Skype coaching session, 05 October 2014. 

The following game was examined, the game 53 moves) is given with annotations and then again without annotations. 

Paul Benson. 

* * * 

Annotated game. 

White: Keres. 
Black: Fischer. 
Event: Candidates Tournament, Bled, 1959. 
Result: 0-1 in 53 moves. 
Opening: Sicilian Defence, Najdorf Variation, B99. 

| Note: Annotations in this game have been supplemented with analysis from game 14 of My 60 Memorable Games, by R. J. Fischer. || 

1. e4 c5 
2. Nf3 d6 
3. d4 cxd4 
4. Nxd4 Nf6 
5. Nc3 a6 

| Controlling the b5 square, so a queen can sit on c7 safely, unless white is prepared to start sacrificing material either on the b5 square itself, or more likely to give up a piece for 2 pawns in a tactic involving wiping out the black queenside pawn formation, after black has expanded with, pawn b5. The Najdorf Variation is a known favourite of white, a consistent repertoire can leave a player open to receiving specially prepared lines created during home analysis. || 

6. Bg5 

| More popular today is, 6. Be3. Whith the white c1 bishop on the g5 square, black is permitted the option of trying for the very sharp Poisoned Pawn Variation, though white can sidestep this if desired. || 

6. ... e6 
7. f4 Be7 

| The Poisoned PawnVariation is initiated with, 7. ... Qb6, it is recommended to study hard before placing this variation in the repertoire. White can prevent the whole idea with, 8. a3, since the blundering, 8. ... Qxb2 9. Na4, traps the black queen, which can then only give itself up for a rook. || 

8. Qf3 Qc7 
9. O-O-O Nbd7 
10. Be2 

| This move was new to theory in 1959. It is believed to have been dreamt up during a fishing trip in anticipation of black arriving at this fairly predictable position. A new idea is about to be unleashed upon a unsuspecting 16 year old. White is fully prepared as one could be back in those days, by working hard with a board and pieces flying around, assisted by a fellow Master. Black, totally unprepared, will have to analyse with the clock ticking towards flagfall. More popular here is, 10. g4, when white will try to open up the black kingside / centre with pawn breaks, even being prepared to let a pawn go in some cases. Play is again sharp and careful preparation is advised. || 

10. ... b5 

| Just what white wanted, let the fun begin! || 

11. Bxf6 

| The first of a set of awkward decisions for black begins. In general, offering an opponent several ways to recapture is a good way to test their tactical and positional understanding, unless of course they possess a lucky 3-sided coin. || 

11. ... Nxf6 

| The correct recapture. If, 11. ... Bxf6, black is allowing a neat trick in, 12. Bxb5 axb5 13. Ndxb5 Qa5 14. Nxd6+, all because the black dark square bishop has relocated to the f6 square. Instead, 11. ... gxf6 12. Qh5, gives black problems, the threat is, Nxe6, as the black f7 pawn is pinned. Attempting to escape with, 12. ... O-O, falls for, 13. Rd3, intending either, Rh3, or, Rg3+, depending on black's 13th move, black is in trouble. Instead, black can protect the e6 pawn with, 12. ... Nb6, but white plays, 13. a3, after which there is a strong threat for white with, 14. f5. || 

12. e5 

| Instead, 12. a3 Rb8, black has counterplay. || 

12. ... Bb7 

| This is the precise position the Russians analysed in the legendary fishing trip. They will have spent more time at home trying to ensure that white does not lose. Once satisfied they can achieve a slight gain in all plausible variations, then the conclusion must be that it is playable. Black is going to go through a few psychological inner battles, as well as having to analyse and reject so much during live play. || 

13. exf6 

| Yes, white has left the queen en prise. Black has no choice, either capture of the white f6 pawn will leave black a piece down, not so much a gift but more of a foist of a queen. Instead, 13. Qg3 dxe5 14. fxe5 Nd7 15. Qxg7 Qxe5, black seems better here. || 

13. ... Bxf3 
14. Bxf3 

| So far white has gained 2 minor pieces for the queen, but black has the a8 rook and the e7 bishop both under attack, one must be let go, but which? || 

14. ... Bxf6 

| Another test passed by black. It is the a8 rook which must be let go. Instead, 14. ... Rb8 15. fxe7, creates problems for black. If, 15. ... Kxe7 16. Nf5+, when capturing this knight allows a royal fork with, Nd5+. So, the king must move with, 16. ... Kf8, but now, 17. Nxd6, gives white good play. Instead after, 15. fxe7 Qxe7 16. Nc6, white wins an exchange. Black suffers similarly if, 14. ... Rc8, is tried. One idea is, 15. fxe7 Qxe7 16. Nf5, threatening both the black queen and, Nxd6+, winning an exchange. If, 16. ... exf5 17. Rhe1, wins the black queen for a rook, and white has a material plus. || 

15. Bxa8 d5 

| Trapping the white bishop. Some black players have chosen to eliminate a dangerous attacker here with the immediate, 15. ... Bxd4. Results after, 15. ... d5, are mixed since 1959, until a curious game from 4 years ago seems to throw new light on the entire queen sacrifice plan. The game Rudd - Neverov, Hastings 2010 continued with black ignoring the bishop with, 15. ... Qc4. Black is not giving white the option to break up the centre and expose his king for a piece. Moreover, it seems that white must lose material in less favourable circumstances. The white d4 knight is doubly attacked, if, 16. Nde2 Bxc3 17. Nxc3 Qxf4+ 18. Kb1 O-O, and while the material using the pawn-count method is equal, black does not have to untangle anymore. The Hastings game continued: 16. Ne4 Qxa2 17. Nxd6+ Kd7 18. Bf3 Qa1+ 19. Kd2 Qxb2 0-1. Improvements for white after, 16. Ne4, are not obvious here, black seems to either have a strong queen infiltration or wins material under more favourable cir! cumstances than previously known. If white tries, 16. Bc6+, 
aiming to disrupt development, 16. ... Kf8, seems sensible. Black might try a riskier, 16. ... Ke7, when, 17. Nf5+ exf5 18. Rhe1+ Kd8, and the active white c6 bishop must retreat. So after, 16. ... Kf8 17. Nce2 e5, and white seems to be losing a piece, the c6 bishop is unprotected if the d4 knight moves. Warning: The analysis given here concerning, 15. ... Qc4, is my own effort, no computer checking has occurred, please analyse carefully yourself before trusting it in a live game. Over to white now to find an answer to, 15. ... Qc4. || 

16. Bxd5 

| White accepts that the best here is to give up the trapped bishop in order to establish the remaining white forces in the centre while disrupting the black development. || 

16. ... Bxd4 

| Black eliminates a dangerous attacker. Black should decline the greedy, 16. ... Qxf4+ 17. Kb1 Bxd4, which is not actually winning a piece as, 18. Bc6+ Ke7 19. Ne2, forks the black queen and bishop, which will be followed with, 20. Nxd4, defending the white c6 bishop. || 

17. Rxd4 exd5 
18. Nxd5 

| Instead, 18. Re1+ Kf8 19. Re5, gives white a comfortable position. The black d5 pawn is not going anywhere, there is no immediate need to capture with the knight, instead, Rexd5, seems a good alternative method for white to win the black d5 pawn. || 

18. ... Qc5 

| Placing the queen on the g1-a7 diagonal, so if the white d4 rook moves the queen will have options to enter the white kingside. || 

19. Re1+ Kf8 
20. c3 

| Doubling rooks with, 20. Rde4, threatening a mate on the e8 square is tempting, but simply, 20. ... g6, is fine for black. White prefers to keep rooks on 2 central files rather than doubling down 1 file. This formation allows for some interesting ideas for white later on. || 

20. ... h5 

| Perhaps something missed by the Russian preparation? Black will achieve both king safety and rook development. The more obvious, 20. ... g6, lets white establish a good grip on the kingside with, 21. g4 Kg7 22. g5. Black might appear to be fighting with, 22. ... Rd8, but, 23. Nf6, simplifies matters. If now, 23. ... Rxd4 24. Re8, threatens, Rg8+ mate, and the only sensible defence is, 24. ... Rd8. Now, 25. Rxd8, keeps up the mate threat, so, 25. ... Qe3+, is necessary and the game finishes with a perpetual check. The black game move indicates a willingness to fight for more than a relatively early draw. || 

21. f5 

| Designed to deter black from playing, 21. ... g6, unravelling. Up until this point white has been steadily building a good position, without any risk whatsoever according to prepared analysis. Perhaps white is beginning to want to try for more than the position has to offer. The advance of this pawn to f5 is not in itself weak, but hinting at the wrong mind set. It is time for white to pull back and accept black is out of danger. So, simply sitting on the game with, 21. Re2, was safe, and, 21. Re5, was also playable. || 

21. ... Rh6 

| Black is almost untangled, it is still slightly favourable for white but that is all. || 

22. f6 

| A critical move. White has been testing black since move 10. The queen sacrifice has neither succeeded nor failed. However, white now wants more from the position than can be justified. Up until now white has been in little danger, but now unnecessary risks are being taken. || 

22. ... gxf6 
23. Nf4 

| This knight is covering the g6 square, preventing both, Rg6, and also assisting a potential white mating net with 2 rooks on the 8th rank, but such a threat of mate is very optimistic. || 

23. ... h4 

| Black places some pressure on the white kingside while waiting to discover if white is playing for mate or to hold what he has. || 

24. Rd8+ 

| Another critical decision. By pursuing this active plan, black is also given activity, and an active queen is difficult to contain at the best of times. Instead, 24. Re2, is logical. || 

24. ... Kg7 
25. Ree8 Qg1+ 

| Now black is fighting for the first time in the game. If white has to start retreating then this indicates the whole concept beginning with, pawn f5, and definitely, pawn f6, was inadvisable. || 

26. Kd2 Qf2+ 
27. Ne2 

| This retreat disrupts the attempted mating pattern, and also offers black use of the g6 square for the rook, the initiative has passed to black. || 

27. ... Rg6 
28. g3 f5 

| Now the black king has a flight square, or for the more optimistic thinkers, a means of entry to some central activity. || 

29. Rg8+ Kf6 
30. Rxg6+ 

| White should avoid the tempting, 30. Rd6+ Ke7 31. Rgxg6 fxg6 32. Rxg6 h3, and white has no sensible means to prevent the black h-pawn from promoting. || 

30. ... fxg6 

| Now the doubled pawns are straightened out, the activity of the white rooks has only assisted black. || 

31. gxh4 

| Now white has doubled h-pawns, both of which can be taken by the black queen. Which should black capture, one leads to a winning position for black and the other does not. || 

31. ... Qxh2 

| The wrong capture, now white can set up a blockade and hold a draw, though black can test the defensive technique for a while. Instead, 31. ... Qxh4 32. Rd6+ Kf7 33. h3, and white has set a little trap. Black is going to need 2 kingside pawns to win the game. If now, 33. ... Qxh3, white has the surprising, 34. Rxg6. If the black king ignores the rook, then the passed f-pawn will be insufficient for black to win. The disastrous, 34. ... Kxg6 35. Nf4+, is a royal fork and white is a piece up. 

Black can improve on this with a deflecting queen check as follows. So after, 31. ... Qxh4 32. Rd6+ Kf7 33. h3, black dislodges the white king with, 33. ... Qh6+. This clever zwischenzug gives the black queen some options not previously available should the white rook snatch the black g-pawn, and they are all about black being able to avoid the royal fork formation of white, Nf4+. The white king has 3 choices. 
(a) Firstly, 34. Ke1 Qxh3 35. Rxg6 Qh4+ 36. Rg3 f4, and black wins the rook. 
(b) Secondly, 34. Kd1 Qxh3 35. Rxg6 Qf1+ 36. Kd2 Kxg6, black wins. 
(c) Thirdly, 34. Kc2 Qxh3 35. Rxg6 Qh2, and both white pieces are attacked, black will win the knight and still have the f-pawn to run. || 

32. Rd4 

| The white rook now controls the 4th rank and protects the h-pawn. The white knight should be quickly re-located on the d3 square, to keep a tight grip on the f4 square. In this way, the black king can not enter the game and the black f-pawn can not advance. The white king is then free to protect the queenside pawns. If black advances, pawn g5, white will trade pawns and the black f-pawn has no means of safely advancing to the f4 square. Action by black would need to be switched to the queenside, but the black queen and 2 pawns will not be able to make progress. White has a blockade, but it is not yet a draw, black will probe for a while until white has satisfactorily demonstrated how to hold the position. || 

32. ... Qh1 

| Threatening, Qb1, a serious invasion which must be prevented. || 

33. Kc2 Ke5 
34. a4 

| Is white thinking of winning this game? This pawn push only creates losing chances for white. Much simpler was, 33. Nc1, since, 33. ... f4 34. Nd3+, a check which wins the f-pawn. Once the white knight establishes itself on the d3 square, the blockade would be almost complete. || 

34. ... Qf1 

| Black just keeps prodding at white, the knight must retreat, it is still just about tenable, but a difficult position to handle in time pressure. || 

35. Nc1 

| Instead, 35. Kd2 Qa1, and black will win a pawn. If, 36. b3 Qb2+, or, 36. Rb4 bxa4. In each case white will have weaknesses on both sides of the board. The black queen will be able to swing from side to side with speed, while the white pieces will not be so nimble as they try to coordinate to protect weak pawns. || 

35. ... Qg2+ 
36. Kb3 

| Making matters really difficult for himself. The white king had 2 other choices. 
(a) Firstly, 36. Kb1, and white will lose the blockade of the f4 square with, 36. ... bxa4. If now, 37. Rxa4 Qf1 38. Kc2 Kf6, avoiding a potential check from the white knight and black is now threatening, pawn f4, next move. Since the white rook is on the a-file white can not play, 39. Nd3 Qe2+, as white loses the knight. 
(b) Secondly, 36. Kd1, seems to hold the position. Black must not blunder with, 36. ... Qxb2 37. Nd3+, a royal fork. Instead, black can try, 36. ... bxa4 37. Nd3+ Kf6 38. Rxa4 a5, setting a trap. If, 39. Rxa5 Qg4+, black wins the important white h-pawn. However, 39. Rd4, establishes the blockade and black can not make progress. || 

36. ... bxa4+ 

| The white queenside now loses all harmony, the black queen is permitted entry, all because the a-pawn was advanced to seek unnecessary activity. || 

37. Ka3 

| The other 2 choices to be rejected were 
(a) Firstly, 37. Kxa4 Qc2+ 38. Nb3, now there is no knight check on d3 available, so, 38. ... Qxb2, black wins a pawn and white now has 2 weak pawns to protect, this is very good for black. 
(b) Secondly, 37. Rxa4 Qd2, a double purpose move. The queen pushes the white knight into further inactivity, 38. Na2, which now allows, 38. ... f4, supported by the queen on d2. This pawn will now cost white material. || 

37. ... Qc2 

| Attacking the unprotected c1 knight and also threatening a mate on the b3 square. White can avoid immediate loss, but at a cost of losing the blockade. || 

38. Nd3+ Kf6 
39. Nc5 

| The active knight seems to be working hard for white, but it is not the queenside which is of greatest importance. Black is now close to winning the white h-pawn, leaving 2 connected passed pawns to quickly run through. One of them will cost white the knight, it then comes down to wether white can make something of his queenside pawns before the remaining black kingside pawn costs white a rook. || 

39. ... Qc1 

| Yet another double purpose move. Firstly, Qa1+, is a strong threat, should white lose the b2 pawn, then the c-pawn anchor point for the rook also becomes weak. Secondly, the queen is supporting the advance, pawn f4, though the real value of such an advance is yet to be revealed. || 

40. Rxa4 Qe3 

| Gaining an attack tempo on the misplaced white knight. There is also a much deeper reason for this queen centralisation, again taking advantage of the uncoordinated white queenside. || 

41. Nxa6 f4 

| Now the white h-pawn has no defence it will be easy prey for the black queen, but there is a subtle point to be taken into account first. || 

42. Rd4 Kf5 

| Now a triple purpose move. Firstly, the obvious defence to the f4 pawn. Secondly the king advance has given the black queen another less obvious but still important option. The third purpose is shown in a line black must avoid, 42. ... f3 43. Nc5 f2 44. Ne4+, when white will give up the knight for the advanced black f-pawn and have a blockade, the rook shuffles between the, d4, and, b4, squares. The flashy, 44. ... Qxe4 45. Rxe4 f1=Q 46. Rd4, is a blockade for white. || 

43. Nb4 

| Instead white can have some activity with, 43. Nc5 Qe7 44. b4 Qxh4 45. b5 Qf6, but black is still winning. || 

43. ... Qe7 

| The white knight has to find a route back to the kingside and place itself on a square where it must give itself up for the f-pawn. The black queen pin is just another time-consuming annoyance for white, the white queenside pawns are not going to roll just yet. || 

44. Kb3 Qxh4 
45. Nd3 g5 

| Black is ahead in the race and will not be subject to having the king harassed, a luxury not available to white, such is the difference of having a queen against 2 pieces in an open position. || 

46. c4 Qg3 

| Another double purpose move. Simultaneously breaking and making a pin speeds up the black plan while slowing down the white plan. || 

47. c5 f3 
48. Kc4 f2 
49. Nxf2 Qxf2 
50. c6 

| The black plan is easy, promote the g-pawn. White similarly is trying to promote the c-pawn. The ease which each plan can be carried out is however far from equal. The black g-pawn needs only 4 moves for success and no support from either king or queen as it runs. The white c-pawn now only needs 2 moves, but it needs support and the white forces are not coordinating. White has to be aware of black queen checks up the c-file, so white must take time with either the rook to block, or with king advances. Had white tried a slower approach with, 50. b4 g4 51. b5 g3 52. b6 g2 53. b7 Qb2, black holds the white b-pawn for a move and the black g-pawn is now unstoppable. || 

50. ... Qxb2 
51. Kc5 Qc3+ 
52. Kd5 

| Trying to assist the c-pawn advance with, 52. Rc4, fails to, 52. ... Qa5+ 53. Kd6 Qd8+ 54. Kc5 g4 55. Rd4 Qc8 56. Kb6 g3 57. c7 g2 58. Rd8 g1=Q+, promoting with check, 59. Kc6 Qc1+, and the c-pawn is lost. || 

52. ... g4 
53. Rc4 

| The desperate, 53. Rxg4, hoping the advanced c-pawn will hold a draw fails to, 53. ... Kxg4 54. Kd6 Qd4+, black utilises a standard method of forcing the attacking king to occupy the square in front of the pawn, enabling the black king to approach and join in to capture the advanced pawn. After, 55. Ke7 Qc5+ 56. Kd7 Qd5+, the white king must block the pawn with, 57. Kc7, and black moves in with, 57. ... Kf5 58. Kb6 Qd4+ 59. Kb7 Qb4+ 60. Ka7 Qc5+ 61. Kb7 Qb5+ 62. Kc7 Ke6, and it is almost over. After, 63. Kd8, a trap has been set by white since, 63. ... Qxc6, is stalemate. However black can improve with, 63. ... Qb8+ mate. || 

53. ... Qe5+ mate 
0-1 

| The mating formation is called the Swallowtail mate. The 2 potential flight squares a knight's jump from the checking queen are occupied by friendly forces of the mated king, giving the impression of the trailing v-formation of a swallow's tail. || 

* * * 

## Unannotated game. 

White: Keres. 
Black: Fischer. 
Event: Candidates Tournament, Bled, 1959. 
Result: 0-1 in 53 moves. 
Opening: Sicilian Defence, Najdorf Variation, B99. 

1. e4 c5 
2. Nf3 d6 
3. d4 cxd4 
4. Nxd4 Nf6 
5. Nc3 a6 
6. Bg5 e6 
7. f4 Be7 
8. Qf3 Qc7 
9. O-O-O Nbd7 
10. Be2 b5 
11. Bxf6 Nxf6 
12. e5 Bb7 
13. exf6 Bxf3 
14. Bxf3 Bxf6 
15. Bxa8 d5 
16. Bxd5 Bxd4 
17. Rxd4 exd5 
18. Nxd5 Qc5 
19. Re1+ Kf8 
20. c3 h5 
21. f5 Rh6 
22. f6 gxf6 
23. Nf4 h4 
24. Rd8+ Kg7 
25. Ree8 Qg1+ 
26. Kd2 Qf2+ 
27. Ne2 Rg6 
28. g3 f5 
29. Rg8+ Kf6 
30. Rxg6+ fxg6 
31. gxh4 Qxh2 
32. Rd4 Qh1 
33. Kc2 Ke5 
34. a4 Qf1 
35. Nc1 Qg2+ 
36. Kb3 bxa4+ 
37. Ka3 Qc2 
38. Nd3+ Kf6 
39. Nc5 Qc1 
40. Rxa4 Qe3 
41. Nxa6 f4 
42. Rd4 Kf5 
43. Nb4 Qe7 
44. Kb3 Qxh4 
45. Nd3 g5 
46. c4 Qg3 
47. c5 f3 
48. Kc4 f2 
49. Nxf2 Qxf2 
50. c6 Qxb2 
51. Kc5 Qc3+ 
52. Kd5 g4 
53. Rc4 Qe5+ mate 
0-1 

* * *