to find anything even remotely resembling 
       a forced win!!   

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

       Black could also (again) try:   
       </=  22...Nxh2!?('!!')  {Diagram?}  
       The main idea of:  - GM Arthur Yusupov.  

       [ GM Larry Christiansen likes this move so much that he awards 
          it an exclam. ('!') ]     

**********

       Now the following is a fairly large improvement over Yusupov's main 
       line of analysis:  

       23.Nxf4!?(Maybe - '?!')  {Diagram?}      
       The only move considered by the analyst in this position.  
       (But taking with the Bishop on f4 seems to win for WHITE! 
        See the variations just below.)  

*******

           ( Or perhaps:  >/=  23.Bxf4!?, ('!')  "+/="  Maybe  -  '±'  {A.J.G.} 
             (I think this could be best ...  but most of the annotators virtually      
              ignore this move for White!!);      

***

             White should NOT play:  </=  23.Kxh2? Qxg3+;  24.Kh1 Qh4+;   
             25.Kg1 Qf2+;  26.Kh1 Qxe2;  "~"  {Diagram?}    
             when Black has a very strong (winning?) attack ... despite his material        
             deficit. - GM Larry Christiansen.  (A VERY wild line!);   

***

             Or White could play:  23.gxf4, {unclear?} {Diagram?}  
              with a very unbalanced position. )     

*******

       23...Qxg3{Diagram?}       
       This now looks to be best.   

           ( Or   23...Nxf4!?;  24.Bxf4 Qh5; "~" )     

       24.Nxg6!?>/=  24...Qxg6; ('!')  {Diagram?}      
       This has to be the correct move.      

*******

           ( Instead  GM A. Yusupov  only gives the following continuation:       
              </=  24...Qxd6??25.Nge7+! Rxe7[]26.c5?? Nf3+!27.exf3!?        
              27...Qf6??{Diagram?} and White is (much) better. ('±')   ( - ChessBase.)      

              {The VERY simple move of:  27...Qxc6;  wins  ("-/+")  for Black!}     

             This line is  SO BAD ... it looks like the result of a mis-print ...  
             or some kind of foul-up during the translation of GM's A. Yusupov's notes! )     

*******

       25.Kxh2{Box?}  {Diagram?}     
       The computer says this is forced.   

*******

           ( </=  25.Bxh2?? Be4; ("-/+") {Diagram?}     
              and it looks like White will be mated in three or four moves.  

***

             Also of no help to White is: </=  25.Ne7+? Rxe7;  26.Bxe7?? Be4;  ("-/+")       
              and once more, White will be check-mated quickly. )     

*******

       25...Qxd6+26.Kh1 Be4;  "/+"  {Diagram?}       
        and Black appears to be just about winning, ("-/+"); in this position.  {A.J.G.}  

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

       Now it could be that ...Nxh2 is better than ...Qh5. But as this very detailed 
       analysis demonstrated, there is still much unexplored territory before anyone 
       could say for certain!  

       (NOTE:  In this version, the analysis of ...Nxh2 has been reduced to just a  
        few lines! The original version is MUCH more detailed.)  ]        

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

 

White's next move is virtually forced. 
(GM Yasser Seirawan gives White's 23rd move an exclam, but that looks to be 
 rather superfluous to me.) 
23.h4,  {See the diagram just below.}     
This is not as bad as it looks  ...  White's strongly centralized pieces may dominate 
the board and the play that ensues. 

   '!' - GM Yasser Seirawan.     

 

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

   White just pushed his Pawn to King's-Rook-Four. How does Black proceed from this position?  (gcg_iva-yus_br91-pos12.gif, 45 KB)

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

 

An incredible position has resulted ...  where Black is involved in a ... 
"do-or-die" ... all-out effort on the King-side.  

*******

     [ </=  23.h3?! Nf2!; "/+" ]   

 

23...Nxh4!?;  ('!' / '!!')   {Diagram?}   
This looks like a tremendously brave and daring sacrifice to me - especially considering the 
situation. (The winner moves on in the WCS Candidate Matches and might eventually face the 
World Champion ... in a BIG-money match! The loser goes home.) Another factor to consider 
was the extremely abbreviated time control this game was played under.  

   '?' - GM Yasser Seirawan.    ('?!' - GM Larry Christiansen.)     

GM Yasser Seirawan criticizes this move and says that >/= 23...fxg3!;  was MUCH better. 
("Black gets a clear advantage without the sacrifice.") Yet analysis does NOT bear this out. 

In fact - my analysis seems to indicate that Black SHOULD try this sack on h4  ...  
I strongly suspect that without this sacrifice, White will find a way to defend ... 
with his surplus of pawns ...  and {eventually} win the game!!  
(The 'Mammoth Book' gives the story that this move was condemned by the group of 
  assembled annotators.  Big deal!  I am reminded of the time about 30-35 TITLED  
  players thought Karpov was winning ...  (a WCM game)  ... until Kasparov finally 
  unleashed a torrent of moves that led - practically by force - to mate.)  

*******

Addendum: I must be a poor potzer who cannot understand anything about chess. 
GM Larry Christiansen covers about the same analytical ground as GM Yasser Seirawan. 
He then seems to conclude that - in the main line after 23...fxg3; - that White is better ...    
refuting what some previous analysts thought about this game. But then he goes on to say    
that he must award Black's twenty- third move a "dubious" appellation.     

   WHAT DID I MISS???!!!????!!?      

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

     [  Variation # 23B1.)   

        After the continuation of: (>/=)   = 23...fxg3{Diagram?}      
      &n