A Very Interesting Game

Sunday night, the 10th of January at the Saratoga Club had some interesting games. There was one upset result, Jeff Hrebenach, rated around 1400 held a draw against Little at 1916. Certainly an upset on the rating difference, but considering my dwindling results over the past two years and Jeff’s jump of 99 rating points in his last tourney maybe not so shocking.
There was one surprise, it can’t really be classed as an upset - Sells with White defeated Steve Taylor. This is the first loss for Steve in the event and it came in what is the most exciting game of the year so far. It is the subject of this post.

Lee Battes (2001) disposed of David Connors (1533) but it was not a routine victory. Connors had definite chances right down to the last few moves. Alan LeCours defeated Ray Alguire and Glen Gausewitz won his game with George Dipre. Both contests were more typical victories by higher rated players.

The make-up rounds for this year’s championship run until 2/21/10.

Saratoga Championship
Saratoga Springs, NY
1-10-2010
White: Sells, P
Black: Taylor. S

A game with just about everything one could ask for in the way of excitement. Sells loses his Queen in an maybe ill-advised capture of a semi-poisoned pawn, and Taylor uses up a huge amount of clock time attempting to make his material plus good. Taylor is in serious time trouble by move 25. He manages to make more than 40(!) moves with less than five minutes on his clock. Taylor’s wonderful quick play tests Sells’ chess skills to the point his 40+ minute edge on the clock is run down to one second. Sells then gave his own demonstration on how to play fast managing a dozen moves with one second remaining. That flurry finally cracked Taylor and his time expired somewhere well beyond move 70!
1. e4 e5
2. Nf3 Nc6
3. d4 exd4
4. Nxd4 …..
The Scotch Game. This debut is not widely popular but is certainly sound. It has been tried by Kasparov no less, Svidler, and many other of the top echelon as a surprise wepon.

4.… Nf6
5. Nxc6 bxc6
6. e5 ….
The Mieses Variation. A line in which Kasparov defeated Karpov in game 16, Lyon 1990. White’s advance e-pawn is a target and Black’s c-Bishop has problems getting into the game. It took only 102 (!) moves for Garry to demonstrate the win against his redoubtable opponent. A clear illustration of the potential for the Mieses line to create long games. In Dortmund 1991 Kasparov needed only 40 moves to beat Piket in the same variation.

6.… Qe7
7. Qe2 Nd5
8. c4 Ba6
9. b3 ….
The British GM Gary Lane says this is the move that revitalized the line. Kasparov rolled it out against Karpov at Tilberg 1991 winning in 44 moves.

9.… g5!?
Karpov tried at Tilberg 9..., g6. That move is probably more common than the text in GM practice. Nikolic played the game move against Kasparov in Linares, 1997 losing in 27 moves.

10. g3 …..
Lane in his book "Winning with Scotch" Henry Holt and Co., NYC, 1993, credits Kasparov again with "revitalizing the Scotch" for finding 10 g3. True the variation refer to does not have the Black K-side pushing forward as in our game. Lane must be trying to say Kasparov did some original work to reinforce the Meies variation. That is accurate, the Scotch had made few appearances at the top level events in the last fifty years until Kasparov tried it in the 1990s.

10.… Bg7
11. Bb2 O-O-O
None of the top flight GM’s have opted for long castling here, a move or two earlier yes, but not here. Most Black players opted to castle short. White enjoyed a 5 to 2 win to loss ratio against long castling in a double handful of games between players in the 2300 to 2500 range.
Interestingly the game position is very similar to Kasparov - Karpov WC Match, Lyon, game 14. There Karpov castled long on move 9 and did not advance the g-pawn until he had built pressure on e5. A question to ponder: How these ideas mesh for Black; pressure on e5, advancing …, d7-d5; and the advance or no of the Black K-side pawns. Karpov, one of the best chess minds working when the cited game was played did not push the pawns. So, think about how the ideas compliment or clash with each other. It is terrifically helpful to have examples between such as Kasparov and Karpov to compare to the game under discussion today between very good local talent. The games referenced and this battle are well worth study if you want to play the Scotch, or if you like 1..., e5; as Black.

12. Qe4?! …..
White is over concerned about the safety of the e-pawn. Standard is to let it go to obtain a Bishop and Knight for a Rook and two pawn; 12 Bg2, Rde8; 13 0-0, Bxe5; 14 Bxe5, Qxe5; 15 Qxe5, Rxe5; 16 cxd5, Bxf1; 17 Kxf1. The results statistically favor White. Again the database has no trials between the top flight GM’s. That makes me suspect all is not so clear as the raw numbers indicate. In the 2300 to 2500 group White won three drew three and lost one in the games found. It appears the two minor pieces can make things very difficult for the Black Rooks. No examples were found of the game move. My computer sees the game as about even after the text.

12.… f5!?
Steve thought a long time about this move. It was the beginning of his time trouble. My game had ended at this point and I was able to watch this most interesting struggle closely. I guessed before he made the text move either 12..., Nb4; or 12..., Nb6; would be the choice. Lane says freeing the Ba6 is a key goal for Black. He wants to shift the White c-pawn which is most easily done by …, d7-d5 at an appropriate moment. Taylor obviously understands this. The game move seems to be wrinkle he came up with to obtain open lines pointed towards the White King at the cost of a pawn or two before executing the sequence; …, Nb4; …, d5. The try 12..., Nb6; may be playable, but the Black King is somewhat exposed after 13 Nd2, d5; 14 cxd5, Bxf1; 15 Kxf1, cxd5; 16 Qd3, c5. The computer sees the game as about even. I believe a human being would think that Black is taking dangerous chances with his King’s safety. The computer prefers 12..., Nb4; to the move played. We all know just how materialistic computers are now. The pawn offer is a powerful ploy in club level play.

13. Qxf5 ….
It is possible to play 13 Qe2, then either 13..., d6; 14 Qf3, or 13..., Rhe8; 14 Kd1, are two alternatives that White can try. Both look to lead to exciting positions where the chances are even. The game move requires very accurate calculation, but that is not beyond Phil’s capacity, and Sells is not willing to call sending his Queen to e4 wrong by bringing her back to e2.

13.… Rhf8
14. Qe4 …..
If 14 Qxh7, Qf7; 15 Bd4, Ne3!; 16 Bxe3, Bxe5; 17 Qxe7, Rxe7; 18 Bxg5, Re8; and Black has the Exchange for two pawns. This is approximately equal materiall. In an endgame the side with the minor piece and pawns might even be a bit better off according to GM Soltis. Here the simple ending is off in the future. Black has two Bishops and two Rooks with a good possibility of getting in …, d7-d5; pressuring the c-pawn, while White has not quite completed his development. That is fertile ground for growing an attack in Queenless middle game.

14.… Nb4
15. Bg2 ….
Up to here Phil had retained the usual edge one hopes to have when playing the White pieces. The text is a routine kind of move and probably played without fully realizing how dangerous things are becoming. With 15 a3, White could push back the Nb4 while making an air hole for the "pig" stuck on a1. Play could go: 15 a3, d5; 16 Qe2, Bxe5; 17 axb4, Qxb4+; 28 Qd2, and White is well ahead. Or, 15 a3, d5; 16 Qe2, Nd3+; 17 Qxd3, Bxe5; 18 Qe2, Bxb2; 19 Qxe7, Rde8; 20 Qxe8+, Rxe8+; 21 Kd2, Bxa1; 22 Nc3, when game heading towards equality.
Both of these guys strike you on first meeting them as scholarly gentlemen, polite and well behaved. After watching this game and several other games of theirs’ I think in their chess playing hearts they wear eye patches and bandanas around their heads, have parrots perched on one shoulder, say things like "Aye Matey" and figuratively fly the Jolly Rodger. At the board they are swashbucklers who court danger with the belief great risk brings great reward.

15.… d5
16. Qxh7 ….
I don’t know if Phil realized he’d have to go down this risky path when he put the Queen on e4. It seems the natural move 16 Qe2, loses after 16..., dxc4; 17 Qg4+, Kb8; 18 0-0, cxb3; 19 Rc1, Rxf2!; and while there are a number of alternatives for White they all lose one way or another. One example: 20 Kxf2, Bxe5; 21 Bxe5, Nd3!, 22 Kg1, b2; winning at least a Rook because 23 Bxb2, is met with 23..., Qe3+; and a deadly fork of the King and Queen is coming, or if Black is alert there is the "old smothered mate" possible after the discovered check by the Knight to h3 and Queen check at g1 with the Knight going to f2 finishing the job. Phil spent some considerable time making up his mind to take the h-pawn.

16.… Rf7?!
Another long think by Taylor. He must have been weighing up the thematic 16..., dxc4; 17 0-0, cxb3; when 18 Rc1, Nxa2; favors Black significantly, or 18 axb3, Bxf1; 19 Bxf1, Qc5; 20 Qh3+, Kb8; when Black recovers the pawn, has the Exchange in his pocket and his pieces are very active. Choosing the text is an error when the other two possibilities are available. Of course this is easy to say with a trusty computer at hand to resolve the complexities of calculation.

17. O-O? …..
After the game Taylor said he thought White was in some trouble after the superior 17 Qh3+. I am not so sure. After 17 Qh3+, Kb8; 18 0-0, Bxe5; 19 Bxe5, Qxe5; 20 Qh5, Rdf8; 21 Nd2, the tactics around f2 do not seem to give Black more than equality; 21..., Rxf2, 22 Rxf2; Qxa1+; 23 Rf1, and White has enough time to meet the several tricks Black has at his disposal. Now White is in trouble.

17.… Rh8
18. Qg6 Rh6
Taylor was down to 12 minutes on the clock when completing this move. Sells had about one hour left on his clock. A big difference, but I’ve seen Steve handle worse successfully.
19. Qxh6 …..
Phil spent a big chunk of his time before playing this forced move. He was no doubt looking for whatever chances might be had going forward.

19.… Bxh6
20. Ba3 ….
The computer suggests 20 Bh3+, and 21 e6, but Black is still winning.

20.… c5
At first it was puzzling why Steve did not just take some more material with 20..., Qxe5. Then a closer more careful examination turned up 20..., Qxe5; 21 Nd2, Nc2; 22 Bc5, and things are getting complicated.. Black is still clearly winning. The clock continues to run down for Taylor, and that could well have motivated him to chose something simple such as defending the threatened piece. Unfortunately, the move gives up a good deal of his advantage.

21. Bxb4 cxb4
22. Bxd5 ….
White has four pawns and a Rook for his Queen. This quite an unusual distribution of material. It is highly unlikely the with the number of pawns and pieces remaining the Queen will not be able to make enough multiple attacks to get some material back.

22.… Rf5!?
The straight forward 22..., Qxe5; 23 Nd2, Re7; takes back one pawn keeps a grip on the e-file. The text loses some piece of Black’s advantage.
23. Nd2 g4?!
Another faction of Black’s advantage slips away. Better 23..., Rxe5; and Black is clearly winning.

24. f4 ….
In the looming time trouble Steve may not have taken this move into account. After this push Black retains some edge but is no longer obviously winning.

24... gxf3 (e.p.)
Taylor used two minutes to decide on this move. There are no viable alternatives. Two minutes is not enough time to search for them. This was the first moment when I thought that the game was getting out of Steve’s control. He got to this position and whatever ideas he had for making a game for himself without the capture on f3 must not have worked.

25. Nxf3 Bb7
Taylor’s clock went below five minutes on this move. Worth a try is 25..., c6, for 26 Bxc6, Qc5+; collects the Bishop. Sells had 42 minutes remaining.

26. g4!? ….
A suspect move. Why not 26 Bxb7+, KxB7; 27 Rae1, avoiding tactics down the long diagonal, threatening to advance the e-pawn and planning to put his King on g2 so to begin advancing the K-side pawns? Since losing the Queen, Phil has kept his composure making a tough and dour defense. This was the first sign he was feeling the pressure.

26.…. Rf4
27. Bxb7+ Kxb7
28. h3 Re4
The Black Rook is very actively posted. That is the price of a slight inaccuracy on move 26.
Time: Sells; 31:00, Taylor; 3:45.

29. Rae1 Be3+
30. Kg2 Qc5
A worthy try is 30..., Qe6; and if White is greedy with 31 Ng5?, Rxg4+!; is very close to dead won for Black. Phil with plenty of time would have probably seen the tactic. If he had continued 31 Re2, then 31..., Qg6; 32 Rfe1?, Rxg4; wins by stripping away the pawn cover around the White King. The best for White is 32 Ree1. A tough decision to swallow in such a tense struggle, and one Phil might have missed.

31. Re2 a5
32. Rfe1 a4
Sells’ played had slowed over the last several moves. Time: Sells; 24:00, Taylor; 2:31.

33. g5 Qb6
34. h4 Kc8
35. h5 Rg4+
36. Kh3 Re4
The game is even. The however is not. Sells; 17:00, Taylor; 1:33

37. Kg2 Rg4+
38. Kh3 Qe6
39. Rxe3 Rxg5+
40. Kh2 Rxh5+
Taylor blitzed out this last sequence and his clock was now showing 1:15. Pretty quick play, but the best is yet to come. It is a nice creative tricky idea; giving up the Bishop for the pawns covering White’s King. Perpetual checks and even mates are in the air now.

41. Kg2 Qg4+
42. Kf2 Rh3
43. Rg1 Qf4
44. Rg2 Rh7

Now we get to the amazing part of the game. Time: Sells; 12:00, Taylor; 00:24!! The Rook was better going to h8, but that is a quibble.
45. e6 Re7
46. Rg8+ Kb7
47. Rg5 Qf6
48. Rge5 c6

There have been no obvious blunders in this one sided time scramble. Sells now had 9:00 minutes remaining, Taylor 00:21.

49. bxa4 Qf4
50. R5e4 Qf5
51. Ke2 Qa5
52. a3 Qxa4
53. axb4 Qc2+
Taylor misses a chance. He could have taken on b4. There was however no time to think about any tactics White could spring, say 53..., Qxb4; 54 Ne5. Rxe6?; 55 Nd3.

54. Kf1 Qb1+
55. Ke2 Qxb4
Sells was now drifting into his own time problems. Taylor made good use of Phil’s thinking time to check that capturing on b4 was safe.

56. Nd2 Qc5
57. Re5 Qd4
58. R5e4 …..
Phil was now well under five minutes on his clock. For the last few moves he had an advantage, not quite a winning one. Here he should have played 58 R3e4, then 58..., Qc3; 59 Rf5, threatens Rf5-f7. By no means a winning idea. It can be met by retiring the Queen to g7, or starting the King to the rescue with Kb7-c7, but that would have set a challenge to Taylor. Steve still had the 21 seconds on his clock. The unanswered question is did he see enough to snap out a move, or would some of the precious seconds be needed to decide which reply is best?

58.… Qa1
Returning to c5 would have been a silent offer of a repetition. The game move made Phil think some more worsening his own clock problem.

59. Nb3 Qh1
60. Nc5+ Kb6
61. Nd7+ Kc7
62. c5 Qh5+
63. Kd2 Qd5+
64. Kc3 Kd8?
This loses. Steve had kept up his speedy play but no one can hold back the flow of time. His clock was showing just a few seconds now. After this wonderful display of almost instant play, Taylor misses the drawing shot. With 64..., Rxe7, he could have secured the half point. If 65 e7, Qd2+; and Black is winning. If 65 exd7, Qxc5+; 66 Kd3, Kxe7; with a textbook draw. Tragic.

65. Rd4 Qa2
66. Nb6+ Kc7
67. Nd7 Lost on Time
The game ran on for six or seven more moves at pace too fast for me to record. Phil’s clock had one second on it as did Steve’s. Taylor’s flag dropped before Sells’ did. What an entertaining battle with many ideas and very good execution even when the clocks were down to near no time at all. An excellent performance by both parties.

More from Saratoga later this week.