Last Tuesday, on May 24, was the ninth and final round of the Pandanet Go European Team Championship. The final round was important in deciding which countries get to play in the European Team Championship tournament in the European Go Congress 2011 in Bordeaux, but it was also crucial in determining which team drops from the A league to the B league next year. The system is such that the last team of the league drops automatically and gets replaced with the winning team of the lower league, and the second to last team of the league plays a qualifying match with the second best team of the lower league. The ninth-round game between Finland and Serbia was to determine which team drops automatically, and which team gets to qualify.
The situation before the round was exceedingly exciting: Finland and Serbia were tied in both game points and board points. If a tie like this were to occur after the ninth round, too, the next tie-breaker would be the number of first-board wins; before the ninth round, Finland and Serbia were tied on this part, as well. In a sense, the first-board game of the match was worth two games. Having known about this situation well in advance, I had been training a lot during three weeks between rounds 8 and 9, my main methods of training having been doing tsumego and reviewing professional games.
The game started as shown here. I had black, and my opponent was Dusan Mitic 6 dan, the first board of Serbia. More advanced players may notice the unusual fuseki choice of black 7, and the even more unusual one of black 9. Instead of 7, black would normally play something in the lower left corner — but having known that the game would be both important and that it would attract a lot of spectators, I wanted to play something more unusual and exciting. It is also for this reason that I chose to play 11 on the upper side, occupying a big point, instead of answering to white 10 in the lower right corner. In a way, you could also say that I was trying to intimidate my opponent!
The right side being more important, it’s correct for white to first extend with 12, and then to play the knight’s move of 14. I chose then to turn with black 15, to make sure that the triangle-marked stone on the lower side will be of some use in the future. You could say that it’s a bit too far away, and that it should be one space to the right — however, since the black group will not be in danger for the time being, I considered it fine enough. Black 17, then, is a nice way to strengthen the upper right corner, effectively preventing another 3-3 invasion. The rest of the moves up to 21 are fairly simple and reasonable.
White, quite predictably, then invaded the upper right corner with 22: this was clearly the biggest unsettled part of the board. The sequence up to 29 was pretty much what I had been expecting, but white 30 surprised me. The following variation shows how white would normally play.
Returning to Figure 3, I of course resisted white’s plan with black 31, capturing the four white stones on the upper side. Up to white 44, I considered the result to be good for black: black had gotten a fair amount of territory, and white still had an unsettled group, even if it could not be called weak. Next I set out to attack the white group on a large scale, however making an oversight in the process.
Black 45 I played to disturb the shape of the white lower right corner group, and also to make sure that the black group on the right side doesn’t get too weak. However, the knight’s moves up to 51 don’t do black any good: they don’t really threaten white, nor do they build anything for black. Mostly, they just serve to make sure that the white left-side position doesn’t get any bigger, and perhaps helps a bit to invade the left side later. Instead of 47, it seems that a move at 48 is better, making white’s shape more uncomfortable.
White’s cut, starting with 54, was effective: the three black stones in the centre are cut and not too strong, and white also got 62 in sente, making sure that the upper-right white group will have two eyes. After 66, it felt like white had the upper hand, and that I would have to do something to turn the game around.
I chose not simply to defend against the forcing white move of 66, but instead struck back with 67, 69 and 71: in this fashion, the centre white group could not easily connect out. 73 also works at keeping the white groups separated, though I liked the double large knight’s move extensions of 69-73-triangle as well. Even further, black 73 would also serve to destroy the white left-side position…
…That is, would I have cared to do something about it. I thought on black 77 for quite a while, considering all kinds of possibilities like black A or black 78. Actually, I most likely thought for too long: I ended up confused about all the possibilities, and got optimistic about the white centre group “maybe” dying — of course that isn’t reasonable thinking. White 78 was a very good refutation of black 77. Black 79 tries to save the situation, but after white has cut with 84 and 86, black is in dire straits, having two weak groups to fend for.
Instead of black 89 here, black could also connect at 90, threatening to play black A next. White would defend, and black would need to add one more stone to the lower-left corner group for it to live; then white would attack the centre, and black would have no way of winning. Instead, I chose to go with 89, sacrificing the lower-left group. The sequence up to black 101 is not too shabby, but the double attack black is dreaming of should not actually work, thanks to a crucial weakness at B. White 104 effectively makes use of it, too. During the game, it actually felt like black was under attack here, not white — yet, there was only one way to go.
Black 5 etc. started the last battle of the game. It would seem that if white instead of 18 had played at 7, black couldn’t have profited from the attack. Were the game to continue until the end like that, white would probably have won by about five to ten points. In the game, black gained a good amount of profit in the upper left corner, and after black 55, six white stones become killable as well: that gave black an extra 7.5 points (black creates a 15-point move on the board for free: with about a 50% chance to get to play the move, it’s reasonable to call the black profit 7.5 points).
In the end — even to my surprise! — black won by 6.5 points. The middle game had felt very difficult for black, and even the double attack had not given a very noticeable profit. Yet, somehow black ended up having the point lead. The rest of the game is visible below. Some mistakes in the endgame do happen, but it’s not too clear they affect the outcome of the game.
In the end, Finland emerged “victorious” with a result of 2-2, including the first-board win! Apparently, later into the game, Finland was cheered on by Dutchmen as well: were a 1-3 result to happen, Serbia would automatically stay in the A league, and The Netherlands would have to play a qualifying game with the second best team of the B league.
Of course, we should not stop at thinking about who drops from the league and who stays; the most important thing for me was being able to play exciting tournament games, belonging to a team representing a whole country.
Finland’s qualifying game with Poland will take place on Tuesday, June 14, on the Internet Go Server (IGS). Stay tuned for another set of exciting tournament games!