Palermo-Najdorf Analysis

by Bill Wall

In the January 1969 issue of Chess Life, Samuel Reshevsky annotated this game in his ?The Art of Positional Play? column.? I wanted to compare his notes and annotations against today?s strong chess engines (latest versions of Komodo, Rybka, Fritz, and Houdini).? Here are some of the comparison annotations between Reshevsky and the chess engines.

Vicente Palermo ? Miguel Najdorf, Argentina Ch (Mar del Plata) Round 7, 1968

????????????????????????????????

1.f4 Nf6 2.Nf3 g6

[2...d5]

3.g3 Bg7 4.Bg2 0?0

[4...d5]

5.d3

[5.0?0]

5...c5

[5...d5]

6.0?0 Nc6

[6...d5; 6...d6]

7.e4 d6

Here, Reshevsky remarked ?A variation of the Closed Sicilian has been reached.? White attempts to attack on the Kingside by pushing pawns, while Black seeks counterplay on the opposite wing.? Black must play carefully to thwart his opponent?s plans.

8.a4

Here, Reshevsky says ?More usual is 8.N-B3 (8.Nc3) which is superior to the text.?? ?However, Rybka 4 gives equal weight to 8.Nc3 and 8.a4.? Rybka 4 says the best move is 8.c4, while Komodo 8 says the best move is 8.c3.? Houdini likes 8.Ne1 and Fritz likes 8.Na3.

8...Rb8N

[8...a6; 8...Bg4; 8...Bd7]

9.Nc3N

[9.c3; 9.h3]

9...a6

[9...Be6]

10.Nh4

[10.Bd2; 10.h3]

10...Nd4

Reshevsky says, ?Preferable was 10...Ne8 11.f5 e6, preventing the advance g4.? However, Komodo likes 10?Nd4 and doesn?t even have 10?Ne8 as one of its top 8 choices.? Rybka likes 10?e6 as the best move and does not have 10?Ne8 among its top 8 choices.? Fritz and Houdini like 10?Bg4.

11.f5 e6

[11...Bd7; 11...b5]

12.g4

[12.Ne2]

12...Nd7

Here, Reshevsky says ?12?Nd5 would have been met by 13.g5 Nxc3 14.bxc3 Nc6 15.f6, with a strangle-hold.?? But Komodo and Rybka see nothing wrong with 12?Nd5, and give it a slight favor for Black.? After 12?Nd5 13.g5 Nxc3 14.bxc3, instead of what Reshevsky commented with 14?Nc6, Black can now play 14?exf5 and have a good game, since 15.cxd4 fails to 15?Bxd4+ and 16?Bxa1.? Komodo, Fritz and Rybka likes 12?exf5 even better than 12?Nd5 or 12?Nd7.? Houdini likes 12?Ne8.? 12...Nxe4 13.Nxe4 exf5 (13...Qxh4? 14.Bg5) 14.gxf5 (here, Reshevsky says 14.Ng5, attempting to retain the piece, would have failed on account of 14?fxg4, followed by h6.? But the engines give this as a favorable line for White after 15.c3 h6 16.Nxg6 fxg6 17.Bd5+ Kh8 18.Rxf8+ Qxf8 19.Nf7+ Kh7 20.cxd4) 14...Qxh4 15.Bg5 Qh5 16.Qxh5 gxh5 17.f6 Bh8 18.c3 Nf5 19.Bf4; 12...Nd5 13.g5 Nxc3 14.bxc3 Nc6 15.f6 Bh8 16.Nf3; 12...exf5 13.exf5 Ne8.

13.g5

[13.Nf3 Ne5]

13...Be5

Reshevsky writes, ?The Bishop must not be permitted to be locked in by f6.?? But Komodo likes 13?Bh8 and Rybka and Fritz like 13?exf5.? Only Houdini favors 13?Be5.

14.f6

[14.Be3]

14...b5

Reshevsky writes, ?Unavailing was 14...h6 15.Nf3 Nxf3+ (if 15...hxg5 16.Nxe5 followed by 16.Bxg5) 16.Qxf3 hxg5 17.Bxg5, with a hopeless position for Black.? But none of the engines evaluate the position as a hopeless position for Black.? Rybka evaluates 17?Qb6 or 17?c4 or 17?Re8 as even for both sides.? Komodo likes 17?c4, followed by 18?b5, and does not show a hopeless position for Black.? Houdini likes 17?Re8 and Fritz prefers 17?b5.? Besides, Black does not have to play 16?hxg5.? Komodo says that 16?b5 is the best move with an even position.

15.axb5 axb5 16.Nf3 b4 17.Ne2

Here, Reshevsky writes, ?To be considered was 17.Nxe5 Nxe5 (17...bxc3 18.Nxd7 cxb2 19.Bxb2 Bxd7 20.Bxd4 cxd4 21.Qg4 and 22.Qh4) 18.Ne2.?? But this is just an equal position, according to Komodo, after 18?b3 19.Nxd4 cxd4 20.cxb3 Qb6.? Also, after 17.Nxe5 bxc3 18.Nxd7 cxb2 19.Bxb2 Bxd7 20.Bxd4 cxd4 21.Qg4, Reshevsky implies that White is better.? But Rybka says the position is even and Komodo says the position slightly favors Black after 21?Qc7 22.Qh4 h5 23.gxh6 Kh7 24.Qf2 e5.? Houdini gives this position a slight edge for Black, while Fritz gives this position a slight edge for White.

17...Nxf3+ 18.Bxf3

Reshevsky writes, ?White should have continued with the attack immediately by playing 18.Rxf3, and if 18?h6, 19.Qe1 hxg5 20.Bxg5 (threatening Qh4 followed by Rh3) 20?Bxf6 21.Rxf6 Nxf6 22.Qh4 with a won position.? If 18?h5 (trying to blockage the position), 19.Rh3, followed by Bf3 and an eventual Bxh5.?? Komodo and Rybka agree that White should have played 18.Bxf3, but Fritz and Houdini like 18.Rxf3.? Even if 18.Bxf3 was played, Reshevsky only discusses 18?h6.? However, Komodo likes 18?Bb7 as best and Rybka likes 18?Qc7 as best.

18...h6

Komodo likes 18?Bb7 as best and Rybka likes 18?Qc7 as best.? Houdini likes 18?Qb6, and only Fritz likes 18?h6.

19.h4

?[19.gxh6 Qxf6]

19...Nxf6

Rehsevsky says, ?Otherwise White could have lined up his pieces on the Kingside at will and crashed through to victory.? Najdorf decides to give the Knight for three pawns, which are normally sufficient compensation.? The sacrifice is not surprising, since he dislikes intensely to be on the defensive.??? But none of the engines likes 19?Nxf6, and consider it a blunder.? Both Komodo and Rybka prefere 19?Bb7 and 20?Ra8 as equalizing for Black.

20.gxf6 Qxf6 21.Kg2

Reshevsky writes, ?Of course not 21.Bxh6 Qxh4, winning material.?? The continuation would be 22.Bf4 Bxf4 23.Nxf4 Qxf4.? However, Reshevsky misses the better move of 21.Qd2 as the engines prefer, and if 21?h5 22.Qg5 with a strong game for White.

21...Bxb2

Reshevsky comments, ?Better than 21...Qxh4 22.Rh1 Qf6 23.Bxh6 Bxb2 24.Bxf8 Bxa1 25.Bxd6, etc.?? This may be true, but Black can improve his play in this variation with 23?Re8 (instead of Reshevky?s 23?Bxb2 line) as Komodo, Fritz, and Houdini recommends, or 23?Rd8 as Rybka recommends.

22.Bxh6

Reshevsky gives this an exclamation (!) point, but that?s not best according to all the engines.? The engines all prefer 22.Bxb2 Qxb2 23.d4 as best for White.

22...Bxa1

Reshevsky writes, ?Inadvisable was 22...Re8, because of 23.Ra7, with serious pressure on the f7 pawn.? If 23?Qxh4 24.Qd2 Bf6 (to prevent Bg5) 25.Rh1.?? But the engines find at least three better moves for White than 23.Ra7.? The strongest that the engines agree is 23.Bg5 Qg7 24.Ra2.

23.Bxf8

[23.Qxa1 Qxa1 24.Rxa1 Re8 25.h5 gxh5 26.Ng3]

23...Be5

Here, Reshevsky comments, ?23...Kxf8 24.Qxa1 Qxh4 (if 24...Qxa1 25.Rxa1, with [White] the better chances in the ensuing endgame) 25.Rh1 Qg5+ 26.Kf2, with the annoying threats of Rh8+, followed by Qg7 or Qh7.?? ??White may have the better endgame, but after 23?Kxf8 24.Qxa1 Qxa1 (24?Qxh4? Is just plain bad) 25.Rxa1 e5, and 26?Be6, and White will still have a hard time in this endgame.

24.Bh6 Qxh4 25.Rh1 Qe7

[25...Qf6 26.Be3]

26.Qd2

[26.Be3]

26...Bd7

Rehsevsky writes, ?Inadvisable was 26...b3 27.Rb1 b2 28.c3 Qb7 29.d4 Bh8 30.e5 d5 31.Bg5, followed by mating threats.? ??The engines have a lot of alternatives in this variation.? First, the engines do not think 26?b3 27.Rb1 is the best move.? Komodo, Fritz, and Houdini favors 27.d4,? while Rybka favors 27.cxb3.? 27.Rb1 is the 5^th alternative in both engines.? After 26?b3 27.Rb1 b2, the engines like 28.Bf4 as best for White.? Now, Reshevsky has 28?Qb7, but the engines say that after 28?Qh4, Black is winning.? After 29.d4, Reshevsky wants Black to play 29?Bh8.? But the engines give a slight lead to Black after 29?cxd4.? Now, Reshevsky wants White to play 30.e5, but the engines give a clear advantage to White after 30.dxc5.? After 30.e5, Reshevsky recommends 30?d5, but that?s not even considered by Komodo or Rybka.? The equalizing move for Black is 30?Qb3.

27.Bf4

[27.d4 cxd4 28.Nxd4 Bf6 29.Be3]

27...Bg7

Reshevsky states that better was 27...Bf6, preventing White?s next move.? The engines only think that 27?Bf6 is slightly better.? Rybka still favors Black with either 27?Bf6 or 27?Bg7.

28.Bg5 Bf6

Rehevsky writes, ?Forced, for if 28...f6, Black?s KB would have remained out of play, and 28...Qe8 29.Qf4 followed by Qh4, would have been unpleasant for Black.?? But the engines say that 28?Bf6 is not forced and they prefer 28?f6!? After 28?f6, Black plays 29?e5.? Rybka says the position slightly favors Black, while Komodo and Houdini say the position is even.? Only Fritz gives a slight advantage to White.

29.Bxf6 Qxf6 30.Qh6 Qg7

At this point, Reshevsky says, ?Black?s task is difficult.? White was threatening to win material with 31.Qh7+ followed by Qh8.? Black should have tried 30?Be8.?? But after 30.Qh6, the Komodo, Rybka, and Fritz think 30?Re8 is best, equalizing according to Komodo, and slightly favoring Black according to Rybka.? Houdini likes 30?Rc8, with the position being equal.

31.Qh2

Reshevsky gives this move an exclamation mark.? Rybka and Fritz think the best move is 31.Qh4, while Komodo favors 31.Qd2.? Houdini likes 31.Qxg7+ as slightly favoring White.

31...e5

Reshesvsky remarked ?Black is compelled to weaken his position.? The text enables White?s Bishop to become active.? Of no help was 31...Rb6 32.Qh4 (threatening penetration of Black?s position by the Queen via e7 or d8) 32?g5 33.Qg3 followed by Ng3-Nh5.?? But according to the engines, Black is not weakening his position with 31?e5.? Rybka says that Black is slightly better and Komodo says it is even.? Also, after 33.Qg3, White cannot play 34.Ng3 as suggested by Reshevsky, since the queen occupies the g3 square.? After 33.Qg3, Komodo and Fritz like 33?Qe5, while Rybka and Houdini like 33?Be8 and an even game.

32.Ra1 Qf6

[32...Bb5 33.Qh4]

33.Qg3

[33.Qh6 Be6 34.Kg3]

33...Kg7 34.Bg4

[34.Ra7]

34...Be8

Reshevsky writes, ?Preferable was 34...Bb5 with the intention of continuing with ?c4.? If 35.Rh1 Rh8, with good drawing chances.?? ?None of the engines pick 34?Bb5 as its first choice.? Their selection of the best move in this position is 34?Bc6, followed by 35?Rh8.

35.Qe3

[35.Rh1]

35...Kg8

Reshevsky gives this move a question mark, but the engines don?t show this as a bad move.? Reshevsky goes on to say, ?Permitting White?s pieces to make inroads into Black defenses.? With 35?Qh4, Black could have stopped White?s progress.? If 35...Qh4 36.Bf3 Qh6 37.Qxh6+ (37.Qf2 Qg5+ 38.Ng3 Bb5, etc) 37...Kxh6 38.Ra6 b3 39.cxb3 Rxb3 40.Rxd6 Bb5, winning the QP ?(41.Nc1 Rc3), after which White would have no winning chances.?? Reshevsky fails to find the best move, at least according to the engines, of 35?Bc6.? The engines give Black a slight edge after 35?Bc6, but only an equal game after 35?Qh4.? After 35?Qh4, Rybka likes 36.Qf3, while Komodo likes 36.Bh3.? Reshevsky looks into 35?Qh4 36.Bf3 Qh6 37.Qxh6+ Kxh6 38.Ra6 and recommends Black play 38?b3.? The engines consider this a mistake and recommend, instead, 38?Bb5 39.Ra7 Kg5, slightly favoring Black.

36.Qh6

[36.Nc1]

36...Bb5

Reshevsky is silent on this move, and they may be the losing move.? Komodo recommends 36?c4, Rybka and Fritz recommend 36?Qg7, and Houdini recommends 36?Bc6.? 36?Bb5 looks like a dubious move to the engines.

37.Rf1

Reshevsky makes no comment on this move, but the engines give a big advantage to White after 37.Ng3, with the threat of 38.Nh5.

37...Qg7

Reshevsky mentions that if 37...Qe7 38.Rh1 and wins, but doesn?t analyze the other alternative move of 37?Qh8, which could possibly draw.

38.Qg5

[38.Qh3]

38...c4

Reshevsky doesn?t comment on this move, and the engines consider it a mistake.? Komodo and Fritz prefer 38?Ba4,? while Houdini and Rybks prefer 38?Re8.

39.Qe7 Qf8

Reshevsky writes, ?Unsatisfacory was 39?cxd3 on account of 40.Rxf7 Qxf7 41.Be6 and wins.?? But Reshevsky fails to find the most resisting move for Black, 39?Rf8 seems to be better than 39?Qf8 according to the engines.? After 39?Rf8 40.Qxd6 cxd3 41.cxd3 Qh6 and Black is slightly better than what was played in the game.

40.Qc7

Reshevsky points out that 40.Qc7 threatens 41.Be6, but misses a better move for White.? According to the engines, best for White is 40.Qf6! cxd3 41.Rh1 Qg7 42.Qxd6 Rf8 43.cxd3.

40...Re8

[40...Ra8 41.Be6; 40...cxd3 41.Be6 Rc8 42.Qxc8 Qxc8 43.Bxc8 dxc2 44.Nc1 Bxf1+ 45.Kxf1]

41.dxc4

The engines don?t like this move as it only equalizes for White.? Komodo likes 41.Qb6 favoring White, while Rybka likes 41.Qa5 favoring White.? Reshevsky is silent on any alternatives.

41...Ba4 42.Ng3

Reshevsky comments, ?Unproductive was 42.Rc1 Qh6 (threatening 43?Qd2) 43.Qxd6 Qg5 44.Kf3 (44.Kh3 Kg7 45.Qxb4 Rh8+ 46.Kg3 Bd7) 44...Rd8 45.Qxb4 Bxc2 46.Rxc2 Rd3+ 47.Kf2 Qh4+ 48.Kf1 Qxg4, with at least a draw, for if 49.Qb8+ Kh7 50.Qxe5 Qf3+ 51.Ke1 Qh1+ 52.Kf2 Rf3#?? But the engines evaluate the position after 42.Ng3 or 42.Rc1 as equal.? After 42.Rc1 Qh6, Reshevsky recommends 43.Qxd6, but the engines consider that as a losing move.??? Komodo recommends 43. Kf3, while Rybka recommends 43.Qb7.? After 42.Rc1 Qh6 43.Qxd6 Qg5 44.Kf3 Rd8 45.Qxb4 Bxc2 46.Rxc2 Rd3+ 47.Kf2 Qh4+ 48.Kf1 Qxg4, Reshevsky claims this is at least a draw, but in all variations for White, the engines show that Black is winning.

42...Bxc2 43.Qa5

Reshevsky writes, ?Not 43.Bd7 Re7 44.Qxd6? Qd8.?? But Reshevsky fails to find the stronger move of 43.Bd7 Qe7, then 44?Rd8 as recommended by the engines. ?If 43.Bd7 Re7, instead of 44.Qxd6, White can play 44.c5 dxc5 45.Qxc5 Rxd7 46.Qxc2, and the game may draw.

43...b3 44.Qd2

Reshevsky mentions no alternative, but the engines think 44.Qc3, followed by 45.Bd1 is best.? The engines consider 44.Qd2 a blunder.

44...Qe7

Reshevsky doesn?t comment on this move, but the engines agree that 44?Ra8 is best, with the position favoring Black.? 44?Qe7 just equalizes.

45.Rh1

Komodo and Rybka evaluate 45.Qh6 as equal, while Fritz and Houdini give 45.Be2 as equal.? Rybka evaluates 45.Rh1 is slightly favoring Black.? Komodo evaluates 45.Rh1 as equal.

45...Qf6 46.Bd7

No comment from Reshevsky, but the engines consider 46.Bd7 as a mistake.? Komodo and Fritz prefer 46.Rf1, while Rybka and Houdini prefer 46.Bd1.

46...Qf4

Reshevsky gives this move an exclamation mark, but misses the even stronger moves of 46?Ra8! with the threat of 47?Ra2 as the engines indicate, strongly favoring Black.? 46?Qf4 only equalizes.

47.Qe2

Reshevsky says, ?Forced, for if 47.Qxf4 exf4 48.Bxe8 fxg3 49.Bc6 b2, with winning chances.?? But this move is not forced.? The engines point out that 47.Qd5 can be played, with only a slight edge for Black.? Also, after 47.Qxf4 exf4 48.Bxe8 fxg3, White doesn?t have to play 49.Bc6 as Reshevsky indicates.? The engines evaluate 49.Kxg3 as the better choice for White, but probably still losing.? After 49.Kxg3 b2 50.Kf4 b1=Q 51.Rxb1 Bxb1 52.e5 with a little bit more resistance.

47...Ra8 48.Rf1 Qg5

Reshevsky gives this a double question mark (??) indicating the losing move.? But the engines don?t consider this as a losing move.? Rybka, Fritz, and Houdini prefer 48...Qg5 as the best move and winning for Black, while Komodo likes 48?Qh4 as the strongest move, with 48?Qg5 also a strong, winning move.? Reshevsky makes the following observation, ?A move that can only be explained by time trouble.? Correct was 48...Ra2!, which would have brought him victory.? After ?49.Rxf4 exf4 50.Kf3 fxg3 51.Kxg3? Ra1, White would have been unable to stop Black?s pawn.? If ?49.Rxf4 exf4 50.Kf3 fxg3 51.e5 g2 wins.? Another try for White is 49. .Kh3, but Black could win after ?50.Kg4 f5+ 51.Bxf5 (51.exf5 Bxf5+ 52.Nxf5 gxf5+, followed by ?Rxe2.? If 51.Nxf5 Qh5+) 51?gxf5+ 52.Nxf5 (52.Rxf5 Ra1 53.Qf2 Bd1+.? If, in this line, 53.Rg5+ Kh8 54.Rh5 Qxh5+, followed by ?Bd1) 52?Qg6+ 53.Kh3 (53.Kh4 Bxe4) 53?Qh7+ 54.Kg4 (54.Kg3 Bd1 55.Qxd1 Qh2+ 56.Kg4 Rg2+ 57.Kf3 Qh3+.? If 56.Kf3 Qf4 mate) 54?Bd1 55.Qxd1 Rg2+ 56.Ng3 (56.Kf3 Qh3+) 56?Qg6+, with mate to follow.?? The engines give 48?Ra2 as the 4^th best move, with just a slight advantage for Black.? Reshevsky only considers 48?Ra2 49.Rxf4 exf4 50.Kf3, which the engines consider 50.Kf3 as a bad move, with Black having the better position.? The engines prefer 50.Kh3 and an equal game.

49.Qf3

[49.Kh3 Qh6+ 50.Kg2 Ra2]

49...Qe7

Reshevsky writes, ?Better was 49?Rf8, and if 50.Be6 b2 51.Bxf7+ Kg7 and Black wins.? White?s best after 49?Rf8 is 50.Ba4, but after 50?f5 51.exf5 gxf5, the chances would be even.?? However, none of the engines selected 49?Rd8 as the best move.? The engines agree that the top choice is 49?f6, and second choice as 49?Qe7.? Reshevsky states that after 49?Rf8, White?s best is 50.Ba4, but Komodo prefers 50.Qc3 as best for White.? Resevsky also says that after 49?Rf8 50.Ba4 f5 51.exf5 gxf5, the chances would be even.? But the engines strongly favor Black?s position after 52?f4.

50.Qg4

[50.Bc6 Ra2]

50...Ra7

Reshevsky comments, ?White?s threat of 51.Nf5 is too strong and Black?s position becomes untenable.? If for instance 50?b2 51.Nf5 Qf8 52.Be6 b1Q 53.Rxb1 Bxb1 54.Qxg6+ Kh8 55.Qh5+ Kg8 56.Bxf7+ Qxf7 57.Nh6+ and wins.?? But after 50.Qg4, Komodo and Rybka say that 50?b2 is the strongest move, and if 51.Nf5, then 51?Bxe4+ 52.Kg3 Ra3+ wins for Black.? Reshevsky also misses the best move after 50?Ra7 51.Nf5.? Reshevsky says that 51?b2 is best, but the engines give 51?Qxd7 and overwhelming advantage for Black.? The continuation might be 51?Qxd7 52.Nh6+ Kg7 53.Rxf7+ Qxf7 54.Nxf7 b2 as winning.

51.Nh5

[51.Bf5 Bd3]

51...Rxd7

No analysis from Reshevsky at this position, but the engines say the best move is 51...Bd3! 52.Rf2 (52.Nf6+ Kg7 53.Rf2 Ra2 54.Nd5 Qd8) 52...Ra2.? The move 51?Rxd7 is not even considered as one of the top 6 moves in this position, and may be the losing move.

52.Nf6+ Kg7

[52...Kf8 53.Qh3]

53.Qh4 Bxe4+ 54.Kg3

Reshevsky points out, ?Not 54.Qxe4 Qxf6 55.Rxf6 Kxf6, with the probable outcome a draw.?? But after 54.Qxe4, Reshevsky misses 54?Rd8 as the strongest move for Black, with a possible win for Black.

54...Qd8

The engines consider this the losing move, and no comment from Reshevsky.? Black can resist better after 54?g5 55.Qxe4 Qxf6 56.Rxf6 Kxf6 57.Qf3+ Kg6.

55.Qh7+ Kf8 56.Nxe4

Reshevsky says, ?Also good was 56.Nxd7+ Qxd7 57.Qh8+ Ke7 58.Qh4+, winning the Bishop.?? But Reshevsky misses the stronger ?56.Qh6+! Ke7 57.Nxe4; 56.Nxd7+ Qxd7 57.Qh8+ Ke7 58.Qh4+ g5 59.Qxe4, winning for White, as recommended by the engines.? Even though Reshevsky says that 56.Nxd7+ Qxd7 57.Qh8+ Ke7 58.Qh4+ wins a Bishop, it is not a clear win after 58?Kf8 59.Qxe4 Kg7.? Black can still hold the position a little longer.

56...b2

This loses for Black faster.? Black can hold a little longer with 56...f5 57.Qxg6 f4+ 58.Kf3 Qh4 59.Nxd6.

57.Qh8+ Ke7 58.Qh4+ Kf8 59.Nf6

[59.Ng5! d5 60.Qh8+ Ke7 61.Qxe5+]

59...Rb7

[59...d5 60.Nxd7+ Qxd7 61.Qh8+ Ke7 62.Qf6+ Ke8 63.Qxe5+]

60.Qh8+ Ke7 61.Nd5+

[61.Nd5+ Kd7 62.Rxf7+ Kc8 (62...Ke6 63.Rf6+ Kd7 64.Rxd6+) 63.Rf8] ?1?0

home