1986-11-01 - Copeland, Graham 0-1 Davies, M

In the rare Alatortsev variation of the Queen's Gambit Graham works up a winning position very quickly. Unfortunately he fails to find the easy kill and his position starts to fade. Eventually he misses a draw but the finish by Martin Davies is one to remember.

[Event "BRISTOL LEAGUE CONGRESS"] [Site "?"] [Date "1986.11.01"] [Round "5"] [White "Copeland, Graham"] [Black "Davies, M"] [Result "0-1"] [ECO "D44"] [Annotator "Mordue,Tyson"] [PlyCount "94"] [EventDate "1986.??.??"] {"Original notes by Alex Easton (AE) in BCT Issue 62, December 1986.} 1. d4 d5 2. c4 e6 3. Nc3 c6 4. Nf3 Nf6 5. Bg5 dxc4 6. e4 b5 7. e5 h6 8. Bh4 g5 9. Nxg5 Nd5 {"The Alatortsev variation. More usual is the main line of the gambit 9... hxg5 10 Bxg5 Nbd7 etc." AE.} 10. Nxf7 {"The main attempt at a refutation. Black gets the initiative and chances of winning the Knight on h8 for his Exchange. Safer for White is 10 Nf3 Qa5 11 Qd2 Bb4 12 Rc1 Nd7 13 Be2 etc." AE.} Qxh4 11. Nxh8 Bb4 12. Qd2 Nd7 {This move does not appear on the database. The main move is 12...c5 with two games featuring 12...Bxc3 with a single occurrence of the odd 12...Ba5. Overall the results in this line are heavily in White's favour.} 13. Be2 {"Probably a better alternative is 13 0-0-0 e.g. 13...c5 14 dxc5 Nxe5 15 f4! Qxf4 (if 15...Nc6 16 g3 Qf6 17 Bg2 or 15...Bxc3 16 bxc3 Qxf4 17 Qxf4 Nxf4 18 c6!? Nxc6 19 g3.) 16 Qxf4 Nxf4 17 Nxb5 or 17 g3." AE. Deep Rybka agrees that 13 0-0-0 is a better move but after 13...c5 it chooses 14 g3! This reinforces a forthcoming f2-f4 and also pushes the Queen off a useful square. If 14...Qg4 then 15 Be2 followed by Bh5+ ensures the Knight escapes via g6 - if 15...Qf5 16 Qc2 does the trick. A particular point is that by delaying dxc5 Nxe5 for a move or so the Black Queen doesn't get the f6 square which which it gets to take the Knight on h8. I assume Graham's move is designed to prevent ...Qg4 and the Bishop can still come to the h1-a8 diagonal via f3.} c5 14. dxc5 $2 {Deep Rybka prefers 14 0-0-0 so as to answer 14...cxd4 with the obvious 15 Qxd4 offering a swap of Queens. After 15...Qxd4 16 Rxd4 Nxc3 17 bxc3 Bxc3 18 Rh4 Bxe5 Black has picked up a couple of Pawns and 19 Rxh6?? loses to 19...Bf4+, but now the engine suggests 19 Bh5+ and 20 Ng6 with a clear advantage. It still looks complicated to me and the text does look more straightforward, but the difference in evaluation between this line and the text is about a point so the silicon chip also has a point.} Nxe5 15. O-O-O Bxc3 $2 {"Premature. Better is 15... Bb7 16 g3 Qf6 17 f4 Nxc3! 18 Bh5+! (If 18 fxe5 Nxa2+ 19 Kb1 Qf5+ or 18 bxc3 Ba3+ 19 Kb1 Bxh1. [The point of 18 Bh5+ is to drive the King to the f-file as 18...Ke7?? 19 Qd6 is mate! TM]) 18.. .Kf8 19 fxe5 [19 Qd6+? Kg8 and Black has two pieces for a Rook, White has pieces en prise all over the board and Black has some very useful checks. TM] 19...Nxa2+ 20 Kb1 Qf5+ 21 Kxa2 Bxd2 22 Rhf1 and I feel Black is better." AE. Deep Rybka and I agree with most of this. However, we disagree with the conclusion because Alex has ended his analysis too early. After 22...Qxf1 23 Rxf1+ Kg7 24 Rf7+ Kxh8 25 Rxb7 if anything White has a clear plus as all his pieces are better posted. There are a lot of vulnerable Pawns here because a majority of them are on the same colour as the opponent's Bishop! Worth analysing further.} 16. bxc3 Qf6 {"16...Bb7 is better, although White is better now - the Black King is exposed." AE. After 16...Bb7 17 Rhe1 followed by 18 Bf1 and White has pressure down the e-file. Deep Rybka suggests instead 16...Bd7 to keep the e6 Pawn protected.} 17. f4 $2 {A big improvement is 17 Qd4! Qxh8 18 Bh5+ and 19 Rfe1 with full mobilisation and all but one of Black's pieces on the edge of the board.} Nxf4 $4 {Clearly better is 17...Nd3+ 18.Bxd3 cxd3 19.Qxd3 Qxh8 and although White has plenty of tempting moves: 20 Qxb5+, 20 Qg6+, 20 Qd4 and 20 Rhe1 there is no immediate win and Deep Rybka actually suggests the passive 20 Kc2. The text opens the d-file for White.} 18. Rhf1 Nxe2+ {I can only assume that in the line 18...Nfd3+ 19 Bxd3 Nxd3+ Martin missed 20 Qxd3! (20...Qg5+ 21 Qd2). Now White Rooks control the open files to either side of the Black King and this should be fatal.} 19. Qxe2 Qg5+ { "Unfortunately Black can't take the Knight because of 20 Qh5+ - a recurring theme." AE. The point is 19...Qxh8?? 20 Qh5+ Ke7 21 Qh4+ Ke8 22 R or Qd8 mate.} 20. Kb1 Bd7 {If 20...Nd3 to block the d-file then White has a field day after 21 Qf3.} 21. h4 $1 {Deep Rybka evaluates this as +16.1 to White. He must be winning then!} Qg7 22. Qh5+ Ke7 23. Nf7 $2 {"A tragedy for White. 23 Qxe5 wraps it up e.g. 23... Qh7+ 24 Kb2 Rxh8 25 c6 mates. Black now escapes with some compensation for the Exchange." AE. The point is not to recapture the Queen but mate after 23...Qxe5 24 Ng6+ Ke8 25 Rf8 mate. Black does have 24... Kd8 but after 25 Nxe5 White is winning the pinned Bishop as well. Graham must have been in time-trouble here as he subsequently makes several other uncharacteristic errors.} Nd3 24. Qxh6 {With the Black King still in the centre it makes sense to keep Queens on and play 24 Nxh6 instead. The endgame is not easy for White, particularly with that big Knight on d3.} Qxh6 25. Nxh6 {White is +2.67 better according to Deep Rybka. What's the worth of that steed on d3 then?} Rh8 26. Rf7+ Kd8 27. Rf6 Ke7 28. Rdf1 {28 Rg6 is objectively a better move. The objection is 28...Be8 and after 29 Rg7+ Kf8 with two pieces attacked, but the engine points out 30 Rxa7 Rxh6 31 Rf1+ Kg8 32 Ra8 winning the pinned Bishop and staying the Exchange up.} Be8 29. g4 {Deep Rybka suggests the subtle 29.c6 based on the extended point 29...Bxc6 30.Rf7+ Kd6 31. R1f6 Bxg2 32.Rg7 Be4 33.Nf7+ Ke7 34.Nxh8+ (the second time this Knight has taken a Rook on h8!) 34...Kxf6 35.Rf7+ Ke5 36.Rxa7 with a winning ending. Of course Black doesn't have to take and, what's more, this is way too much for a human being in time-trouble to calculate accurately.} Nxc5 30. Kc2 $6 {Missing a subtle tactical point. 30 g5 defending vulnerable units was much better. According to Deep Rybka White's plus is a mere 0.59 now.} Ne4 31. R6f4 $2 { Once again 31 g5 is the move. After 31...Nxf6 32 Rxf6 White still has a positional advantage but no longer a material one.} Rxh6 32. g5 {A zwischenzug to defend the h4 Pawn. Black could try 32...Nxg5 but he has better due to White's faulty 30th move.} Rh8 33. Rxe4 Bg6 34. Rff4 e5 35. Rg4 $6 {"The Rook wasn't attacked! 35 Kd2 saves a tempo although it may not matter. 30 Kc2?! has had its rewards." AE. Possibly 'just desserts' is a better phrase. Now White has to think about drawing this ending. Alex's point about the tempo loss not being important is correct. After 35 Kd2 Bxe4 36 Rxe4 Ke6 37 Ke3 Kf5 38 Kf3 Rd8 39 Re2 Rd3+ 40 Re3 transposes into the game.} Ke6 36. Kd2 Bxe4 37. Rxe4 Kf5 38. Ke3 Rd8 39. Kf3 Rd3+ 40. Re3 e4+ 41. Ke2 Kf4 42. Rxd3 cxd3+ {A critical moment. Worth pausing and working out what you would play as White.} 43. Ke1 $4 {"43 Kf2 e3+ 44 Ke1 Kf3 also loses - Black Queens and stops the g-Pawn. 43 Kd2 also loses tragically to 43...Kf3 44 g6 e3+ 45 Kxd3 e2 46 g7 e1=Q 47 g8=Q Qd1 mate. But 43 Kd1 draws - 43...Kf3 44 g6 e3 45 g7 Kf2 46 g8=Q e2+ 47 Kd2 e1=Q+ 48 Kxd3." AE. The difference between 43 Kd2 and 43 Kd1 is that ...e3 is not check and Black has to spend an extra tempo playing ... Kf3-f2 so his King no longer controls e4 in the mating net. Now there comes a completely different mating net and one worth remembering!} Ke3 $1 44. g6 d2+ 45. Kd1 Kd3 46. g7 e3 47. g8=Q e2# {White finally emerges a significant amount of material up, but the object of the game is to checkmate and Black has done just that in a very picturesque fashion.} 0-1

The games viewer used here was developed by http://chesstempo.com and is licensed under CC BY 2.0.