1985-01-20 - Mordue, Tyson 1-0 Tarrant, G
Yet another Bristol League Congress game. This could have been the upset of the event. After a strong opening Tyson dissipates his initiative and then blunders a piece! However, the consequence is that Tarrant's Kingside is shattered and White gets a big Pawn on f6, a real bone in the throat. Tarrant's extra Knight canters into the White Queenside and is unexpectedly trapped. Black misses his best line and Tyson gets to keep his f6 Pawn. Adjournment analysis shows Tyson the best way forward and when Tarrant, instead of staying passive, tries to go active he is executed immediately with a tidy finish. Another game well worth the study.
[Event "BRISTOL LEAGUE CONGRESS"] [Site "?"] [Date "1985.01.20"] [Round "1"] [White "Mordue, Tyson"] [Black "Tarrant, G"] [Result "1-0"] [ECO "B00"] [Annotator "Mordue, Tyson"] [PlyCount "103"] {In the previous two League Congresses I had tied for first place but lost out on tie-break for the title of Bristol Champion. The tie-break used was Sum Of Opponent's Scores (SOS) and, as I'd drawn my first game in each of those events, I'd always been playing catch-up. Indeed one of those first rounds games had been, with Black, against the opponent in this encounter so I was determined to win this particular one. Although the game started on Friday evening, because of the adjournment it wasn't until Saturday afternoon that actually I won my first round game! This game was in BCT issue 50, Summer 1985 with the introduction "Notes by APA (Alan Ashby), exclamation marks and question marks by TM.} 1. e4 e6 2. d4 b6 {The Queen's Fianchetto Defence, also known as Owen's Defence which Tony Miles, amongst others, had helped popularise in the mid 1970s. Nothing to do with the French Defence which by 2.. .d5 immediately challenges in the centre. In the game Black snipes at the White position from a distance in the opening. He continues to do this for the whole game, yet when he finally plays the grand centralising move ...Qd4, his position immediately collapses!} 3. Nf3 Bb7 4. Bd3 Nf6 5. Nbd2 c5 6. O-O {The most common move here is 6 c3 to bolster the centre but this might transpose into a French Defence with ...d7-d5, e4-e5. Of course this might still happen. I was giving Black the chance to go into a sort of Sicilian with 6...cxd4 7 Nxd4, which featured in three games on the database, although this was unlikely as my opponent wasn't a Sicilian player and I was. Ironically, in the previous year's event in Round 3 my opponent Dave Osborne had chosen a variation of the Sicilian featuring a rapid Queenside fianchetto and he got chopped up very quickly. There was also one other possibility which my opponent actually played into.} c4 $6 {This was played in Marzolf-Iglesias, Metz Open 2006 which continued 7 Nxc4 Nxe4 8 Re1 Nf6 9 Bg5 d6 and Black had a passive Hedgehog-like position. After a long struggle where Black had an extra Pawn at one point the game was drawn in 69 moves.} 7. Bxc4 Nxe4 {I was intrigued to see that Deep Rybka already dislikes the Black position so much that it considered not taking the e4 Pawn!} 8. Nxe4 Bxe4 9. Re1 {Originally I gave this an ! I wanted to push the Bishop back before I played my next move. However, it seems that the immediate 9 d5 is more accurate. See the next note.} Bb7 10. d5 $1 {"Black's sixth move successfully forced the exchange of Black's c-Pawn for the White e-Pawn, but Black had probably underestimated this move which cuts his position in two. The exchange ...e6xd5 would always leave Black with a weak isolated d-Pawn, and Black will also have difficulty developing his Queen's Knight since ...d7-d6 leaves the e-Pawn en prise. Perhaps 9 d5 was even stronger to slow down Black's chance of playing ...Na6." APA. Deep Rybka's evaluations of the positions after 9 d5 and 10 d5 confirm that Alan was right. In fact I'd already looked at 6...c4 previously, concluded that it was a poor move and thought that the inclusion of 9 Re1 improved the position!} Be7 11. Bf4 {11 d6 does not serve White well here. It closes the d-file from White's perspective, gives Black back the c6 square, opens the line of one Bishop and pushes the other to a better square on f6. I have a clear memory of a game Rumens-Basman, published in Chess magazine, in the mid 1970s where White played d6 in an analogous position, got an awful position and only won when Black blundered in time-trouble.} O-O 12. Qd3 Qc8 $5 {Deep Rybka likes this move attempting to get pieces out via a6.} 13. Ng5 {At the time this looked good but White's initiative starts to fade here. It may simply be a case of right square, wrong piece. Instead after 13 Bg5 Bxg5? 14 Nxg5 g6 15 dxe6 followed by 16 N or Rxe6 wins. Better is 13...Bc5 after which Deep Rybka gives 14 Bh4 with a +1 evaluation. Also there is 13...f6 upon which Deep Rybka analyses 14.dxe6!? d5! 15.Bxd5 Rd8 16.c4 Bxd5 17.cxd5 fxg5 18.Qf5 Rf8 19.Qe5 Bf6 yet still gives White +1 despite his piece minus. Worth investigating further, but this is very ironic in view of what happens in the game where White shortly gets a Pawn to the sixth rank and remains there until the finish. } Bxg5 {"Black is forced to concede the two Bishops, sinc3 13...g6 allows 14 dxe6 dxe6 15 Nxe6 winning material." APA.} 14. Bxg5 Ba6 {"When an opponent has the two Bishops it is always a good idea to try to exchange off one of them, so the move played is far better than the alternative of 14...Na6." APA.} 15. Bxa6 $2 {Deep Rybka claims that White keeps a plus of nearly 1 after 15 Re4 f5 16 Rh4 Bxc4 17 Rxc4 Qb7 18 dxe6 dxe6 18 Rd4!} Nxa6 16. c3 $2 {Obviously to keep the Knight out of b4 but White would do better to play 16 Bd2 then follow the plan outlined in Alan's note to my 20th. What happens now is that I target the right square from the wrong direction and lose not only the advantage but worse.} Nc5 17. Qg3 {17 Qd2 is better but I was too focused on the idea of Kingside pressure. White has a Queenside majority and should try using it.} f6 18. Bh6 Rf7 19. Rad1 $6 {"Black has defended well. White's best try is now 19 dxe6 dxe6 to keep the position open for the Bishop, though White would have little if any advantage." APA.} e5 {According to Deep Rybka Black now has a small advantage of +0.14. However, White has a space advantage, a Bishop against a Knight, more actively placed major pieces so he must be better. Regardless of the differences in evaluation I didn't improve my position with a simple one-move blunder.} 20. f4 $4 {"A terrible oversight. The plan of c2-c4, b2-b3 then regrouping with Bh6-c1-b2 and finally f2-f4 would keep up the slight pressure." APA.} Kh8 {This routine unpin wins a piece because White has just cut off the Bishop's retreat. It was rare that I blundered in those days and this was down to a simple lapse in concentration. However, Deep Rybka's evaluation is interesting as it barely gives Black more than +0.5. White gets two Pawns for his piece, one of them a bone in Black's throat for the remainder of the game, while Black's solid Kingside majority of Pawns becomes an almost useless minority wreck.} 21. fxe5 gxh6 22. exf6 Qf8 $6 { White is very lucky here that Black can't take on f6 so he gets to keep his 'bone in the throat' for a little while and not have to resign yet. The fairly simple variation goes 22...Rxf6? 23.Qe5 Kg7 24.Rf1 Qf8 25.Rd4! - threatening 26 Rg4+ and the first time in the game that White's superior manoeuvring space comes into play. It won't be the last. - 25...h5 26.Qg5+ Rg6 27 Rxf8 Rxg5 28 Rxa8 and wins an Exchange and the endgame. Black misses the same trick twice in a few moves here. He has control of g5 and should use it to activate his Queen by 22...Qg8! 23 Qe5 Qg5!} 23. b4 {I hadn't yet invented the adage "A Knight is always bad on Knight two!" then, but it's going to take Black time to get that steed back in the game. Positionally White stands excellently and Deep Rybka only has Black at +0.79.} Nb7 24. Rf1 {The inclusion of the last few moves has made no difference to the tactic that refutes 24...Rxf6??.} Re8 { Better is 24...Qd6 25.Qf2 Raf8 26.Qd4 Rxf6 27.Rde1 - with the idea of the effective 28.Re8!? - 27...Kg8 and, of course, Black has improved his position considerably.} 25. Qh4 Nd6 26. Rd4 Re5 27. c4 Nf5 {"With 27...Ne8 followed if necessary by ...Re5-g5-g6 [utilising the control of g5 again. TM] victory would be well in sight. But the move played lets White back into the game." APA. I originally gave this move an ? but it's actually okay. The problem is the sequel. Of course Alan's suggestion is okay as well. However, I'm confident that when Tarrant played 27...Nf5 he was already committed to playing 28...Nxd4.} 28. Qh5 $2 {And this move I gave an ! when in fact 28 Qf4 with the same idea is more accurate.} Nxd4 $4 {Instead 28...Rxf6 29 R4f4 Qxb4! and now 30 Rxf5?? Rxf5 31 Rxf5 Qb1+ wins. Also 28...Qe8 29 R4f4? Ne3 attacks the Queen and f1 Rook at the same time. Tarrant, who had begin to run short of time to the control at move 42, assumes that with the text he's simplifying but he's made a subtle oversight.} 29. Qxe5 Nc2 30. a3 $1 {To Tarrant's horror he finds that I'm trapping the Knight. The steed seems to have plenty of space but it's an optical illusion, and one that I had played for. Ironically the trap is a Pawn to Rook Three, a move that I usually criticise for various reasons. No reason to complain here but what is the assessment of the position once I've recovered the piece? At this point I got up from the board and left Tarrant to stew. He spent a lot of his remaining time over the next move and, although it's the simplest choice, he was clearly disconcerted and didn't make the best move.} Nxb4 $6 {"Black must be careful. The safest line is probably 30...Nxa3 31 Qc3 a5! (better than 31...Nxc4 32 Qxc4 Rxf6 when White has 33 Rxf6 Qxf6 34 Qc8+ or 33 Qd4) 32 Qxa3 (32 bxa5 Qc5+) 32...axb4 with at least a draw. Also worth considering is 30...b5 31 Qc3 Nxa3 32 Qxa3 bxc4." APA. In the line 30...Nxa3 31 Qc3 Nxc4? 32 Qxc4 Rxf6 33 Rxf6 Qxf6 34 Qc8+ Kg7 35 Qxd7+ Deep Rybka says White is +0.5. Black may have an extra Pawn but the h-Pawns are doubled and useless. The alternatives 33 Qc3 or 33 Qd4 are regarded as equal. In the same line 30...Nxa3 31 Qc3 a5 32 Qxa3 axb4 seems well met by the remarkable 33 Qa1! when Deep Rybka says White is +1 or so. His Pawn structure is superior, Black is hamstrung by the weaknesses of both back rank and the a1-h8 diagonal and White's pieces are better placed as well. Yes, even on a1 the White Queen has more scope than the black one and the open a file helps. Of course Black has 32...Qxb4 but after 33 Qg3 we have the same structure as in the game and Black will have exactly the same problems. In the alternative 30...b5 31 Qc3 Nxa3 32 Qxa3 bxc4 33 Qc3! lands Black in the same pickle. Deep Rybka says White is nearly +1.4 here. In summary Black's best option is the one that leads to that Queen and Pawn ending. Deep Rybka and other engines are great at major piece endings with lots of open lines. They can easily calculate further than humans and judge positions more accurately than they do. This is a prime example although I can see Tigran Petrosian smiling at the line with 33 Qa1. Hence it's no surprise that some of Alan's suggested lines are not so sound in the light of modern computer analysis and, by the same token, why Tarrant took as long as he did and still didn't find the best move. Overall White's superior Pawn structure coupled with superior mobility stands him in good stead as I discovered during my adjournment analysis. Black's only plus is that he has an outside passed Pawn, so that limits White options to the extent that he may not, for example, be able to go into King and Pawn endings easily.} 31. axb4 Qxb4 {At this point Deep Rybka said that White is +1 or so. A handy plus in a position where the material is actually even. At this point I thought it best to avoid any back-rank tricks. Black can't do the same as he has two h-Pawns! In the moves leading up to the time-control at move 42 I gradually improve my position without being too committal. Essentially Black merely plays ...Qf8 and ...d7-d6. The latter concedes the e6 square to White. Having said that the advance d5-d6 may have been troubling Tarrant and e5 was a very good square for the White Queen. It so happens that e6 is even better.} 32. h3 {Not 32. Qb8+ Qf8 33.Qxa7 Rxf6 34.Rxf6 Qxf6 35.Qxd7 as 35...Qd4+ will be a perpetual check.} Qc5+ 33. Kh1 Qf8 34. Qb2 d6 35. Qd4 Qd8 36. Qf4 Qf8 37. Kh2 Kg8 38. Qg4+ Kh8 39. Qe6 Kg8 40. Rf3 Kh8 41. Rf4 Kg8 42. Rg4+ Kh8 {"With time pressure over, Black has a solid, if passive, position." APA.} 43. Rf4 {The sealed move. The adjournment merely convinced me that although Black is badly tied down if I was going to win this then I had to advance the Kingside Pawns. I'd deliberately refrained from doing so before the adjournment in case of tactical surprises. It turns out that there aren't any and, for the moment, the best rank for White Rook is the third. The winning tactic also turned up quite a few times in my analysis as I suspect it had in my opponent's. I assume Tarrant decided not to advance his Queenside Pawns because they would become weak. It only takes White a handful of moves to support f6 with a Pawn on g5 so King marches to the Queenside become feasible. However, it's more likely that any attempt to play ...b6-b5 is met by c4-c5 and the creation of another passed Pawn should stretch Black to breaking point. Worth looking at though.} Kg8 44. Rf3 Kh8 45. g4 Kg8 46. Kg3 Qd8 47. h4 {With every advance of a White Pawn Deep Rybka increases its evaluation of White's position. Now it's nearly up to 1.5.} Qc7 48. g5 hxg5 49. hxg5 {Now that White has secured his f6 Pawn he has numerous options with his Queen and Rook. The engine goes up to over +2.0 in its evaluation.} Qc5 $2 {"Black must remain passive with ... Qd8-f8, though White might still be winning with Kg4-f5, Rg3 [and] g6. The line played allows White to finish spectacularly. APA." To emphasis that point 49...Qd8 50 Qe2 h6? - attempting to undermine f6 - 51 Re3 Ke8 52 Qh5! and the invasion down the h-file finishes Black off. So Black can't even move his h-Pawn. My plan was to put the Rook on e4, play Qh5 and then g5-g6 wins. The variations were ...Rxf6, Qxh7+ Kf8, Qh8 mate or ...hxg6, Qxg6+ Kf8, Qh6+ Kg8, Rg4+ winning or, assuming the Black Queen is on c7, ...hxg6, Qxg6+ Kh8, Rh4+ Rh7, f7! wins. In all these line the f6 Pawn's control of g7 is vital as it is in the game's finish.} 50. Re3 Qd4 $4 51. g6 $1 {At which point my opponent exclaimed "Oh, no! I haven't fallen for that trick, have I?" Simply, yes. White is winning the Rook and I've been careful to make certain that my King isn't on a square where the reply was 51...Qxf6 with check.} hxg6 52. Qe8+ {Tarrant must lose the Rook with check. He had overlooked after 52... Rf8 the switchback 53 Qxg6+ and mate next move.} 1-0
The games viewer used here was developed by
http://chesstempo.com
and is licensed under
CC BY 2.0.
