Skype coaching session, 19 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: Shirov, 2718. 
Black: Kramnik, 2772. 
Event: Corus Wijk aan Zee 2001. 
Result: 0-1 in 53 moves. 
Opening: Caro-Kann Advanced, Botvinnik-Carls Defence, B12. 

1. e4 c6 
2. d4 d5 
3. e5 c5 

| An interesting choice. The usual Caro-Kann strategy is to postpone immediate central tension, develop pieces, find some king safety by castling, and start the fight as the middlegame begins. Here black is deliberately seeking to initiate liquidation of the closed white centre if allowed. || 

4. dxc5 

| If white tries to keep a dark square central pawn chain with, 4. c3, simply, 4. ... Nc6, placing pressure on the centre is fine for black. If now, 5. Ne2 Bf5, or, 5. Nf3 Bg4, the black c8 bishop is developed before the natural black move, pawn e6. || 

4. ... e6 

| Black chooses to steer the game into a French Defence Advanced a tempo down, but with white having committed to capturing on the c5 square. The question arises, why not play the French Defence in the first place and then have tempi equal by following mainline theory? It is all about skirmishing with repertoires. The French Defence player is not guaranteed the opponent will choose the Advanced Variation against it. Consequently a Caro-Kann player wishing to avoid the standard white treatment in the CK Advanced with black playing, 3. ... Bf5, steers the game into a different opening system perhaps in the hope the white opponent will not be familiar with strategies of the French Defence. If this thinking is sound, then perhaps the opening moves, 1. e4 e6 2. d4 d5 3. e5 c5 4. dxc5, with an extra tempo for black could be renamed the French Attack? || 

5. Bf4 

| A new move, which seems not to have been adopted by other titled players since this game, a little clue there. Instead, 5. Be3 Nh6, is fine for black. In this type of position white should not try to keep the snatched c5 pawn, but instead proceed with simple developing moves and wait for the fight to begin in the middlegame. || 

5. ... Bxc5 
6. Bd3 

| Making an offer that most players with black would probably accept. Perhaps black reasons there is no point in unnecessary tactics against an opponent well known for very sharp play. || 

6. ... Nc6 

| Declining, 6. ... Qb6, attacking both the f2 and b2 pawns. It seems, 7. Nh3, would be forced, after which, 7. ... Qxb2 8. Nbd2, gives white a good lead in development for a pawn. If, 7. Qe2 Qxb2, and black is winning the a1 rook. Finally if, 7. Qd2 Qxb2 8. Qc3 Bb4, and white is losing queen for a bishop. || 

7. Nf3 f6 

| Good old-fashioned French Defence strategy of removing the white pawn centre, which still applies even when being a tempo down. || 

8. Nbd2 fxe5 

| Black avoids the unnecessary complications of, 8. ... g5 9 Bg3 g4 10. Nh4, when black can not win a pawn with, 10. ... Nxe5 11. Bxe5 fxe5, as, 12. Qxg4, leaves material equal but with the black kingside wide open. || 

9. Bxe5 

| If, 9. Nxe5 Qf6, gives black good play. There are no tricks involving, 10. Qh5+ g6 11. Nxg6 hxg6, as the black queen backwardly defends the black h8 rook. || 

9. ... Nf6 
10. Qe2 O-O 
11. O-O-O 

| Very aggressive. Instead a much calmer approach is, 11. O-O, when white will centralise rooks and can plan for a central break with, pawn c4 after completing development. || 

11. ... a5 

| More French Defence strategy. The arrangement of the white knights has a touch of the Tarrasch system about it, and if the d2 knight is destined to enter the game by the b3 square, then the advance of this pawn is often awkward for white. || 

12. Nb3 

| Assisting black to gain an initiative, despite black being a tempo down from move 4 onwards. Instead white could set up some opposite wing action with, 12. h3, intending pawn g4. Perhaps the bolder amongst us might consider an immediate, 12. g4, though this would require some careful analysis before being offered. || 

12. ... Bd6 
13. Kb1 Bxe5 

| An important softening of the dark squares which haunts white for a long time. || 

14. Nxe5 a4 
15. Nc1 

| A sad retreat creating disharmony in the white queenside. White might have been planning, 15. Nxc6 bxc6 16. Nd4, aiming to win the black e6 pawn with, 16. ... Qb6 17. Nxe6 Re8 18. Bf5, but now black shows some lateral thinking, literally, with, 18. ... c5 19. Rhe1, and now a 4th and crucial attack hits the white e6 knight, 19. ... Ra6, and the white knight on e6 is lost. After the white knight retreats to c1 black forces a weakness in the white queenside which persists almost to the end of the game. || 

15. ... a3 

| Clearly white will not capture this pawn, and permitting black to capture on the b2 square seems equally horrible, so the process of elimination forces the white reply. || 

16. b3 

| The black a3 pawn is now both a pain in the side and a pain on the side for white. || 

16. ... Qa5 

| As subtle as the proverbial flying brick. The intention is, 17. ... Qc3, and mate on the b2 square. This cheeky plan is rather difficult to prevent without creating more queenside problems for white. In fact, the weakening of the white queenside must lead white to wonder about placing the king on the queenside in the first place. || 

17. c4 

| Giving the white queen a defensive role along the 2nd rank. There are quite some interesting ideas to be found by both players had white tried, 17. Qe3. Clearly, 17. ... Qc3 18. Bxh7+ Kxh7 19. Qxc3, black loses the queen is not going to occur. Instead black can try for tactics with, 17. ... Nxe5 18. Qxe5 Ng4, with threats of, Nxf2, or, Rxf2, as appropriate. The big idea is to push the white queen off the a1-h8 diagonal and invade with, Qc3. White has a potential defence to mates on the b2 square if the c1 knight can transfer to the d3 square, but if black can find a way to eliminate a defending knight on d3 then the mate on the b2 square is on again. So after, 18. ... Ng4 19. Qh5 Qc3 20. Bxh7+, the d3 square is free for the c1 knight. Note, white should not play, 20. Qxh7+ Kf7 21. Bg6+ Ke7 22. Nd3, as now, 22. ... Nxf2, threatens to remove the defending knight and mate will follow on the b2 square. So after, 20. Bxh7+ Kh8 21. Bg6+ Nh6, and now the black knight can not be used to eliminate the white defending knight. Play would continue, 22. Nd3, but now black sets about removing the vital white defender with, 22. ... Rxf2, which threatens mate in 2 moves beginning with, Qxc2+. So white must play, 23. Rc1, but after, 23. ... Rd2, black is threatening, Rxd3, and, Qb2+ mate. White has one last trick with, 24. Rhf1, which threatens an immediate, Rf8+ mate. Black simply plays, 24. ... Bd7, and black is safe. Now, 25. Qe5, guards the potential mate on the b2 square but allows, 25. ... Rxd3, black protects his own queen and undermines the defence of the white queen, black has just won a piece and wins. 

Guess you all noticed why 17. Qe3 would not work for white. || 

17. ... dxc4 
18. Nxc4 Qc7 

| The black queen sits on the c-file, planning to support a dark square invasion which white manages to prevent, but at some cost. The black queen and white c4 knight are also in that relationship of mutual restriction on the, a5, b6, d6, e5, squares. The black queen does not need access to these squares, however the white knight would really appreciate to be able to return to e5 if possible. || 

19. Rhe1 Nd5 

| Another non-subtlety, a triple fork with, Nc3+, is threatened. || 

20. Qd2 

| Now white has lost all possibilities of kingside activity, the battleground is now permanently the white queenside. Opposite wing castling is supposed to offer chances of mutual attacking options, and while white can definitely castle long in some French Defence set-ups, something is radically wrong here indicating castling kingside was advisable. || 

20. ... Ncb4 

| Creating tandem knights, that is where 2 knights support each other, meaning should one knight be captured then the option of the other knight recapturing is available, which is of great use if the original knight on a square has significant influence on the position. White now realises the game is critical and must sift the least painful of many options. || 

21. Ne2 

| Forced. The c3 square needs defending and this costs white a pawn. Instead trying to prevent black from achieving, Nc3, fails as follows. If, 21. Be4trying to prevent the black plan with a threat of material gain, black ignores it and continues, 21. ... b5 22. Ne3 Nc3+ 23. Ka1 Nxd1 24. Bxa8 Qe5+ 25. Kb1 Nc3+ 26. Ka1 Ne4+, winning the white queen. Instead if, 21. Ka1 b5 22. Ne5 Nc3, snares the immobile white d1 rook. || 

21. ... Rxf2 

| A safe and solid pawn gain. White must now start to confuse matters as best as possible. || 

22. Be4 b5 

| Intending to displace the white c4 knight which has only one flight square. || 

23. Qd4 

| Going for complications, Instead, 23. Ne3, would keep more tension in the position. || 

23. ... Rxe2 

| The position begins to simplify a little and white is not going to have sufficient material left to play for tricks. || 

24. Rxe2 bxc4 
25. Bxd5 Nxd5 
26. bxc4 Ba6 

| The "lazy French bishop" at last makes a move, and an important one. The pin on the white c4 pawn interferes with white getting the rooks mobilised. || 

27. Rc1 Nf4 
28. Rd2 e5 

| The start of a slow march. Two minor pieces and a passed pawn will overpower a rook, 3 attacking units against 1, but simplifying to such an ending is not easy to organise. || 

29. Qe3 Bxc4 
30. Ka1 Qf7 

| A triple purpose move. Escaping the pin down the c-file, over-protecting the f4 knight and planning a tactic down the h8-a1 diagonal beginning with, 31. ... e4, followed with, 32. ... Qf6+. || 

31. g3 

| If, 31. Qxe5 Nd3, black wins an exchange. Instead ganging up on the black bishop with, 31. Qc3, allows simplification with, 31. ... Ne2 32. Qxc4 Nxc1 33. Qxc1 e4, after which the black e-pawn is strong as well as creating strong opportunities for the black queen on the a1-h8 diagonal. || 

31. ... Nd3 

| Heading for an ending a solid pawn up. Instead, 31. ... Ng6, black would still have a strong advantage, but the position would still have complexities. || 

32. Rxd3 Bxd3 
33. Qxd3 h6 

| Again the threats down the h8-a1 diagonal allow black the tempo-gaining, 33. ... e4, as, 34. Qxe4 Qf6+, with mate on the b2 square to follow. || 

34. Qb3 

| White must remove queens and try to hold the rook and pawn ending while effectively being a king down. Leaving queens on the board would allow the black queen and rook to create strong threats to the white king while supporting the advance of the e-pawn. || 

34. ... Qxb3 

| The black strategy now is to advance his king to support the advance of the passed e-pawn, white will however be able to stop the king entry. Black must then use the passed a-pawn to tie down the white king while making progress in the centre. At the right moment black will aim to liquidate the queenside pawns, possibly at the cost of the passed e-pawn if need be, in the understanding that the white king will be too far from the kingside action, where the black king and rook will outplay the lone white rook and win both white kingside pawns. If black does not have to lose the e-pawn then it will become a question of finding a means of achieving king support for the e-pawn. White will suffer from not having the king able to fight in the centre. || 

35. axb3 Kf7 
36. Rc6 a2 

| Very important in the process for black to win. The white king and rook can combine to win the a2 pawn, but should they do so then the black e-pawn will be too far advanced to be caught by the white king. Furthermore the black a2 pawn guarantees that the white king can not move from the a1 square while the black rook is on the a-file. || 

37. h4 

| White is trying to create a blockade on the kingside, the black king must not be allowed to support the passed e-pawn. Instead, 37. b4, attempting to create some queenside confusion in the moves to the time control might have been worth a try, it should not work, but one never knows... || 

37. ... Ra3 
38. b4 e4 

| Black should ignore the white g3 pawn, and just concentrate on finding a means of forcing the white rook to give up the control of the black 3rd rank. Once the white rook retreats, then the black king can advance and the e-pawn will prove to be the decisive factor. || 

39. Rc4 e3 
40. Re4 Kf6 
41. g4 

| White tries to prevent further black king entry, but this will not succeed. || 

41. ... h5 

| A simple way to destroy the blockade. No matter how white responds to this, the black king advances onto the f5 square. || 

42. gxh5 Kf5 
43. Re7 

| The white rook now has 5 squares on the e-file to restrain the advance of the black e-pawn from behind or give a series of checks to the black king. Here the e7 square is chosen so if black mishandles matters then the g7 pawn might be captured and the passed white h5 pawn might become a strength. || 

43. ... Kf4 
44. b5 

| Instead, 44. Rxg7 e2 45. Re7 Re3, one reason for placing the black rook on the a3 square is revealed, the e-pawn is shielded and will promote. || 

44. ... Rb3 

| Time to remove the white b-pawn before it becomes too dangerous. Now that the black king is advanced the a2 pawn is not needed, the focus has now been transferred onto the black e-pawn. || 
45. Rf7+ 

| The checks begin rolling in. Instead, 45. Kxa2 Rxb5, and at the right moment black will have, Re5, again shielding the e-pawn and forcing promotion. || 

45. ... Ke4 
46. Re7+ Kd3 
47. Rd7+ Kc2 

| Now white has to be aware of potential mating nets. || 

48. Rc7+ Kd2 

| Black somehow misses, 48. ... Rc3, which wins very quickly. There are 2 possible mates after, 49. Rxc3+ Kxc3 50. b6 e2 51. b7 e1=Q+ 52. Kxa2 Qb1+ 53. Ka3 Qb3+ mate. Instead after, 48. ... Rc3 49. Rxg7 Kb3, black threatens mate on the c1 square, so, 50. Rg1 Rd3 51. b6 e2 52. b7 Rd1+ 53. Rxd1 exd1=Q+ mate would occur. || 

49. Kxa2 Rxb5 
50. Rxg7 e2 

| Now the black plan is to shuffle the king backwards towards the centre when the white rook checks can be blocked with, Re5. || 

51. Rd7+ Kc2 
52. Rc7+ Kd3 
53. Re7 Rxh5 
White resigns, 0-1 

| White can not protect the h-pawn, and if, 54. Rd7+ Ke4 55. Re7+ Re5, the e-pawn promotes. || 

* * * 

## Unannotated game. 

White: Shirov, 2718. 
Black: Kramnik, 2772. 
Event: Corus Wijk aan Zee 2001. 
Result: 0-1 in 53 moves. 
Opening: Caro-Kann Advanced, Botvinnik-Carls Defence, B12. 

1. e4 c6 
2. d4 d5 
3. e5 c5 
4. dxc5 e6 
5. Bf4 Bxc5 
6. Bd3 Nc6 
7. Nf3 f6 
8. Nbd2 fxe5 
9. Bxe5 Nf6 
10. Qe2 O-O 
11. O-O-O a5 
12. Nb3 Bd6 
13. Kb1 Bxe5 
14. Nxe5 a4 
15. Nc1 a3 
16. b3 Qa5 
17. c4 dxc4 
18. Nxc4 Qc7 
19. Rhe1 Nd5 
20. Qd2 Ncb4 
21. Ne2 Rxf2 
22. Be4 b5 
23. Qd4 Rxe2 
24. Rxe2 bxc4 
25. Bxd5 Nxd5 
26. bxc4 Ba6 
27. Rc1 Nf4 
28. Rd2 e5 
29. Qe3 Bxc4 
30. Ka1 Qf7 
31. g3 Nd3 
32. Rxd3 Bxd3 
33. Qxd3 h6 
34. Qb3 Qxb3 
35. axb3 Kf7 
36. Rc6 a2 
37. h4 Ra3 
38. b4 e4 
39. Rc4 e3 
40. Re4 Kf6 
41. g4 h5 
42. gxh5 Kf5 
43. Re7 Kf4 
44. b5 Rb3 
45. Rf7+ Ke4 
46. Re7+ Kd3 
47. Rd7+ Kc2 
48. Rc7+ Kd2 
49. Kxa2 Rxb5 
50. Rxg7 e2 
51. Rd7+ Kc2 
52. Rc7+ Kd3 
53. Re7 Rxh5 
White resigns, 0-1 

* * * 
 __,_._,___