Sunday, June 24, 2012

The Interzonal, part 1



The 1991 Candidates Cycle represented a changing of the guard. Of the great players from the 1960s and 70s, only Viktor Korchnoi qualified, and the only former champion to secure a place was FIDE favorite Anatoly Karpov (Tal and Smyslov played in the Interzonal, but did not qualify). Replacing reliable standbys like Larsen, Portisch, Polugaevsky, etc. were a whole series of emerging stars from the new generation, names that were soon to achieve international recognition: Boris Gelfand, Viswanthan Anand, Nigel Short, and of course, Vassily Ivanchuk. It still seemed implausible that anyone could unseat the eternal challenger, ex-champion Karpov (who had contested 7! straight world championship matches and qualified for another), but time was taking its inexorable toll on the great Tolya, and prospects seemed bright for the new contenders, if not in this cycle, then certainly in the next one.



The tournament began badly for Ivanchuk, as he lost to New Zealand Master Murray Chandler in the first round in an upset of over 100 rating points. Chandler managed to diffuse Ivanchuk's Winawer French, forcing the Ukrainian star into a passive position that appeared lost before resourceful play brought Ivanchuk within reach of a draw and then he lost the thread, and with it, the game. Not losing heart though, Ivanchuk clawed right back into contention, rattling off an impressive string of five straight victories!

Ivanchuk-Cabrilio
Manila Interzonal 1990
Round 2



Here, Ivanchuk won with the help of a nice little combination:

23. Rxe4! Nxe4
24. Qe7

And black was forced to resign in 6 moves.

Zapata-Ivanchuk
Manila Interzonal 1990
Round 3



In this difficult position, white blundered with:

29. Be6?

Allowing Ivanchuk to obliterate his king with the nice sequence:

29. ... Nh5
30. Nb4 Bxg3+
31. fxg3 Rf2+
32. Kh1 Nxg3+
33. Rxg3



And now, rather than 33. ... Rxg3, allowing 34. Nxc6+ with a hint of counterplay, Ivanchuk calmly finished the game with the elegant stroke:

33. ... Rh2+

Ivanchuk-De Firmian
Manila Interzonal 1990
Round 4



In this fascinating position, Grand Master De Firmian has sacrificed his knight on e4 to shatter white's kingside and build up a terrific initiative (This plan has since appeared in many games). Faced with a complex opening innovation, Ivanchuk nevertheless arrived at an amazing rejoinder, here playing:

16. c6!

The point being that 16. ... Bxc6 will be met by the discoordinating 17. Bf4.

16. ... 0-0
17. Bd2 Bxc6
18. Qa5 Rb6
19. e3 Rfb8
20. Rc1 Nxe4
21. Nf3



And here, facing the ever-present threat that his initiative will run out and his two pawns will not be enough to compensate the lost piece, De Firmian missed the best move, 21. ... Nxd2, instead playing the immediate:

21. ... Qe7

To which Ivanchuk duly responded with:

22. Bb4

And after

22. ... Qb7
23. Bd3

Ivanchuk had fully unraveled. Black's momentum waned over the next few moves, and Ivanchuk promptly seized the initiative and used his extra piece to win.

In the next round, playing black against Pedrag Nikolic, Ivanchuk turned in a strangely reckless performance, as if his string of success had gone to his head. He did not lose his form though, and when Nikolic missed the best continuation to punish him, Ivanchuk recovered to pull out yet another win.

Nikolic-Ivanchuk
Manila Interzonal 1990
Round 5

1. d4 f5
2. g3 Nf6
3. Bg2 g6
4. c3 c6
5. Bg5 Bg7
6. Nd2 d5



The position is a Dutch defense, ostensibly a Leningrad, which typically resembles a King's Indian Defense, except that Ivanchuk has been forced to play d5 by Nikolic's slightly unusual setup with c3 (otherwise Qb3 would make it impossible for black to castle). This has brought about an odd sort of hybrid between the Leningrad and Stonewall Dutch variations. The position is a bit shaky for black, with holes and several points of entry for the white pieces (if Nikolic can coordinate an invasion). On the other hand, it is difficult for white to fully mobilize here, and black certainly has his chances...

7. Nh3

The king's knight aims for f4, reserving the f3-square for the queen's knight.

7. ... 0-0
8. 0-0 Qe8
9. c4 Ne4



Ivanchuk plays resourcefully, aiming to preemptively diffuse white's mobilization through an exchange.

10. Nxe4 dxe4
11. f3 exf3

Ivanchuk later recommended the fascinating and complex 11. ... e5!?

12. exf3 Qf7
13. Re1 Qxc4



Snatching the c4-pawn was a very risky move, requiring great precision to hold the balance. Essentially, Ivanchuk has just opened the position for white and that with all of his queenside pieces still in the box! 13. ... Re8 was more conservative, probably still allowing white some advantage, but with far less severe consequences than in the game.

14. Bxe7 Re8
15. Kh1 Na6?

Amazingly, this simple developing move seems to be a serious mistake overlooked by several of the GM commentators. With his thirteenth move, Ivanchuk has bitten off a bit more than he can chew, and now he must seek to simplify the position. 15. ... Qxd4 seems to be best, angling to force the queens off. Then 16. Qb3+ Qd5 and it seems that white has no better than to exchange: 17. Qxd5+ cxd5 18. f4 Nc6 19. Bxd5+



White retains an annoying initiative here, with threats like Ng5, but black can defend via Bd4 and Kg7, and his position is fundamentally sound and ought to gradually drift toward equality. Instead, with the game continuation, Ivanchuk allows white's initiative to grow with the queens still on so that the attack packs a lot more venom.

16. Bf1 Qf7
17. Ng5 Qd5



And now, it seems that white is winning! After 18. Rc1! b5 (otherwise Bc4 skewers queen and king) 19. Rxc6! Bb7 (19. ... Qxc6 20. Qb3 +-) 20. Rd6 Qxa2 21. Bxb5 the black position crumbles. Nikolic didn't find this continuation though, instead opting for:

18. Qc1?

This threatens Bc4, but it remove the check threat along the a2-g8 diagonal, which takes the bite out of the tactical threats.

18. ... b5
19. Bg2 Bb7
20. f4 Qxd4



And suddenly, miraculously, black is up a pawn with chances of staying that way. Note that 21. Bxc6 loses to the diabolical 21. ... Qd7!

21. a4 Rab8!

This move holds the black position together, as we shall see in another move.

22. axb5 cxb5



And now, 23. Bxb7 looks like it should work but doesn't! After 23. ... Rxb7 24. Rxa6? comes 24. ... Qd5+ 25. Kg1 Bd4+ 26. Kf1 Qh1+ 27. Ke2 Rbxe7+ and the roof caves in on white.

So Nikolic instead tried:

23. Rxa6 Bxa6
24. Qc6?

This renews the threat along the a2-g8 diagonal, but it turns out that Ivanchuk has an escape:

24. ... h6!
25. Qe6+ Kh8
26. Nf7+ Kh7



And preposterously, the black king slinks away to blow razzberries at the white army.

27. Ne5 Bxe5
28. fxe5 Qc4
29. Qf6 Bb7
30. e6 Rg8



Nikolic maneuvers every which way to try to get at the black king, but Ivanchuk seems to have an answer for all comers.

31. h4

This is pure dreaming though.

31. ... Bxg2+
32. Kxg2 Qd5+
33. Kh3 Rbc8
34. h5 g5
35. Qe5



This last move is tantamount to raising the white flag, and indeed, the game is over:

35. ... g4
36. Kh4 Rc2

With mate to follow.

0-1

So, after five rounds, Ivanchuk found himself at +1.5 on a 4-game winning streak, but in the next round, he faced the legendary Hungarian master Lajos Portisch. I'll pick up the story in a future installment!

Saturday, June 23, 2012

As in Life, so in Chess

Over the course of the last few posts, we've seen vivid demonstrations of Ivanchuk's immense talent. Now I'm going to look at his personality a bit and begin to recap his playing career.



Vassily Ivanchuk exploded onto the chess scene in 1988 at 19 by scoring a scintillating 7.5/9 at the New York Open, besting a field of strong grandmasters, including Vladimir Tukmakov, Bozidar Ivanovic, and Boris Gulko! By the end of the year, his rating had skyrocketed into the world's top 10, and he soon achieved international fame by winning the 1991 Linares Super Tournament ahead of a score of the world's best players, beating World Champion Garry Kasparov and perennial challenger and ex-champion Anatoly Karpov in the process. By the onset of the World Championship candidates cycle that year, it was widely anticipated that Ivanchuk was a future world champion in the making.

Alas, Ivanchuk's mercurial temperament was soon to surface. It seems that he is something of a savant. While he has enormous intuitive chess talent, his emotional equilibrium is unpredictable, subject to mercurial twists and turns that affect his playing objectivity and judgement. As he describes it, playing under the weight of expectations,

"Sometimes... gives me an extra incentive to put in more work, but at other times... success goes to your head. Then you don't always objectively accept defeats, but instead think, 'How's that possible? How could I make such bad moves?" and I'm not always able to recover quickly."



Elaborating on the subject of nerves, he relates:

"I know myself--if a tournament is very important, then that's it, I can't prepare for it--neither at the computer nor at the chessboard. When the tension drops a little then the desire to play chess returns... Why is it like that? I don't know."

Moreover:

"It's not only about the games themselves, but also the time between games. For me the big problem is the latter--waiting, getting butterflies in your stomach..."

This problem of nerves has plagued Ivanchuk throughout his career, such that it is never surprising to see him at the top of a tournament cross table or the bottom, depending on how his emotional equilibrium is withstanding the stress.



At Reggio Emilia in April of this year (2012), Ivanchuk gave a lamentably vivid demonstration of the agony he endures when his nerves fail. As the top rated player in a 6-person double round robin, Ivanchuk seemed to start out with confidence, scoring two wins and two draws in the first four rounds. In the fifth round, playing with the black pieces against American GM Hikaru Nakamura, Ivanchuk put together a masterful positional game, goading his opponent forward and then counterattacking ferociously to achieve a winning position, and then... he blundered, first losing his advantage, and then letting the game crumble into a loss. In the 6th round, playing with black again, this time against the lowest rated player in the event, Ivanchuk played passively, almost indifferently, and slowly lost a game that he seemed never to be fully present in. In the 7th round, he played more actively, but at a critical point he suffered a complete mental lapse and dropped a piece:

Vitiugov-Ivanchuk



Black to move, Ivanchuk played 30. ... Nxc4?? 31. Nxc4 and only now realized that the planned 31. ... b5 and Qb6, to recapture on c4 with the rook does not work because of 32. Ne3, blocking the check.



Alas, the story did not end there.

In round 8, Ivanchuk showed up a mere shadow of himself and seemed to invite all hell to rain on his head:

Ivanchuk-Caruana



Such positions do not arise of themselves! From here, Ivanchuk's position just steadily eroded:

21. ... e4
22. fxe4 dxe4
23. Be2 f4
24. axb6 Qxb6
25. Qb3 Qa7

And now, trying to be active, Ivanchuk played the losing:

26. Bb5?



26. ... cxd4
27. Bxd7 dxe3
28. Nc4 Rd8
29. Ba4 Rd2

And now, obviously beside himself, Ivanchuk suffered a complete and almost unprecedented meltdown:



30. Qxb7 Qxb7
31. Rxf4 gxf4
32. Nxd2 exd2
33. Bxf4 Rxf4
34. Bc6 Qb6+

0-1

One could cry watching such a thing.

Amazingly, having vented at the universe, Ivanchuk came back to calmly draw in the 9th round and then won in the tenth! At was as if, having lost everything, he could finally set his nerves aside and return to playing chess.

In terms of pure personality, Ivanchuk seems as temperamental and eccentric on the board or off. As World Champion Anand described him:

"He's someone who is very intelligent... but you never know which mood he is going to be in. Some days he will treat you like his long-lost brother. The next day he ignores you completely. ... I have seen him totally drunk and singing Ukrainian poetry and then the next day I have seen him give an impressive talk."

Ivanchuk tends not to look at the board when playing, but instead, stares blankly around the room, as he reviews variations solely in his mind.



He plays chess prolifically, seemingly to avoid having many periods of "waiting" where his nerves can play on him, and he is one of the most active grandmasters in the world. He also plays regularly with amateurs; as he puts it:

"If someone asks me to play a game then it's not hard for me to give them that pleasure."

Non-chess hobbies that he pursues tend to be rather fanatical. For a time, he played endless games of Internet checkers:

"I would ... go onto the site and play checkers for hours on end, until you no longer get up from your desk, and have square eyes. ... it takes me a huge amount of effort to stop."

Unfortunately, he also finds chess somewhat antisocial, relating that:

"In general, I feel as though I don't have enough contact with people... In chess there's constant competition and it's important not to reveal your weaknesses, which has hurt me a little in life as I don't talk enough with people... there has to be a certain distance between competitors."

Elaborating on this, he opines:

"I'd say that for a top-level chess player, sincerity and openness are negative qualities."



Of course, Ivanchuk is also famous for his eccentric, seemingly autistic moments, such as when he was awarded an oversized, ceremonial winner's check for finishing first at the 1994 Intel Grand Prix, and, left holding it on the stage, he looked at it bemusedly for a moment and then proceeded to attempt to fold it down to pocket size and put it into his wallet. The tournament organizers had to intervene!

On another occasion, he was been observed out in a park near his hotel around midnight, dressed in shorts despite freezing temperatures and howling like a wolf, perhaps attempting to exorcise the agony of his loss the day before...

In this same spirit, I'd like to wrap this up with another priceless Ivanchuk quote. Whens asked whether chess playing came naturally to him, he replied:

"It's like a centipede. If it thought about how to place its feet it would be difficult for it to move, but instead it just moves, and everything works out well for it."

Ivanchuk's Counterattack

I'm going to cover one more Ivanchuk game before delving into a little of his career history and reviewing his 1991 candidates matches and his numerous appearances in the FIDE World Cup. Ultimately, I hope to do pieces on each of the participants in the upcoming (March 2013!) FIDE Candidates tournament, and Ivanchuk will be the first.

So far, we've seen Ivanchuk deliver some incredible blows in the midst of wild attacks. Now I'm going to examine a more recent game where the Ukrainian GM demonstrated an amazing instinct for counterattack--something quite a bit more challenging on the nerves!



Aronian-Ivanchuk
Bilbao Masters 2011
Round 4

1. Nf3 d5
2. c4 e6
3. d4 Nf6
4. Nc3 Nbd7
5. Bg5 Bb4
6. cxd5 exd5
7. Nd2!? c6



After just 7 moves, we have a strange, untheoretical position. The black bishop placement on b4 resembles a Ragozin Queen's Gambit Declined, but the Nbd7 is uncharacteristic of this variation. Meanwhile, Aronian's Nd2 move seems like a slightly odd attempt to profit from Ivanchuk's mixing of systems. The pawn exchange and subsequent c6-advance have brought about a Carlsbad exchange structure... The opening has clearly given rise to a position that will force the players to innovate. How refreshing!

8. e3 Nf8
9. Bd3 Be7
10. Qc2 Ne6
11. Bh4 g6



Ivanchuk preempted white's threats down the long diagonal preparatory to castling. The position here remains original and rather more dynamic than many traditional QGD variations.

12. h3 0-0
13. Nf3 a5

With his last move, Ivanchuk took space on the queenside and discouraged queenside castling for white. Here, white was obliged to settle on a plan. One option was to castle kingside and aim for central play with f3 and e4. Instead, Aronian decided to throw down the gauntlet and attack:

14. g4!?



14. ... b5
15. Ne5 Bb7
16. f4!? c5!?

Now the game steers toward insane complications as both players burned their bridges.



Ivanchuk welcomed the capture of his b5 pawn, perhaps planning to respond with 17. ... c4 or even cxd4, opening lines in the center with the idea of plunking a knight on e4. Aronian opted to eschew the pawn capture though, and launched a direct and violent assault on the black king.



17. f5! cxd4
18. fxe6 dxc3
19. 0-0



Castling here was more about connecting the rooks and bringing a rook to the f-file than about king safety, but white's king could hardly hope to find security on the other wing! Now, however, Ivanchuk immediately took aim by blasting open the center to clear the long diagonal for an all-out counterassault with the fascinating:

19. ... d4
20. Nxf7 Qd5
21. Bxf6

Aronian ignored black's looming check on h1, seemingly content that his king would be able to flee and his own attack would continue.

21. ... Qh1+
22. Kf2 Qxh3!



This move was killer. Rather than following up with checks, encouraging the white king to flee to the queenside, Ivanchuk simply threatened Qxe3#.

23. Ke1 cxb2

Clearing the dark-squared diagonal a5-e1. There will be no security for the white king on the queenside!

24. Qxb2 Bb4+
25. Kd1 Bf3+



The poor white king has come under a harrowing barrage! From this position, 26. Kc2 Rfc8+ 27. Kb1 Qxf1+!! 29. Bxf1 Be4+ wins for black, so Aronian had to shed an exchange.

26. Rxf3 Qxf3+
27. Be2 Qxf6

Ivanchuk finally recovers the piece he shed to initiate his offensive, and boy has it paid interest!

28. g5 Qg7
29. Qxd4



Here, Ivanchuk had to consider the possibility of 29. ... Qxd4 30. exd4 Bc3 with a complex ending that may well have held drawing chances for white. Certainly, it was preferable to keep the queens on the board, but Ivanchuk was down to 1 minute at this point to Aronian's 10 and there were still 10 moves left until the time control. Nevertheless, Ivanchuk decided to play on for the attack

29. ... Rxf7
30. exf7+ Qxf7
31. Rc1 Rb8
32. a4??



Amazingly, Aronian cracks in Ivanchuk's time pressure. After the more circumspect 33. Rc2, black's advantage would have been modest. With a4 though, white gives up the massive check on b3 and his position instantly crumbles.

32. ... Qb3+
33. Rc2 Qb1+
34. Rc1 Qb3+
35. Rc2

Ivanchuk repeated moves to creep closer to the control, now down to a mere seconds on the clock (not counting the increment).

35. ... Rc8
36. Bc4+



The only way to save the rook is to drop the bishop.

36. ... bxc4
37. Qd5+ Kf8

For a minute or so, Aronian hunted a perpetual.

38. Qd7 Re8

But there is nothing here. White is completely busted.

0-1

One more time we have seen Ivanchuk whip up an attack with great energy and wild creativity. In this case, he matched tactical acumen with great intuition, sensing the resources of his own defense while foreseeing the flimsiness of white's position.

Friday, June 22, 2012

The Maddening Knight and Bishop



Every once in a while in my chess preparation I set aside a little time to review the knight and bishop mate. It seems to me that this mating pattern is one of those bugbears that every chess player (shy of strong masters perhaps?) simply prays and gambles will never become relevant in a game. I felt that way (more or less) myself for a long time--why bother studying it? It'll never come up. And then I had that playing experience that calls conventional wisdom into question. So, here's the first example from my own praxis evaluated here on this blog.

For this game, I really want to look at the ending and the middlegame, but I'll include the opening with minimal commentary for those readers who like to see where a game has come from:

Joshua-(Expert)
Summer Swiss, Natick, MA 2006
Round 1

1. e4 c5
2. Nf3 Nc6
3. d4 cxd4
4. Nxd4 Nf6
5. Nc3 e5
6. Ndb5 d6
7. Nd5 Nxd5
8. exd5 Nb8
9. Qf3!? Be7?!
10. Qc3 Na6
11. Be3 Bd7
12. Nxa7 f5



Black has responded poorly to white's opening experiment, but having lost a pawn, he now develops a little bit of counterplay. I suddenly got nervous, began seeing all kinds of nonexistent evils in the position, and blundered.

13. f3?

Better is 13. Bxa6 bxa6 14. 0-0 f4 15. Nc6 where white will consolidate a small advantage.

13. ... Bh4+
14. g3 f4
15. Nb5?



It's hard to understand what I was afraid of. 15. Bf2 fxg3 looks scarier than it is, eg. 16. hxg3 Bxg3 17. Bxg3 Rxa7. Black has recovered his pawn and gained a passed h-pawn, but it has come at the cost of piece coordination, and white will obtain significant play against his king. White is better. Instead:

15. ... 0-0
16. Bf2 Rc8



Now white has to walk a bit of a tightrope.

17. Qb3

The computer prefers 17. Qa3, aiming for counterplay with a proposed perpetual after 17. ... Rxc2 18. gxh4 Rxf2 19. Qxh4 Kg1 etc. My opponent would no doubt have played 17. ... Be7, allowing 18. 0-0-0, and at least I'm still alive kicking, although black definitely has all the fun here.

17. ... Qa5+?!

My opponent angles for activity, but this move allowed 18. c3!, a move that my meager tactical vision consigned to the scrap heap, as I was worried about the future of my b5 knight. It turns out that the queen placement on b3 opens up tactical possibilities based on Na7-c6, due to the threat of pawn takes knight, pawn takes pawn check! This is the kind of lingering tactical idea that computers are chill with and human nervous systems can hardly tolerate... Incidentally, 17. ... Be7 was more or less equal after something like 18. 0-0-0 Nc5. Black will have the fun, but white should hold.

18. Nc3? fxg3
19. hxg3 Rxf3

Now white is in trouble, and we can see the shortcomings of my piece placement.

20. Bb6



Rather than attempting to counter strike with something like 20. Bd3, threatening Bxh7, I realized that I could force the queens off at the cost of my g-pawn (which was going down anyway) and I figured I would have better chances of holding with his queen off the board. This turns out to be rather misguided...

20. ... Bxg3+
21. Kd1 Qb4
22. Qxb4 Nxb4

This "endgame" is hardly less shaky than the middlegame. c3 is still a squealing weakness, and the black kingside passers a long-term problem.

23. a3



This is the resource I was counting on. White needs a tempo or so to try to catch his breath and unravel. If black complies, the plan actually works: 23. ... Na6 24. Ne4 Bg4 (black must counterattack; d6 cannot be defended) 25. Be2 Rcf8!



Threatening the demonic Rf1+! Therefore, white must play: 26. Kc1 Bf4+ 27. Kb1 Be3 28. Rh4 Bxb3 29. Rxg4 Rh3 30. Nxd6



And here, white is still a bit shakier than black, but the passed d-pawn suddenly offers realistic counter chances and prospects of a dynamically even game. I hadn't seen all of this with clarity of course, but I had the sense that the queen exchange had given me a new lease on life...

Instead, my opponent sank into a deep thought, fidgeted several times like a cat about to make a jump, and then slammed

23. ... Rfxc3!

onto the board with a huge exultant grin.



I admit, I was initially surprised and pleased to be winning an exchange. Then I was steadily demoralized as I took stock of what my opponent had just done to my position. At the cost of a rook for a knight and pawn, black has taken a commanding hold over the board, demonstrating that white's pieces are miserably uncoordinated, that his king is laughably naked and that the black pawns are going to have their say. The knight's ability to jump to e4 was one of the few remaining resources for white in the whole position. In a sense, black has just eliminated white's best piece! My opponent was very proud of this exchange sacrifice after the game, and he had every right to be; it was definitely the best move in the position, and one that I had not even considered.

24. bxc3 Nxd5
25. Ba5 Bg4+
26. Kc1



White's pieces scatter like roaches exposed to light as the position goes from desperate to pathetic.



My only saving grace here was that my opponent had consumed gobs of time wending his way through the opening and deciding on his exchange sac, and now had precious few minutes to reach 40 moves.

26. ... h5
27. Bh3

Disregarding admonitions about avoiding exchanges as the inferior side, I had a sudden notion that I might be able to work my king toward d3 and suppress the central pawns if I got my opponent's light-squared bishop off the board. It's really just a fantasy...

27. ... Bxh3
28. Rxh3 h4
29. Kd2 Bf4+?!



Black starts to lose the thread. 29. ... e4 was fairly decisive, and even a move like 29. ... g5 would have maintained the full advantage.

30. Ke2

Sliding toward the white squares.

30. ... g5
31. Rg1

And somehow, white is crawling back into this game!



Of course, 31. Rd1 is objectively stronger, but I was the midst of hatching a hair-brained scheme. I had decided here that my only chance of holding was to maximize piece activity. Black's last several moves made his bishop passive, and I came to the conclusion that if I could only activate my own bishop, I might be able to pick off a few of his pawns.

31. ... Kf7
32. c4?!?

And here it is, a ridiculous idea. White jettisons both of his c-pawns to activate his bishop. This is definitely a losing plan, but recall that my opponent was running very low on time and getting increasingly nervous, lest he make a mistake. I had calculated a sequence resulting in the exchange of a pair of rooks, and I thought that the resulting position would give me good chances to pick off several of black's pawns. It was certainly worth a shot!

32. ... Rxc4
33. Bd8 Rxc2+
34. Kf3 Rc3+
35. Kg4 Rxh3
36. Kxh3



Objectively, white is dead lost, but his suddenly active pieces now have an opportunity to begin collecting pawns.

36. ... Kg6
37. Rd1 Ne3
38. Rxd6+ Kh5



And now came the ridiculous inspiration. White will get rolled off the board by the combined force of the h and g pawns unless he does something drastic, hence:

39. Bxg5 Kxg5
40. Rb6 Nc4

Here we had reached 40 moves, but my opponent didn't know it and kept blitzing out his replies.

41. Rxb7 Nxa3
42. Rg7+ Kf5?
43. Kxh4



White has achieved more than he could realistically have hoped for. Black's split-second decision to jettison his h-pawn instead of playing passively, eg. 43. ... Kh5 44. Rh7+ Bh6 has suddenly centered the entire game on a discussion of whether black knows how to mate with bishop and knight! At about this point, my opponent looked at me and queried as to whether we had reached 40 moves. When I answered in the affirmative, he got up for a long stretch and breather while I sat and pondered.

The question I faced here was whether I should sac the rook for the pawn at the first opportunity, trusting that my opponent would not know what he was doing, or whether I should maneuver for a while, seeing how he played it. It seemed to me that there would probably be multiple opportunities to sacrifice the rook (preferably at a point where my king had achieved greater centralization) and that my opponent's play would give me cues about his confidence. If he just calmly advanced the pawn, disregarding the threat of the rook sac, I would know that he knew the endgame (or had a good poker face), and I was probably done for. On the other hand, if he attempted to shelter the pawn with his pieces, I might be able to torture for a long time...

43. ... Ke4
44. Kg4?

Argh. It turns out that this straightforward move is a mistake, as I cut off my own king's path to the front of the pawn, making the defense significantly more difficult. Preferable (probably) was 33. Kh3 Nc2 45. Kg2 Ke3 46. Kf1

44. ... Nc2
45. Re7 Ne3+

And here my opponent shows some understanding of knight and bishop coordination--not a good sign. He has built a wall along the g-file, cutting off my king.



Gloomily, I retreated.

46. Kh5 Kf3
47. Kg6 e4
48. Re8

A pointless move, but what else is there? I was getting ready to sac the rook... or resign.

48. ... Ng4?!



But what's this??!! The most obvious move was 48. ... Nc7 or Ng7, covering all the squares to the first rank and forcing the rook sac. Instead, my opponent seems to be preparing to play his knight to e5 to shield the pawn. Seeing the possibility of queening and thus winning without being obliged to demonstrate an understanding of bishop-knight mating, he leaps for it. With a new lease on life, I suddenly sank into thought. If my opponent is honestly nervous about knight and bishop mate, I thought, perhaps I should hold off sacrificing the rook and maneuver still further. The contortions that black will have to go to keep the rook from taking the pawn will be difficult and stressful, and he might make a mistake!

49. Kf5 Ne5
50. Ra8

Of course not 50. Rxe5?? winning the two pieces and losing the game.

50. ... Nd3
51. Ra3

I begin the process of torturing him.

51. ... Bd2
52. Rb3



My play here is risky; objectively I should have sac'd the rook already, but it remains virtually impossible for black to promote the pawn without allowing the rook to capture.

52. ... Bb4
53. Ke6

And here, after a long think, my opponent suddenly sat up and confidently slammed down the move Nc5+.



Of course, this is a lovely impossibility, and I duly pointed out to him that he had just dropped his king, and we had to stop the clocks and have the tournament director assess a time penalty to his clock. I think this is what ultimately did him in. He was so flustered to have made such an oversight that he could no longer summon the courage to even consider the knight and bishop mate.

53. ... Kf2
54. Kd5 e3
55. Rb1 Bd2



Black harbors dreams of playing 56. Nc1 and forcing through the pawn. Of course, this overlooks the threat of Rb2, forcing a king move that will enable the white rook to redeploy to attack the pawn from the rear.

56. Rd1 Nb2
57. Rh1 Na4

This knight move is quite strange. Obviously, 57. ... e2 was the move, but again that faces the rook sac that black just can't bring himself to encourage.

58. Kd4 Nc3?



Black has finally completely lost it. I almost began to pity him. His eccentric knight maneuver, perhaps conceived in an attempt to bring the knight around to g1 and thus shield the pawn, has let the white king in to d3 where it will finally be able to join the fight and prevent the black pawn from queening. Now black has to work hard to achieve the bishop and knight endgame without allowing a rook for piece and pawn exchange.

59. Kd3 Nd5??



Ayieee! The stress has finally broken him and he completely cracks up. Relatively best was 59. ... Kg2 after which black has to invest half a dozen moves to force the original rook sac. Instead, he has allowed the terminating stroke:

60. Rh2+

After which, Rxd2 wins the bishop and pawn for the rook, draw.

1/2-1/2

Not objectively a great game, and certainly not an accurate one, but it was one of my more dramatic chess battles. The lesson here? Well, I won't indulge in a lecture on gambler's fallacies. Knight and bishop mate endgames are very rare, so it's probably not worth studying the patterns for most players. That said, there's little worse than finding yourself with an obviously winning position and not knowing how to win it...