Mr. Nakamura was very nice during the whole tournament, and I discovered he has been a part-time resident of my new city as well for the last number of years. That partially solves the mystery (to me!) of his ICC handle, CapilanoBridge. I didn't succeed in getting the gossip on what he's doing in Vancouver at all however! The tournament was held at the Grand Pacific hotel in Victoria, which was a very nice location. It's right downtown, and a pretty swanky hotel. The deal for the Canadian Open at the same location in July looks like a pretty good one if you're planning on making the trip. I would advise though that you pay attention to the time controls used: game in 90 minutes, with 30 second increment. It is deceptively quick, and if you hadn't read the details (I'll sheepishly put my hand up here), it was quite the surprise to find at move 40 that your 2 remaining minutes were all you still had until the end of the game! So maybe because of that, or because he's a fast time control cyborg (he's widely believed to be have been the best in the world at 1-minute chess, and perhaps 3-minute as well) Mr. Nakamura moved *very* fast in all of his games. He didn't spend much time at the board, so if you were following online, what seemed to be him thinking was more likely him noticing it was his move and casually strolling back to his board to make his next move virtually instantly. The experience was very much like playing a computer - virtually instant moves that you didn't foresee, that you then realized were better than anything you expected, and when washed and repeated, you quickly run into problems with running out of time! It was definitely a unique opportunity for all of us lucky to be paired with him, especially as when asked later if he might play again next year, he laughed and just responded "Unlikely!". Alright, to the game: I browsed through some of his recent games, but that was more to get a feeling for what he was unlikely to play! He's a top player, so any preparation he's done isn't going to be whipped out for the likes of us. Although I still expected to be surprised, as his other games all had an anti-theoretical feel to them.

1. d4 Nf6 2. c4 Nc6 Well that definitely was a surprise! I've actually played this position once before in a tournament, I followed the same development as in that game, but likely erroneously. 3. Nf3 3. Nc3 e5 3... e6 3. Nf3 is the most common move here, as 3. Nc3 allows black to play e5, and although white is of course fine, one might even argue 1. d4 is designed to prevent an easy e5 from black without conceded something in return (weakness, space or material). It's important black hasn't yet played d6, as his bishop quickly finds a productive use outside his pawn chain. The other obvious option is d5, but it's also not clear how to take advantage of the knights in the middle without overextending yourself. I'm sure it's a good move, but not one you want to be figuring out at the board against someone who is more familiar than you with the position. More like one you play back immediately and confidently in a way that somehow says "yes, I am prepared for this and I'm going to go for your throat, are you sure you remember it as well as you think you do?". Strike fear into their hearts my friends! 4. g3 4. a3! 4... d5 It seems 4. a3 is a more common move, and this does make sense. Playing Nc3 allows Bb4 and a sort of nimzo-indian/queen's gambit declined position will be reached where black has committed to an early Nc6. In both of those openings, there are rarer lines where Nc6 is played, and as they are not so common, their positional ideas are probably not as widely understood. This seems like a decent tradeoff playing as black. Instead a3 does seem somewhat slow, but after d5, a queen's gambit position is likely reached where there is no pressure with Bb4, and white has (perhaps) lost time with a3, but reasons black will necessarily need to lose time as well by moving his knight on c6. If white proceeds with say a minority attack on the queenside, perhaps a3 would have been necessary anyway! But also, perhaps black's knight is en route to a different home than usual (normal for him to live on d7 for a while) via e7 or perhaps even a5! Looking in the database, I've found Nakamura has defended this position several times as black against other GMs (no losses, several wins, it seems I'm in good company ;), but has preferred yet another idea: d6. Since white has wasted time on a3, the bishop is doomed to be a little passive for the foreseeable future, so use the time you've gained to try for the e5 break instead. This definitely takes the game away from queen's gambit waters and toward something more like a Benoni pawn structure. In this kind of pawn structure, there is no obvious benefit to having played a3. Presumably this is also why black doesn't just play d6 on the third move - now if d5 he has the option to play exd, freeing at least one his bishops and unbalancing the pawn structure. There are many gaps in my chess understanding. One problem for me has long been that I've never found a suitable response to the Nimzo-Indian. For some reason, my brain just does not lend itself to thinking of the natural plans and moves that white usually goes for in the main lines (4. e3 and 4. Qc2 both it seems!) and never having put any time in to fix this, I've toyed with other ways to avoid this. One I've tried has been 4. g3, but that uncovered a new problem: my brain also does not understand Catalan positions! I've watched many wonderful Kramnik games, but it seems I always get into trouble trying to go that route. So I decided to go for an exchange pawn structure. It's in no way challenging to black's position, but at least for the moment, I've been burned too many times to try playing a position after say Bg2 dxc4 Qa4 Bb4+ Bd2 Nd5. I just know what kind of trouble I get myself into! So instead I decided to head for a pawn structure I am better acquainted with, even if it's not very aggressive with my particular piece placements. 5. cxd5 5. Bg2 dxc4 6. Qa4 Bb4+ 7. Bd2 Nd5 5... exd5 6. Bg2 Bd6 Black's bishop looks well placed here, as it can't be harrassed by a pawn anymore. Perhaps one day it'll help in an attack on white's king. 7. O-O h6 This move I didn't expect as castling is certainly very natural. However I think it's better. White definitely has a problem finding a home for his dark squared bishop now. Perhaps I should have played Bg5 instead of castling myself, but then I expect black would have played h6 anyway - taking the two bishops for himself. That's not ideal from my point of view, but given how hard it is to find a decent home for him, and the problems that causes in my own co-ordination in hindsight it seems like the better option. I might be biased by the resulting play though - perhaps the truth may be that white is just best off allowing dxc to happen and accepting that it gets messy. 8. Nc3 O-O Ok, so I'm mostly developed, except for the bishop, but where exactly should it go? I did briefly consider Bf4, but it seemed like it would risky in the long term once black exchanges. I think white is quite safe in the short term, but one day black might reorganize with Kh8 and Rg8 and g5 will dangerously open my king. It's also not clear what white's plan should be. Putting a knight on e5 seems to be only temporary - one day f6 will kick it and then where? Black has very good light square control and there are no enticing looking pawn breaks. Maybe I'm solid enough, but creating weaknesses and removing possible active plans seems like a bad idea. So instead I opted for the minority attack, with a minor threat to go after his bishop with Nb5 if he plays Re8. If Nb5 immediately, Be7 seems fine as Bf4 a6 Nxc7 Ra7 will lose the knight on c7 after Bd6. Bxc7 Qd7 leaves two pieces under attack. With the rook on e8 however, Nxc7 would be a fork, so I would win material. 9. a3 Re8 He plays it anyway! So I figured why not take the bishop pair. This is arguably not a great decision as my own dark squared bishop is such a terrible piece, but playing b4 doesn't really look like it's getting anywhere. It's already quite hard to find a plan for white which makes sense! At least in this case I can hope the bishops will one day have their day in the sun, and I have pretty clear plan: keep the center and kingside together while I improve that unopposed bishop on the queenside and slowly pursue the minority attack. 10. Nb5 Ne4 11. Nxd6 Nxd6 Here I noticed that Ne5 was a nice tactical opportunity. I'm threatening the d5 pawn, threatening to take his knight and ruin his pawn structure, and if he exchanges knights, I'll open the position slightly (advantage: bishops!) and regain the pawn, with some uncomfortable queenside pressure. I was quite pleased with myself, but somehow overlooked his obvious response! 12. Ne5 Ne7

I think he had time to blink after writing my move down before responding, but my stomache sank a little - of course Ne7 is the best move, it takes care of all of those threats, while keeping the position more closed, and while improving his knight to non-awkward square! I initially felt that my knight was now the worse placed one on e5 - it can't stay, and it just retreats to f3, I've pretty clearly just wasted time while helping him. And I wasted more time calculating possibilities that were never even on the cards. Oy! So what to do instead? Well, playing for b4 and b5 only makes sense once there's a pawn to attack on c6. Furthermore, his knight on d6 is so well placed that it helps prevent that b5 break, and is also ready to jump into the c4 square. Since he can easily back it up with b5, and follow with a quick a5 perhaps, it's possible I would be the one who would get in trouble over there! The only other pawn break I could see would be to play for e4. I can't just play Re1 and e4 as black would gladly trade his knight for my light squared bishop leaving his unopposed with my airy kingside. So I would need to support it with the move f3, but that will create weaknesses on the e-file if something goes wrong. The move didn't quite feel right yet, so I opted instead to simply restrain his d6 knight while developing my bishop and "await further events" as the russians say. 13. b3 c6 Well at least now the minority attack has a target. Of course my light square bishop also looks pretty ineffective. 14. a4 a5 I decided putting the bishop on a3 was the better option - it doesn't seem to have a future on b2. In that case It would be helpful in protecting d4 if I played for the e4 centre, but it also is no longer defending e3: As soon as I play f3, it would be met by f6 and Nef5 and then hopping into e3. So back to the queenside: 15. Ba3 Nef5 16. e3 Ne4 Nef5 put a little pressure on d4, so I simply defended it some more to free my queen and remove any hidden tactics based on removing the defender of the e5 knight. So now I am almost fully developed - the rooks are the last thing to bring into play. But where should they go? a Rook to b1 makes sense, as then the b-file can be opened and I'll have a target at b7 to attack. It's entirely possible my pawn on a4 will even end up weaker in that case, but I'm not sure what else to do! Unfortunately black's pesky knights seem to be controlling all the useful squares and it's hard to find a reorganization without walking into a fork. Furthermore, what is black up to? Normally in these pawn structures he attacks on the kingside, and it seems as though pieces are gradually migrating that way. Maybe he's ready for an h-pawn thrust with h5 on the next move now. I will have to watch out for a weakening on the f2/e3/g3 complex as it's common for a knight sacrifice to occur around those spots. h5 meets all those aims: advancing it to h4 will weak all those squares a bit, and will also further open lines around my king. The move Ng5 is in the air to start invading on the light squares. An alternative plan might be to continue reorganizing with Nfd6 and following up with g5 and f4, again aiming to weaken my king and create a weakness on e3 or d4, and an attack as well! I figured since I was a bit scared of waiting around, and it's not clear my rooks can be improved at the moment, I needed to seize the moment and start something active on the queenside. 17. b4 axb4 18. Bxb4 Nfd6 So either he's going for plan #2 listed above, or, there is a new plan to target my a4 pawn. 19. Qe2 f6 20. Nd3 b6

Alright, so my knight has been kicked, and it seems black intends to target a4. I can't easily get rid of the pawn as if I push a5, he'll push past with b5, leaving a5 still weak, and himself with his own protected passer on b5, and a gigantic steed on c4. Both his pawns are a little backward right now, but c6 is very hard to attack as he can block the c-file with his knight. Likewise, b6 will be defended by a knight from c4, and since I have no way to get my bishop or knight over there, I'm out of luck. I did notice however that his kingside is a little bit airy with the light squares, and I figured he probably too wants to develop his bishop so that he likely double his rooks on the a-file. Clearly I should get my knight to f4 to pressure those light squares, and the d5/e4 squares. I can't do that immediately because of Ba6 winning an exchange, but I can remove that threat with tempo with Rfc1. Black will play Nc4 as noted above, and then Nf4 is safe. So: 21. Rfc1 Nc4 22. Nf4 Qd7 22... Bb7 23. Qg4 After I played Nf4, he slowed down a bit. He looked a bit suspicious about his kingside it does look like white finally has a bit of something to nibble on: If he goes Ba6, he abandons the h3-c8 diagonal and white's queen can start moving over with Qg4. Suddenly, Nh5 is in the air and white has some counterplay. But on the other hand, simply developing the bishop with Bd7 will allow the exchange sacrifice Rxc4 where white is very happy to remove one of those knights and pick up at least one pawn for it. Likely the the c6-pawn will also drop and white appears to have excellent compensation. So instead after thinking a bit, Qd7 is the answer: control the g5 square myself first, while leaving the B free to come to a6 in the event of Rxc4. Here however I noticed that maybe Rxc4 is possible anyway! Black can't take with dxc as white gets good compensation with the open diagonal, and two pawns. Instead he pins the rook with Ba6, and white has the sneaky move Bh3: If black continues Bxc4 Bxd7 Bxd3 the knight on f4 can recapture that bishop, Nxd3 and white will have two pieces for the rook! So black will have to spend a move moving his queen. This means that will also need to abandon the g4 square and white will have time to move his forces over there. Is there enough compensation? I thought likely not, but since I'm down to about 8 minutes to finish the game, it's going to be a lot easier if I need to make him think and if I'm the one with the threats. I tried to calculate what happens after the likely moves Rxc4 Ba6 Bh3 Qf7 Bf5 Bxc4 Qg4: White owns the light squares on the kingside and the black light squared bishop has been distracted far away. It appears that at the very least white can regain the exchange with the moves Bg6 or Be6 - that seems like I would still be worse there - black's remaining pieces are still better able to attack my weaknesses (a4 in particular) and I'm less able to respond. However, maybe the opposite colored bishops would offer some help in making a fortress. More interesting was the prospect of invading with the knight. Maybe something like Bg6 Qb7 Qf5 is possible... Then there is the threat of Bh7+ Kh8 Ng6+, that looks very dangerous indeed! And it seems like there are a number of options for discovered perpetual checks to boot. Those are probably unacceptable for black to give in to, so it's likely he'll choose something unclear instead. No time left to really calculate, so I'll just go for it! (Qh5 Qf7 I also considered briefly but it seemed like I would remain passive and with little time I didn't feel comfortable trading queens. The computer however, gives Qh5 Qf7 Qxf7 Kxf7 f3 Ng5 e4 Ne6 Nxd5! (nice one) and it's hard to see how black can fruitfully play for a win. He shouldn't have much trouble drawing after cxd exd Ne3 dxe6+ Bxe6 as the opposite color bishops result, and the queenside a and b pawns will be liquidated. Amazing how far off my feeling about a particular position really is!) 23. Rxc4 23. Qh5 Qf7 24. Qxf7+ Kxf7 25. f3 Ng5 26. e4 Ne6 27. Nxd5! cxd5 28. exd5 Ne3 29. dxe6+ Bxe6 30. Rc7+ Kg6 31. Be1 Nxg2 32. Kxg2= 23... Ba6 23... dxc4? 24. Qxc4+ Kh7 25. d5!|^ After I played Rxc4 he stopped to think for a while. Clearly he hadn't looked deep enough before and he used quite a bit of time here decided what to do. In the end he appeared quite angry that Ba6 was still the best option and his king was going to have to suffer. The next few moves were banged out quite violently at blitz pace! 24. Bh3 Qf7 24... Bxc4?? 25. Bxd7 Bxe2 26. Nxe2 Red8 27. Bxc6+- 25. Bf5 Bxc4 26. Qg4 c5

Yes, this seems best, cutting off at least one white attacker. The bishop has no where pleasant to go, so Be1 seems forced, if a bit scary. cxd exd might follow, and black might soon be able to respond with his own exchange sacrifice: Rxa4! removing the defender of the Be1 with perhaps a counterattack to follow. 27. Be1 cxd4 Ok, so those threats above might come to pass, or perhaps the rook on e8 is just being improved. White has access to the b4 square again though, so the bishop can return from the dead. White also doesn't need to recapture immediately as perhaps he'd like to make a bishop move intermezzo, or even just go for the light squares and leave the pawns altogether! Unfortunately, he has no time left to consider any of these options really. I tried calculating as quick as I could, but it's complicated and I can't see a clear attack or draw, so I just recaptured as it seemed somewhat safe, and intuitively I felt that white didn't yet have to be forceful. This is likely the critical position to consider though. 28. Bg6 Qe7 leads to: a1) exd the computer evaluates this as very slightly better for black after b5 a5 b4 clamping down on that Be1 (how fitting that it still is a terrible piece - and that my decision to try to improve it seems entirely wrong!) White has nothing better than taking the rook as on Qf5 Ng5 defends everything with tempo and white is toast. So in full: 29. exd b5 30. a5 b4 31. Bxe8 Rxe8 32. Qh5! Rd8 and a kind of dynamic balance is reached: we both have weak light squares but black's passed pawn is obviously the more dangerous and the Be1 is dominated. But black's rook must defend d5, and his queen must prevent Qe6+, so it's not yet clear how to make progress (but no doubt Nakamura would find a way to outplay me - even without my low time!) a2) Bxe8 Qxe8 (Rxe8 exd the computer calls even! I guess because there is no passed b-pawn yet - black must lose time going back with Ra8 to push it past and white can improve in the meantime with h4... for example Ra8 h4 b5 a5 b4 Rb1! now works: Nc3 Bxc3 Qf5 guards against c2, and black can't take a5 due to Rb8+) exd b5 a5 Ng5 (with threat of Qe4 winning) h4 f5! Qd1 Qxe1+ Qxe1+ Nf3+ and black wins the a-pawn. b1) Bf5 g5 (dxe?? Be6+ wins for white) exd b5 Be6+! Kg7 a5 and white gets one pawn back. Instead of now giving up the dark squares the computer actually recommends black liquidate to an ending by force: f5 Bxf5 Nf6 Ne6+ Qxe6 Bxe6 Nxg4 Bxg4 Rxe1 Rxe1 Rxa5. Again, black has a passed pawn, white doesn't, white's king is the little weaker. Tough defence, but to my eyes not a no-brainer. Good chances for a black win. b2) Bf5 g5 exd Ra7 Kg2 is hard to progress against. White threatens Ne6, followed by Qh5 in some lines. For example: b5 a5 b4? Be6+ Kg7 Bxd5 Bxd5 Nxd5 and white will capture on b4 the discovered check after Qb7 Nxb4 has no threat, white simply blocks with Qf3. And so white should be fine here with two pawns for the exchange and no black passed pawns. Playing Kg7 instead runs into Ne6+ Kg8 Qh5 Qf7 (what else to stop Qg6+-h6+-g6+?) Bg6 Qxe6 Bxe8 and again white is more than fine! 28. exd4 28. Bg6 Qe7 29. exd4 (29. Bxe8 Qxe8 30. exd4 b5 31. a5 Ng5 32. h4 f5! 33. Qd1 Qxe1+ 34. Qxe1 Nf3+ 35. Kg2 Nxe1+ 36. Rxe1 Rxa5-/+) (29. Bf5 g5 (29... dxe3?? 30. Be6++-) 30. exd4 b5 (30... Ra7 31. Kg2~~) 31. Be6+! Kg7 32. a5 f5 33. Bxf5 Nf6 34. Ne6+ Qxe6 35. Bxe6 Nxg4 36. Bxg4 Rxe1+ 37. Rxe1 Rxa5=/+) 29... b5 30. a5 b4! 31. Bxe8 Rxe8 32. Qh5 Rd8=/+ 28... g5 This is similar, however the black queen watches over h5 instead of b4. Here again white has a tough choice: take back the exchange, or play for attack with Bb4. Well in hindsight the "attack" is much harder and messier than I expected and it's easier to defend than expected after taking back the exchange, so clearly Bg6 is the best move. :) Not such an easy choice with ~3-4 minutes for my skill level. Clearly I can improve here... Bg6! Qe7 Bxe8 Qxe8 Ng2 Black threatened h5 winning the knight. The N can reemerge on e3 and easily guard all the light squares, and in fact also is useful from g2 guarding the bishop on e1! An odd if effective co-ordination. The a4 pawn is actually poisoned - not because after Rxa4 Rxa4 Qxa4 Qe6+ Kh7 white gets the pawn back with Qxb6, but because after f3! the knight on e4 is actually trapped and white wins. Without winning a pawn, white should hold on easily. Black might even need to take care soon after h4 and Ne3-f5 to ensure his own king has adequate cover.

Alas, I played the optically better move on instinct, without calculating much. 29. Bb4? 29. Bg6! Qe7 30. Bxe8 Qxe8 31. Ng2= Rxa4?? 32. Rxa4 Qxa4 33. Qe6+ Kg7 34. f3+- 29... h5! And now I realized I was in trouble. h5 decoys my knight unto an unfortunate square - no as much light square influence and a target for a future Be2 skewer. 30. Nxh5 Nxf2 If I play Kxf2 Be2 wins back the knight and the game. So that only leaves Qf3 as a possible move, but it seemed to me that Be2 again would have the same effect, so with my clock ticking below the 3 minute mark (he was still north of 45 minutes) I resigned. However, this came as a surprise as he had expected: 31. Qf3 and in the case of Be2 32. Nxf6+! is possible. Now if Kh8 Qxf2 Qxf6 Bd7! seems like excellent drawing chances. White has one pawn for the exchange and will liquidate the b-pawn leaving no weaknesses or passed pawns. It's hard to see how black wins here, although it may be possible. Instead Qxf6 Qxd5+ Kh8 Qd7! threatens mate and limits black's options. Qg7 leads to endgame where I have two pawns for the exchange, but it's not so easy to get rid of the b-pawn. More play left, but the computer at least believes it may be balanced. White will also have more counterplay with his now-passed d-pawn. I think I agree. However black has an even better move to stop mate: Qh6! I can't take back the knight due to Qxh2+, and black is suddenly all over my own king. Since d5 doesn't work: d5 g4! Bc3+ Kg8 Be6+ Rxe6 kills the attack and black has his own mate with Nh3+ building a net around the white king. So I feel justified in resigning after all, although I definitely regret not going for Bg6! 0-1 [Kenney]