1984-05-28 - Rossiter, Philip ½-½ Mordue, Tyson

Tyson - Another game to prove that not all the games on this page are going to be Mordue wins. This game has never been published before although Alex Easton did refer to it in his report in BCT on the 1984 Cotswold Congress. It's a battle royal in a Sicilian Najdorf. Promising junior Phillip Rossiter makes a typical sac on b5. Tyson counters with his own sac on e5. A tremendous clash of the pieces follows and at one point Tyson has all his pieces en prise! If you like complicated tactical games then this one is for you.

[Event "COTSWOLD CONGRESS, GLOUCESTER"] [Site "?"] [Date "1984.05.28"] [Round "5"] [White "Rossiter, Philip"] [Black "Mordue, Tyson"] [Result "1/2-1/2"] [ECO "B87"] [Annotator "Mordue,Tyson"] [PlyCount "82"] {Philip Rossiter, aged 15, was one of England's best young juniors at time of this game. He achieved a couple of IM norms but never the title. This game is one of the most spectacular that I have ever played. After a tense start it explodes into action and at move 30 I have all my pieces en prise!} 1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 a6 6. Bc4 e6 7. O-O {More common is 7 Bb3 b5 8 0-0 but we transpose anyway.} Be7 8. Bb3 b5 9. Be3 O-O 10. a3 Bb7 {From this point on White has the sacrifice Bxe6 fxe6; Nxe6 Q-moves; Nxf8 B(probably) xf8 . Obviously sacrifices on e6 are a common possibility in the Sicilian Defence, yet in this line this one has been played very rarely indeed. The reason why is that it's simply not that good. Black gets two active minor pieces for the Rook and it's difficult for White to get any meaningful activity for his Rooks without opening the position up, but if he does so Black's pieces will probably be even more active! In my own praxis I've only faced this particular sac twice and won both games with no particularly serious problems faced. Hence no surprise that Rossiter turns it down now and continues to do so.} 11. f3 Nbd7 12. Qe2 Qc7 13. Rad1 {The game Juliao-Sanguinetti, Pan-American Champs, Winnipeg 1974 continued with 13 Rfd1 Rfe8 14 Rd2 Rac8 15 Rad1 Qb8 16 Qf2 Ne5 17 Nce2 Nc4 18 Bxc4 Rxc4 19 c3 d5 with an advantage to Black. Shortly afterwards White incorrectly sacked an Exchange and lost in 46 moves. There are similarities to our game. Rossiter soon gets as far as doubling the Rooks and I did intend trying for ...d6-d5 but it never happened.} Nc5 14. Ba2 Rad8 15. Bf2 Rfe8 16. Rfe1 Ba8 {The sort of pass move that was prevalent in the new computer chess programs being produced at that time. Deep Rybka says the position is completely level and doesn't blink a microchip at this move.} 17. Qe3 g6 {Either with the idea of playing ...e6-e5 (followed by a rapid ...d6-d5) and ruling out Nd4-f5 as a reply, or intending . ..Be7-f8-g7 or even both. Deep Rybka says this is an error and gives 18.b4 Ncd7 19.Bxe6 fxe6 20.Nxe6 Qb8 21.Nxd8 Bxd8 22.Qd4 and evaluates the position as +0.42 for White. I stand by my earlier comments about this sort of position and as Rossiter didn't play this I assume that he was of the same opinion.} 18. Rd2 Rd7 19. Red1 Bd8 $2 {A very odd move that I can't explain. I can only assume that the Bishop was headed for b6 but if so it was better to preface the idea with 19...Qb8. Deep Rybka suggested 19...e5 followed by 20...Ne6 giving White a plus of 0.35. Personally I'd prefer playing ...d6-d5 as soon as reasonably possible after ...e7-e5 rather than give up the d5 square completely. Obviously 19...R8d8 is a possibility too and also 19...Bf8 although Deep Rybka wasn't keen on either move.} 20. Ndxb5 $1 {Another typical Sicilian sacrifice. White gets usually three Pawns for his piece, an active position and three connected passed Pawns for the endgame. In this case the capture on d6 attacks the e8 Rook and the c5 Knight so Black will be obliged to play ...Rd7xd6, giving us a material imbalance of Rook and three Pawns for two Knights. Given that the position is wide open the Knights aren't going to be very useful and Deep Rybka says White is better by 1.2 to 1.3. Black now has be careful yet at the same time ensure he gets active play for his own pieces, a precarious balancing act but one Sicilian players are used to. In this game both players give a high-wire performance to be proud of, and it certainly caught the attention of the other competitors.} axb5 21. Nxb5 Qb8 { Black is going to have to play ...Rxd6 anyway so the Queen chooses b8 so as to have the option of ...Qxb2 later.} 22. Nxd6 Rxd6 23. Rxd6 Ncxe4 $1 {Sensibly deciding to return one of the Knights for two Pawns, activate the Bishop on a8, and introduce tactical possibilities on the a7-g1 diagonal. The b2 Pawn hangs as well. Rossiter, a well-known time-trouble addict even then, now has a lot more calculating to do having spent a while over his 20th move.} 24. fxe4 Nxe4 {Deep Rybka gives some time 24...Bc7. It's a nice thought, particularly in view of a possible alternative for White next move. The reply is obviously 25 e5 and I'd be reluctant to play 25...Bxd6 giving White an advanced passed Pawn and surrendering the dark-squared Bishop, especially with the weakening ... g7-g6 having been played. Deep Rybka agrees and analyses 24...Bc7 25.e5 Qb7 26. Qh3 Qxb2 27.Rxe6 fxe6 28.Bxe6+ Kh8 29.exf6 Qxf6 when White has two extra Pawns and a plus of 1.46. Not what Black is looking for.} 25. Ra6 $2 {Rossiter holds on to both his material and control of the b6 square. This is understandable. If say 25.R6d3? Nxf2 26.Kxf2 then 26...Qxh2 is ticklish but 26...Bb6 is great for Black. Deep Rybka suggests 25.Rxd8! Rxd8 26.Rxd8+ Qxd8 27.Qd4 and says that White is +1.40. Obviously 24...Bc7 would have avoided this. The engine also suggests 25 Qb3 and if Black avoids the Queen swap with 25...Qc8 then it insists on multiple exchanges with 26.Rxd8 Rxd8 27.Rxd8+ Qxd8 28.Qd3 with a similar evaluation to the above line. However, it's not clear that the Rook is on a good square on a6, even if there is the possibility of Rxa8, and Deep Rybka says that now White is only +0.45. With ...Bb6 ruled out I decided to kick White from a different direction.} Bg5 $6 {Deep Rybka prefers 25...Bc7 giving 26 h3 Bh2+ 27 Kh1 Bf4 and this does seem more accurate than my more direct choice. White could have replied to this with 28 Qb6 for example - if 28...Qe5 29 Bd4 with a strong position, extra material not withstanding.} 26. Qe2 $6 {A slightly passive move but it deters Black from playing 26... Qxb2?? because of 27 Rxa8 Rxa8 28 Qxe4 with an extra piece.} Bf4 27. g3 {After 27 h3 Bh2+ 28 Kh1 Bg3! 29 Be1 (not 29 Bxg3?? Nxg3+ forks - the point of inserting 27...Bh2+ 28 Kg1) 29...Qf4 and Black has an active position with many pieces aimed at the White King. Indeed the pressure persuades Deep Rybka not only to play 30 Rxa8 here but also at move 27! The text looks clear enough but Black switches to another new angle. It was around here that I began to understand that although Rossiter was one of our bright new things it might be the case that I was better at tactics than he was.} Ng5 $1 {With the idea of 30 gxf4 Nf3+ 31 Kf1 Nxh2+ 32 Ke1 Nf3+ with a perpetual check and White can't afford to walk on to the h1-a8 diagonal as ...Nd4+ wins his Queen, Rxa8 being answered by ...Qxa8+! What's more Black threatens both 30...Bf3 skewering and 30...Nh3+ 31 Kf1 Qb7 when 32 Rxa8? is no good after 32... Qh1+ 33 Bg1 Qxg1 mate. I was fairly sure that Rossiter had missed 29...Ng5 and that any result was now possible including a Black win! White's extra material on the Queenside is of little account and Black's King is much safer than the White monarch.} 28. Rd3 $2 {Coping with some threats but not all of them. Deep Rybka analyses 28.Ba7! Qxb2 29.gxf4 Nh3+ 30.Kf1 Qxa2 31.Rad6 Qxa3 32.Rd8 Bc6 33.Be3 Rxd8 34.Rxd8+ Kg7 35.Qd3 and gives White +1.3 or so. For those who don't want to allow...Qxb2 it offers 28.Rb6 Qc8 29.Kf1 with a +1 but it's still a messy position. After 29...Bf3 30 Qc4 Bxd1 31 Qxc8 Rxc8 32 gxf4 Nf3 33 c3 Nxh2+ 34 Kg2 Ng4 Black has regained all his material and the engine still insists on +1 for White. After the text Black dives in on b2. Is it poisoned of not? Well, Deep Rybka says Black is 0.4 to 0.5 better if he throws in ...} Qb7 $1 {...first. Now mate is threatened and a6 is en prise. If the immediate 28...Qxb2 29 Rxa8 deals with most problems immediately.} 29. Rd5 {If 29 Bd5 Nh3+ 30 Kg2 Qxa6 31 Bxa8 (not 31 gxf4?? Nxf4+ wins the Queen! Nor 31 Kxh3? Bxd5 32 gxf4 Bc4 and Black comes out the Exchange up.) 31...Nxf2 32 Kxf2 Bc1! and after 33...Bxb2 Black has a slightly better position due to White's broken Pawns.} Qxb2 30. Qb5 $1 {At this moment all of Black's pieces are en prise (!!), a factor that didn't go unnoticed in the tournament hall. Instead both: A) 30.Rxa8 Rxa8 31.Rd3 Nh3+ 32.Kg2 Nxf2 33.Qxf2 Qxa2 after the recapture on f4 Black is taking on c2 or a3 with an extra Pawn. Or B) 30.gxf4 Nh3+ 31.Kf1 (not 32 Kg2 Nxf4+winning the Queen) 31...Bxd5 32.Bxd5 Nxf4! again coming out a Pawn up. So the text is White's only practical chance. Obviously Black can't trade Queens so he must continue going forward somewhere. } Qc1+ $2 {Instead 30...Nh3+ keeps an advantage. After 31.Kf1 (Not 31 Kg2 or h1 Qxb5 and the d5 Rook is pinned.) 31...Qa1+ 32.Kg2 (or 32 Bf2 Rb8 33 Rb6 Qf6!? 34 Rxb8+ Bxb8+ 35 Kh3 Qd8! 36 Kxh3 exd5 yields Black a small plus as well) and now Black is effectively a whole tempo up on the game. After 32... Rb8! play continues 33.Rxa8 Rxa8 34.Kxh3 (34 gxf4? Nxf4+ and 35...Nxd5 wins material. This is the difference between this line and the game.) 34...Qxa2, again with a slightly bigger plus to Black of 0.46. I think that in my analysis I missed the game move 32 Bb7.} 31. Kg2 Bxd5+ {If 31...Rb8 32 Rxa8! gives White the advantage back.} 32. Bxd5 Rb8 33. Bb7 {White also has 33 Ra8.} Be3 {This move forces a draw although it needs some accurate calculation. There is an interesting alternative that I didn't consider which goes 33...Ne4 34.Qb4 Nxf2 35.Qxf4 Qxf4 36.gxf4 Rxb7 37.Kxf2 and once again Black has a better ending because of White's split Pawns. However, Deep Rybka spotted a fascinating alternative: 34.Rxe6!? fxe6 35.Qc4 Nxf2 (or 35...Ng5 36.gxf4 Nf7 37.Qxe6 Qxc2 [not 37...Rxb7? 38.Qc8+] 38.Bd5 Rf8 39.Qd7 Qf5 40.Qxf5 gxf5 41.a4 with a drawn ending apparently. As Black I might be worried about losing this because of the passed Pawn supported by the Bishops.) 36.Qxe6+ with a perpetual check.} 34. Qe5 $1 {White decides to make Black take the draw. I have play around the White King after 33 Bxe3 Qxe3, and swapping rooks as well with 34 Ra8 doesn't bode well. Queen and Knight are better than Queen and Bishop especially when play is confined to one side of the board, and the c2 Pawn goes very quickly with checks.} Bxf2 35. Qxb8+ Kg7 36. Qe5+ f6 37. Qc7+ { 37 Qe2?? Qg1 mate is an embarrassing finish.} Kh6 38. Kxf2 Qd2+ 39. Kf1 {If 39 Kg1 Qe1+ 40 Kg2 Qe2+ 41 Kg1 Nh3+ 42 Kh1 Qf1 is mate. If Black tries 39 (41)... Nh3 then 40 (42) Qc5 is the answer.} Qd1+ 40. Kf2 Qd2+ 41. Kf1 Qd1+ 1/2-1/2

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