The fifth in my series of memorable games contains several elements of the previous installments. Like game 4, it contains an absolutely miraculous comeback, from a position that seems totally lost to a brilliant victory. Like game 1, it ends with a queen sacrifice. And like game 2, it ends with a long sequence of moves whose point I did not actually grasp until near the very end.
What is it about this game and game 2 that attracts me so much? In some ways, they are very bad for teaching purposes because in general you should try to plan ahead. These games are the exact opposite. They are games where I was improvising on the fly, charging ahead like Luke Skywalker in his X-wing Starfighter, trusting in the Force to bring him through to safety.
But what made them so memorable was exactly that — the sense that the pieces themselves were guiding me. It’s a sense that is expressed in a poem by Ralph Waldo Emerson, who wrote: “He builded better than he knew / The conscious stone to beauty grew.” In other words, the architect built something better than he could have planned, because he let the stones do the talking. It is such a rare thing to have this feeling over the chess board, and when it happens it is worth remembering. When I finally saw for myself what the capstone was going to be and where it was going to fit … it was just goosebumps all over.
For ChessLecture subscribers, this game will be familiar; I gave a lecture on it called “A Yankees-Red Sox Moment.”
Colin Chow — Dana Mackenzie
Western States Open, 2012
My opponent in this game was 12 years old, and one of the many rising stars in Bay Area chess. I said in my lecture that his rating was “2175, going on 2275,” and it turned out I was right. A few months after this game he tied for second in the World Open Under-2400 section and his rating went up over 2300!
1. e4 e5 2. Nf3 Nc6 3. Bb5 Nd4 4. Nxd4 ed 5. O-O g6 6. d3 Bg7 7. Qe2 …
Position after 7. Qe2. Black to move.
FEN: r1bqk1nr/pppp1pbp/6p1/1B6/3pP3/3P4/PPP1QPPP/RNB2RK1 b kq – 0 7
Only seven moves in, and we’ve already hit a critical point in the game! I had never seen this move, 7. Qe2, before, and in fact it has never been played in the entire ChessBase database! The move I played was the standard move that I would play against 7. f4 or 7. Nd2 or 7. c3, but I failed to appreciate how this position was different.
After this game I analyzed it quickly to see how the opening had gone so wrong, and I realized that 7. … Qe7! was the proper answer here, with the idea of preparing … c6 and … d5. The very next day, I was paired against one of Chow’s friends. He whipped out the same first seven moves, perhaps expecting me to repeat my mistake. Not a chance! I played 7. … Qe7 and won that game, too. That was pretty cool, to face the same opening novelty twice in the same tournament and win both games!
7. … Ne7? 8. f4! …
Usually I’m skeptical of this move in the Bird Variation, but here it is completely justified because the pawn can push on to f5.
8. … c6 9. Ba4 d5 10. f5 …
Threatening f6. I don’t want to play … f6, hemming in my bishop and weakening my kingside. I considered 10. … Qd6 for a long time, when after 11. Bf4 Be5 Black is okay. However, Rybka says that White gets a significant advantage after 11. Bg5!, again threatening f6. Somewhat surprisingly, the computer thinks that the plan I played was the best defense — trading on f5 and blockading the f-pawn with my bishop. Even so, that plan has major drawbacks, weakening my kingside and opening the e-file. If that’s the “best defense,” then my position is already pretty desperate.
10. … gf!? 11. ef Bf6
Of course, I can’t take on f5 because of the pin on the e-file. That is what makes the 7. Qe2 line different from all the other variations.
12. Bh6! …
Position after 12. Bh6. Black to move.
FEN: r1bqk2r/pp2np1p/2p2b1B/3p1P2/B2p4/3P4/PPP1Q1PP/RN3RK1 b kq – 0 12
Even though 12. Bh6 is a somewhat obvious move, I still think it’s a very good one. At first I thought the point was to keep me from castling, which didn’t scare me because I don’t want to castle kingside anyway. However, the real point is that the weakest spot in Black’s position, the f7 pawn, now cannot be defended by any piece other than Black’s king. Eventually the king must leave the center, and the f-pawn will be a goner.
12. … b5?
Black does not have time for such time-wasting pawn demonstrations. It is essential to complete my development as quickly as possible with 12. … Qd6 13. Nd2 Bd7. If White tries to keep my king in the center with 14. Qh5, then I must abandon my f-pawn to the fates: 14. … O-O-O! 15. Qxf7 Nxf5. Aha! It turns out that Black isn’t the only player with a weak f-pawn.
Of course I knew that 12. … b5 was strategically dubious, but I think I was still in a state of denial about the need to sacrifice the f-pawn.
13. Bb3 a5 14. a4 b4 15. Nd2 Kd7
At this point I finally came to the realization that there was no time left to lose. But my king’s desperate flight is too little, too late.
16. Qh5 Qg8
Position after 16. … Qg8. White to move.
FEN: r1b3qr/3knp1p/2p2b1B/p2p1P1Q/Pp1p4/1B1P4/1PPN2PP/R4RK1 w – – 0 17
Here I was expecting White to play 17. Nf3? Kc7 18. Ng5 Bd7 and Black miraculously saves the f-pawn (19. Qxf7? Bxg5! or 19. Nxf7? Be8). But my opponent now comes up with what should have been, under any normal circumstances, the game-winning move.
17. Ne4!! …
At this point you must be wondering how Black could possibly have won this game. It doesn’t even look as if I’ll survive until move 25! Patience, patience. All will be revealed…
17. … de
There’s no point in declining the sacrifice. If 17. … Be5 18. Ng5 White is a tempo ahead of the line I just showed you, and he comfortably wins the f-pawn with a huge advantage.
18. Bxf7 Qd8 19. Bf4! …
My young opponent plays another exquisite move. The threat is, of course, 20. Be6 mate. Throughout this part of the game I was just playing move to move, trying to do anything to stay alive.
19. … c5 20. de Ba6 21. Rfe1 h6
I saw e5 coming, and this was the only way I could see to avoid losing a bishop. I was expecting 22. e5 anyway, winning a pawn and giving White an incredible four connected passed pawns on the kingside. Surely this is winning. Nevertheless, my opponent once again finds an even better move. There’s no question that at this point he saw all the way to the position on move 27 and thought he was winning a full rook.
There are so many ironies in chess. By being so damn brilliant and passing up the routine, workmanlike win after 22. e5, my opponent eventually ended up losing.
Position after 21. … h6. White to move.
FEN: r2q3r/3knB2/b4b1p/p1p2P1Q/Pp1pPB2/8/1PP3PP/R3R1K1 w – – 0 22
22. Be6+! Kc6 23. Qf7! Bg5
Just as bad is 23. … Rf8 24. Bd5+ Kb6 25. Qe6+ Ka7 26. Bxh6. I’m still just playing one forced move after another and praying.
24. Bd5+ Kb6 25. Qe6+? …
A little move-order slip. I didn’t even realize it until a couple days ago when I went over the game with a computer. White should play 25. Bxg5 first, and after 25. … hg 26. Bxa8, winning a full rook. (26. … Qxa8 27. Qe7 or 26. … Qc7 27. Qf6+!) Rybka rates this line as 4 pawns behind for Black, while after the move White played I am “only” 2 pawns behind. I’m still lost, but it’s the difference between being put-away-the-pieces-and-go-home lost, versus let’s-just-play-a-few-more-moves-and-see-what-happens lost.
25. … Ka7 26. Bxg5 hg 27. Bxa8 …
Position after 27. Bxa8. Black to move.
FEN: B2q3r/k3n3/b3Q3/p1p2Pp1/Pp1pP3/8/1PP3PP/R3R1K1 b – – 0 27
Here it is, the turning point of the game. White surely got to this position when analyzing his 22nd move and saw that Black cannot recapture the bishop. If 27. … Kxa8 28. Qxa6+ and if 27. … Qxa8 28. Qxe7+. It would seem that resignation is in order for Black. Except…
Except that Black has one little move that White missed.
27. … Qc7!
Threatening perpetual check with … Qxh2+ and … Qf4+. In order to prevent the perpetual, white has to give up the bishop on a8 after all.
In my ChessLecture I called this a “Yankees-Red Sox moment.” If you’re a baseball fan, you know the moment I’m talking about. The 2004 American League championship between the New York Yankees and Boston Red Sox. The Yankees demolished the Red Sox with three straight wins. No baseball team had ever come back from a 3-0 deficit in a seven-game series. The Yankees were ahead by one run going into the ninth inning of game 4. They had the best relief pitcher in the history of baseball, Mariano Rivera, on the mound. Game over, series over, right? Then Rivera gives up a meaningless walk, and Dave Roberts steals second base.
That was the moment. I’ll never forget the look that Yankees catcher Jorge Posada had on his face. It was his throw to second base that had come oh-so-close to catching Dave Roberts. His look was 50 percent disgust and 50 percent “wait a minute, that wasn’t supposed to happen.” Of course, it shouldn’t have mattered! The Yankees were ahead with three outs to go! Even if they gave up a run, they could still win in extra innings! Even if they lost, they would still have three more chances to beat the Red Sox!
But as we now know, Roberts did score the tying run. The Red Sox won the game in extra innings, on Big Papi (David Ortiz)’s home run. They won games 5, 6, and 7, too. Then they beat the St. Louis Cardinals in the World Series for good measure. And none of it would have happened if Dave Roberts hadn’t stolen second base. What that play meant was that the Red Sox hadn’t given up yet.
You can say exactly the same things about 27. … Qc7. White’s probably a bit disgusted right now because it turns out Black wins the bishop on a8 after all. Still, objectively, it shouldn’t matter. White is still completely winning. He will have plenty more chances to beat the Red… oops, I mean, Black pieces. But this move is Black tapping White on the shoulder and saying, “Remember me? I’m still here. And I’m not giving up.”
28. e5 Rxa8 29. f6 Nc6 30. Qd5 Kb6
Surprisingly, the king has turned into a useful piece. It defends c6 and c5 so that Black’s queen can go elsewhere. Also, by some perverse miracle, the formerly hunted king is now quite safe behind a barricade of pawns.
31. e6 …
Hard to know what to say about this move. On the one hand, it’s the obvious move, and it should have won. The trouble, though, is that White voluntarily opens up the b8-h2 diagonal and breathes life into Black’s position. He gives Black the initiative, and this is a matter of great importance. As I said in my ChessLecture, as long as you have the initiative you cannot lose the game. From this move until White’s resignation, twelve moves later, Black is the one calling all the shots. What a change from the first 27 moves of the game!
31. … Rh8
Position after 31. … Rh8. White to move.
FEN: 7r/2q5/bkn1PP2/p1pQ2p1/Pp1p4/8/1PP3PP/R3R1K1 w – – 0 32
Once again Black threatens perpetual check. White has two obvious moves to prevent it: 32. h3 or 32. g3. Which one is the best?
A little bit of space so that you can think about it …
…
The better defense was 32. h3! for two reasons, tactical and strategical. Let’s start with the tactics. In some ways 32. h3 appears worse, because Black can strike at the h-pawn right away with 32. … g4. I’m sure this is why Chow didn’t play it. However, if White has nerves of steel he should realize that he does not need to defend on the kingside. His passed e-pawn decides: 33. e7! gh 34. e8Q! As perilous as the position seems, Black has no way to checkmate White on the kingside — in fact, he only has one check, 34. … h2+ 35. Kh1, and then the attack is over. Black’s pawn is actually a “traitor pawn” that gets in the way of his queen and rook.
Thus Black cannot play 32. … g4, his only attacking move, and his initiative dies after one move. Chess is a battle for the initiative. With sufficient courage, White could have won this battle and won the game.
The second reason for preferring 32. h3 is strategical. By playing 32. g3, White would create a huge weakness on f3, and also open up the long diagonal. Those two factors give Black very serious counter-chances, as we are about to see.
32. g3? Qh7 33. Qg2 …
An important moment psychologically, if nothing else. For the first time in the game White has been forced to retreat.
33. … Bb7
Wresting control of the long diagonal. 34. e7 would now be met by 34. … Ne5!
34. Qf2 d3!?
Another tricky move to comment on. The computer likes 34. … Qc7 with the idea of 35. e7 Ne5 36. Qe2 Qc6! 37. Qb5+ Qxb5 38. ab, with a very double-edged position.
Psychologically, I was not ready yet to take my queen off the h-file, and my overriding objective was to get my knight to f3 somehow, some way. 34. … d3 seemed like the way to do it.
Also, as I commented in my lecture, from here to the end of the game I was basically able to predict every move of White. I was practically certain that White was going to play 35. c3? Chess becomes much easier when you can predict your opponent’s moves! I call it “moving your opponent’s pieces.”
35. c3? …
Logical… and wrong! White, of course, wants to prevent … Nd4. As it turns out, his last chance to win the game was to allow that move and play 35. cd! Nd4 36. Re3! Nf3+ 37. Rxf3 Bxf3. It looks as if Black has gotten the last laugh, because White cannot recapture on f3. But it’s actually White who laughs last, because after 38. f7! my bishop cannot escape! Any bishop move would be met by 39. f8Q. Black’s best move according to the computer is 38. … Qf5, and in this insanely complicated position White wins (supposedly, according to the computer) with 39. Rc1 Rc8 40. d4!
Whatever. All that this line really proves is that we’ve reached a position only computers can play correctly. For practical human chess, 35. c3 is forced. After this the computer says that Black has already equalized.
35. … Qc7
Again, sticking to the plan of trying to get the knight to f3. Now … Ne5 is threatened.
36. Qf5 …
Again, totally predictable. White continues to bar the door to Black’s wandering steed.
Even though there is in general little that one can learn from “computer chess,” I can’t resist showing you one equalizing variation that Rybka comes up with, where White again ignores the threat and plays 36. f7!? Now, of course, Black plays 36. … Ne5 37. Rxe5 Qxe5, and it’s White to play and not lose.
Position after 37. … Qxe5 (analysis). White to move.
FEN: 7r/2q5/bkn1PP2/p1pQ2p1/Pp1p4/8/1PP3PP/R3R1K1 w – – 0 32
The first thing to notice is that if 38. f8Q?? Qd5! wins. If 39. Qxh8 Qh1 mate. In fact, White’s only way to stop mate is to sacrifice his newly minted queen on c5, but after 40. Q8xc5+ Qxc5 Black is winning. However, 38. f8N!! miraculously saves the game. The main point is 38. … Qd5 39. Nd7+ and Black has no good place to put his king. If he goes to a7, a6, or c7, then White is able to trade off one of Black’s attackers with either Nxc5+ or Qxc5+. And if Black’s king goes to c6, then he blocks his own bishop-queen battery! White can then play 39. … Kc6 40. cb and start battering down the pawn wall. For these reasons, Black’s best reply is simply 38. … Rxf8 39. Qxf8, but then chances are roughly even after 39. … Qd5 40. Qf1! (only move) Qh1+.
As I said, this line has little relevance to the game, but it’s one of the most surprising under-promotions I’ve ever seen in a position that could actually have arisen out of a tournament game (rather than a composed position).
Okay, back to the game. After White’s 36. Qf5 I continued my attempts to get my knight to f3:
36. … d2! 37. Re2 Qd6
Now with a new threat — to queen my pawn. Again, White’s response is totally predictable.
38. Rd1 …
Position after 38. Rd1. Black to move.
FEN: 7r/1b6/1knqPP2/p1p2Qp1/Pp6/2P3P1/1P1pR2P/3R2K1 b – – 0 38
But now what is Black supposed to do? It looks as if White has plugged all the holes in all the dikes. The knight still can’t go to d4 or e5, and the d-pawn looks like a goner.
38. … Nd4!!
Yes, we can! I consider this move a fitting bookend to White’s brilliant knight sacrifice earlier in the game.
39. cd …
Or 39. Rexd2 Ne2+ 40.Kf1 Nxg3+ with at least a draw.
39. … Qxd4+
And here I get lucky again. Only it’s sort of luck and sort of not. If White plays 40. Qf2! then I have nothing better than a draw by repetition with 40. … Qd5! 41. Qg2! (the only way to stop mate on h1) Qd4+! (I can win the exchange with 41. … Qxg2+, but the endgame is dead lost) 42. Qf2 (White has to save his queen) Qd5 etc. I don’t know whether Chow saw this (I believe he was now down to a minute on his clock), but in any case I think he still firmly believed that he was winning. Why take a draw if you are winning? So he played the losing move:
40. Rf2? …
And now, even though I had only a couple minutes left on my own clock, I saw how the game was going to end.
40. … Qe3!
This move is in some ways even better than the next one, because it’s the quiet move that sets up the fireworks. One threat, of course, is 41. Rxd2 Qe1+. The other threat is less obvious, and my opponent failed to see it.
41. f7? …
Notice that the time control was on move 40, so White now had all the time he wanted to decide on his move. Even so, he still did not see my winning combination. The best try for White here was 41. Kf1. The move I would probably have played is 41. … Ba6+, but the computer says that the best move (and the only clearly winning move) is the rook sac 41. … Rxh2! I have to say, I doubt I would have found that.
Position after 41. f7. Black to move.
FEN: 7r/1b3P2/1k2P3/p1p2Qp1/Pp6/4q1P1/1P1p1R1P/3R2K1 b – – 0 41
Do you see what White missed?
41. … Qxg3+!!
I think that the late Jerry Hanken was really onto something when he wrote his famous “Parting with the Lady” article for Chess Life. There’s something about queen sacrifices that makes them psychologically harder to spot (and also, psychologically, harder to respond calmly to) than the sacrifice of any other piece. In this case, it’s just a very simple two-move combination: if 42. hg Rh1 mate. Even so, my opponent never saw it coming.
42. Kf1 Ba6+ 43. Re2 Bxe2+ White resigns
It’s mate in two after 44. Kxe2 Rxh2+ 45. Kf1 Qg2 mate.
What a stupendous finish to a stupendous game!
It’s true that I got lucky. White could so easily have played the checks in the correct order on move 25, or pushed the correct pawn on move 32, or even acquiesced to a draw on move 40, and then the result would have been different. Even so, I feel that this game is one of my finest creations. First, it’s an incredible “story game.” There are so many wonderful moments: The flight of Black’s king across the board. Chow’s beautiful knight sacrifice and his wonderful combination that should have won a rook. The “Yankees-Red Sox moment.” The suspense-filled battle on the edge of a cliff from moves 33-38, as Black tries to get his knight to e5 or d4 and White tries to deny it. And finally the swift, stunning ending, the knight sacrifice and the “quiet move” on move 40 and the queen sacrifice out of nowhere. Blood was shed in this game. Tears were wept. This was chess at the highest level of art.
My other comment is that yes, you will always need luck to win a game that you should have lost. But there are certain things you can do to help that luck along. Keep fighting for the initiative, keep making annoying little threats that keep your opponent off balance, and most importantly, never give up.
{ 4 comments… read them below or add one }
Very interesting game indeed!
This game is so dope!
Thanks! With many games you can say, “Well, I’ve seen a game like that.” But with this one, I think you’d have to search long and hard to find another game quite like it.
Just found your website and I’m reading through your interesting blogs. We come from the same “Fischer Boom” chess generation.
This is a crazy game!! I’m familiar with the surge of adrenaline when you realize you’re not going to lose in the next 5 moves, and maybe you have some counter-play. But I can’t even imagine how you felt when this ended.
I played Colin Chow in the first round of the National Open several years back and managed a draw though he was rated several hundred points higher.
Looking forward to reading the rest of your articles!