1984-10-27 - Mordue, Tyson 1-0 Ashby, Alan
A win from the Yeovil Open which was Tyson's first tournament win outside Bristol. In a Sicilian Scheveningen he builds steady pressure against Alan Ashby's King. There are some useful insights by Graham Burgess in the notes. At the end Black misses one chance to cling on, then gets crushed by the local superiority of the White army.
[Event "YEOVIL OPEN"] [Site "?"] [Date "1984.10.27"] [Round "2"] [White "Mordue, Tyson"] [Black "Ashby, Alan"] [Result "1-0"] [ECO "B42"] [Annotator "Mordue,Tyson"] [PlyCount "69"] [EventDate "1984.??.??"] {This was my second win over Alan Ashby, arguably Bristol's strongest player at the time, and it helped me on the way to sharing first place, my first Open congress victory. Original notes by Graham Burgess marked GKB from his report on the event in Bristol Chess Times, Issue 46, November 1984.} 1. e4 c5 2. Nf3 e6 3. d4 cxd4 4. Nxd4 a6 {"This flexible approach keeps Black's options open and avoids some of White's sharper options. However, White can delay playing Nc3 so as to reserve the possibility of a Maroczy bind with c2-c4, especially if the Black Queen moves to c7. In that case White's next moves would be c2-c4, Nb1-c3 and Rc1 threatening Nd5." GKB. This assumes that one of Black's moves has been ...Nb8-c6 in which case after ...e6xd5, c4xd5 said Knight is pinned. Playing Nc3-d5 is usually quite strong if allowed in a Maroczy bind. However, there are exceptions. The game Cuellar-Karpov, Leningrad Interzonal 1973 featured Cuellar playing Nd5 at a time when Karpov had five units controlling that square, yet he still couldn't take it. Later Cuellar over-pressed with the attack and went on to lose.} 5. Bd3 Nf6 6. O-O d6 7. Be3 Be7 8. f4 O-O 9. Nc3 {"Thus Tyson declares his intentions and [Black] can develop his Queenside as he wishes." Playing a Maroczy bind with f2-f4 usually doesn't go so well, so by transposition we now have a Scheveningen variation of the Sicilian Defence.} Qc7 10. Kh1 b5 11. a3 Bb7 12. Qf3 {The Queen can also go to g3 via e1.} Nbd7 13. Qg3 Rfe8 $6 {"Somewhat passive. Alan rejected 13.. .Nc5 because he feared f4-f5. The position is somewhat delicate but our (Tyson's and my own) analysis implied that Black was then okay. In any case Tyson intended to play 14 e5." GKB. The point of 13...Rfe8 would be to give extra support to the e6 square after f4-f5 and an exchange on e6. Nevertheless some players would keep the Rook on f8 to counter the influence of White's f1 Rook down the newly-opened file. As ...Rfe8 weakens f7 then I'd definitely want to play e4-e5 as the likelihood is that there will be a swap on e5 and the f-file will be opened and I'd want the Rook back on f8! In other words Black should wait to play ...Rfe8 until after f4-f5 and only then if he is worried about the e6 square. Deep Rybka's preferences are 13...Nc5 and 13.. .Rac8, the latter being a standard move in many Sicilian variations and, as the two Queens are potentially vis-a-vis, defending Her Majesty might be a good tactical point. There is also a regrouping with ...Rac8 and then ... Qc7-b8-a8 which features occasionally. In fact it appears that 13 Qg3 may not be an accurate move. 13 Rad1 or 13 Rae1 may be better. In the only game I could find with the position after 13...Nc5 on the database, namely Balais-Ballo, French team event 1998 White chose not to play 14 e5? dxe5 15 fxe5 because 15...Nh5! wins the e5 Pawn! Instead White retracted his previous move with 14 Qf3 and after 14...Rac8 15 f5 e5 16 Nde2 Qc6 17 Ng3 Nxd3 18 cxd3 d5 Black had a nice position and went on to win in 36 moves.} 14. Rae1 Bf8 15. Bc1 $6 {"Typical Tyson -simple and aggressive. He wants to force through e4-e5, the Bishop is in the way, so he moves it. However, 15 Bd2 is probably more accurate since c1 is then more easily available for a Rook in the event of the c-file being opened." GKB. Deep Rybka agrees saying the difference is worth 0.15 in White's favour, and evaluating the position after the text at about -0.1.} Nc5 16. e5 Nxd3 {Note the value of Black's 14...Bf8. If this Bishop was still on e7 then White could now play 17 exf6 threatening mate on g7, and after 17... Bxf6 18 Qxd3 he comes out a piece up. However, if Black swaps on e5 before taking on d3 then after exf6 Black can swap Queens and avoid material loss. This motif comes up frequently in Scheveningens with ...Qc7.} 17. cxd3 Nd7 { Black deliberately refrains from playing ...d6xe5 in the hope that White will play the poor e5xd6. As playing e4-e5 is a major move in Scheveningens it was always unlikely that White would do so hence Alan swaps eventually. Note that Black playing a subsequent ...d6-d5 releases all the pressure on e5 and gives White a free hand on the Kingside. See Mordue-Dobrov, Caerleon 2006 for a prime example of a GM getting it wrong and being overwhelmed.} 18. Ne4 {"The Knight must be eliminated otherwise it jumps in where it feels like. However, White has another Knight, the triumphant march of which can be arranged. Often in the Sicilian each Black Knight has to 'mark' its White counterpart. In this game, by exchanging one Knight on d3 for a Bishop Black has made this [marking] more difficult, and has also been forced to exchange off his light-squared Bishop, so often the source of Black's counterplay in such positions. Thus, unable to find significant counterplay and outnumbered on the Kingside, Black's defence is practically impossible." GKB. A wonderfully insightful comment. The swap ...Bb7xNe4 also happened in Mordue-Dobrov referred to above, except White still had his light-squared Bishop and this was the piece that recaptured on e4. Dobrov had problems with Bxh7+ in some lines shortly afterwards. Also in that game both Black Knights failed to track their White counterparts. The other White Knight found a home on f5 and the crowning move of the game was the sacrifice Nf5xg7. In that game I would have been drawing on my experiences in other Scheveningens I'd played and the current game would have been one of them.} Bxe4 19. Rxe4 Nc5 {"On 19...dxe5 20 Rxe4 Nc5 21 Rg4." GKB.} 20. Re3 dxe5 21. fxe5 Rad8 22. Ne2 $6 {More accurate is 22 Qh4 to avoid Black's next. Deep Rybka evaluates this position as level, but one of the engines faults is underestimating building attacks against the King where concrete threats appear beyond it's calculating horizon.} Nb3 23. Ref3 Nxc1 $2 {Probably a rushed move because Alan was getting short of time. 23...Qc2! is a nice move to play in this sort of position. Right now it attacks the Knight on e2, an incidental, and it also prepares ...Nxc1 followed by ...Qxb2 but, possibly most important if White plays d3-d4 for any reason, Her Majesty defends h7 from a safe distance should that square come under siege.} 24. Rxc1 Qb7 {"Essential was 24...Qb8 so as to tie White's Queen down to the defence of the Pawn on e5. Even then White maintains a very strong position by advancing the Pawn to d4 to relieve the Queen of her temporary immobility. It should be mentioned that Alan was becoming very short of time." GKB. Deep Rybka has no difference in its evaluation of either 24...Qb8 or 24...Qb7, which is roughly 0.25 to White. Graham's point is valid subject to White's d3-d4, but possibly better than both other Queen moves is 24...Qb6 looking to come in at d4 at a convenient moment and maybe hold up d3-d4 completely.} 25. Nc3 Rc8 26. Rcf1 Rc7 27. Ne4 Rec8 28. Qg5 $5 {"28 Nf6+ Kh8 29 Nxh7! leads to a winning endgame for White since White's attack wins the piece and a few Pawns back, but fails to mate the Black King which can run to the Queenside. At any rate that was the conclusion reached in the post-mortem. Tyson's move, which prevents ...Rc1, makes mate unavoidable." GKB. Deep Rybka's first reaction to 28 Nf6+ Kh8 29 Nxh7 was to surrender its Queen with 29...Qxf3. It will take a while for White to win the game but we can take this as an endorsement of the Knight sac. The analysis of taking the Knight goes as follows: [28.Nf6+ Kh8 29.Nxh7] 29...Kxh7 30.Qh4+ Kg8 31.Rh3 f5 32.exf6 gxf6 33.Qh8+ Kf7 34.Qxf6+ Ke8 35.Qxf8+ - at this point Deep Rybka gives White +13.63! and the rest of the line is a typical piece of engine analysis with the Black King being chased by a Queen and Rook until it gets to the wrong square! - 35...Kd7 36. Rh7+ Kc6 37.Rc1+ Kb6 38.Qf2+ Ka5 39.Qe1+ Kb6 40.Qg1+ Ka5 41.b4+ Ka4 42.Qd1+ Kxa3 43.Ra1+ Kxb4 44.Qe1+ Kc5 45.Qe3+ Kd6 46.Qf4+ Kc6 47.Rc1+ Kb6 48.Qd4+ Rc5 (if 48...Ka5 49 Ra1 mate) 49 Rxb7+ wins ... finally! As this is a not-too-difficult of analysis up to 35 Qxf8+, I must assume that either I didn't fancy the grind after 29...Qxf3, or that I thought I was heading into a Queen ending when all the Rooks come off at say move 37 or thereabouts. Again probably a grind coming up. Whatever 28 Nf6+ is probably the perfectionist's choice. 28 Qg5 is a practical gamble with the odds heavily in my favour. As Graham points out it prevents ...Rc1. It also contains a very subtle idea which is 29 Nf6+ Kh8 30 Rg3 followed by 31 Qh6!! The threat is mate on h7 and 31...gxh6 is met by 32 Rg8 mate while 31...gxf6 32 Qxf6+ is mate next move. It seems that the only defence is the immediate 28...Kh8, whereupon it's not easy for White to continue his attack without weakening his back rank, e.g. 29 Rg3 Qa7 30 Nf6 Qf2! or 30 Qh4 Rc1. Or 29 Qh5 h6 30 Ng5 g6 31 Qh4 Bg7! and White can't take on f7 three times because of ...Rc1 mate. However, Black has to work all this out and he was short of time. Regardless of White's missed win the above lines seem to indicate that Black has placed his Rooks on exactly the right squares to meet White's attack. However, Alan's next move undoes all the previous work, and it just happens to be a Pawn to Rook Three! In the game with Dobrov referred to earlier I also had a chance to sacrifice on h7. My analysis correctly showed that Black could give up his Queen, and a Pawn, for a Rook and Knight to break the attack. So I chose not to sac and instead maintained the pressure. This approach paid off when my opponent blundered a few moves later. I did not recall the current game at that time but the parallels are intriguing.} h6 $4 29. Nf6+ Kh8 30. Rh3 {Now there is no defence to the threat of 31 Rxh6+ no matter how Black plays. The local superiority is overwhelming.} gxf6 31. Rxh6+ Bxh6 32. Qxh6+ Kg8 33. exf6 { "Black is reduced to spite checks." GKB. I was actually surprised by the next move as I was expecting a resignation.} Qxg2+ 34. Kxg2 Rc2+ 35. Kh1 1-0
The games viewer used here was developed by
http://chesstempo.com
and is licensed under
CC BY 2.0.
