Blogging from 32,000 feet

Hello again!! Thanks to the miracle of modern technology I am writing my first blog post from midair — specifically from American Airlines flight 1861 from Dallas to San Jose. Another of the miracles of modern technology is that I just lost 15 minutes of work. I don’t know whether it has to do with the Internet connection or hitting the wrong key on my new netbook computer that I’m still getting used to. Anyway, the post that I had been working on for 15 minutes just vanished into thin air, leaving only the title, “Blogging from 32,000 feet.”   🙁

Oh well, here we go again. This time I’ll save my work a little bit more often.

I promised you some positions and analysis from the U.S. Senior Open, so here we go. The first position is from my round four game against Dale Sharp, in which I was White. The full PGN of the game is here.

Black to move.

The opening was a King’s Gambit, in which Black tried to protect the gambit pawn with … g5 but was not entirely successful. I also went on a somewhat dubious adventure, taking his rook on a8 in the hope that I would be able to rescue my knight somehow with an advantage in material.

Here Dale surprised me by playing a bishop sacrifice, 17. … Bxc2. What do you think about this sacrifice? Is it sound? Should White take the bishop, or is there a better alternative?

While you’re thinking about that, let me discuss Dale’s thinking, which is interesting because it points out one of the benefits of a postmortem analysis: the fact that you can see how another player approaches the same position. I had spent most of my time analyzing 17. … Kb8, to which I was planning to reply 18. a5 Kxa8 19. a6 Bc8 20. ab+ Bxb7. White has lost the knight, but as compensation he has exposed the Black king, and in fact I think White stands much better.

But Dale said that he hardly even looked at 17. … Kb8. “Too slow!” he said. Instead, he had been thinking about how to activate his worst piece, the knight that is currently penned in on h6. So he had been looking for the best place to move his f5 bishop to, when it suddenly occurred to him that he could just sacrifice it on c2! If I took the bishop, he would then play 18. … Nf5 with “the mother  of all fork threats,” as he explained it.

Here I made the mistake of believing my opponent’s analysis. I didn’t like the knight coming in either, and so I took it off with 18. Rxh6.  The next few moves were pretty obvious: 18. … Bd3+! (a useful intermezzo, pulling the queen away from the defense of g2) 19. Qxd3 Qxg2+ 20. Ke1 Bxh6. Now I played 21. Qb5, threatening various checkmates on the queenside, but he played 21. … Qh1+ and I had to admit that his attack was quicker. I played 22. Qf1 Qxf1+ 23. Kxf1 Kd7 and we got to this position.

After the smoke cleared.

Black will win his piece back, and the endgame is about equal. I managed to mess it up and was very fortunate to draw, but that is another story.

Back to the original position. After 17. … Bxc2 I should have taken the sacrifice! After 18. Qxc2 Nf5 19. Re1 White is just fine! Yes, Black wins back the exchange with 19. … Ng3+ 20. Kg1 Nxh1 21. Kxh1, but after 21. … Qh5+ 22. Nh2 (diagram) White’s king is as snug as a bug in a rug.

After 22. Nh2 (analysis). A much better position for White.

White’s threats on the queenside are very menacing, and Black’s attack on the kingside is easy for White to fight off. For example, if 22. … Bh6, threatening … Bf4, White can play 23. Bf3 Qh4 24. Re4 Bf4 25. Rxf4! Qxf4 26. d5 winning.

Moral: NEVER TRUST YOUR OPPONENT!

My game in round 5, against Tim Rogalski, was also very interesting, even though in the end I blew the game with a horrible blunder. It’s a hard game to analyze with the computer, because the computer program Fritz is completely convinced that its namesake variation, the Fritz Variation of the Two Knights Defense, is rubbish. Perhaps it is rubbish if your opponent plays like a 2800+ computer. However, it is well documented that humans do not play like computers, and so I still think the variation is quite playable against human opponents.

Because the computer does not really “get” this opening, I’ll have to do a more careful analysis after I get home. But I’ll show you some of the highlights, along with some quick and dirty analysis. The full PGN is here.

Rogalski — Mackenzie, U.S. Senior Open 2009

1. e4 e5 2. Nf3 Nc6 3. Bc4 Nf6 4. Ng5 d5 5. ed Nd4 (the Fritz Variation) 6. c3 b5 7. Bf1 Nxd5 8. Ne4 Qh4 9. Ng3 Bb7!?

Through White’s 9th move we were following the famous correspondence game Berliner-Estrin. Berliner played 9. … Bg4 and won, but there are serious doubts about the soundness of Black’s piece sacrifice after 10. f3. With the text move, Black offers a piece sacrifice in a different way, essentially disregarding the material situation in order to achieve maximum piece coordination.

The line was no surprise to my opponent, who said he plays the Two Knights as Black and has always wanted to play the 9. … Bb7 line in a tournament, but has never gotten the chance.

10. cd O-O-O 11. h3 h5

My first long think. My opponent thought this move was forced, but in fact 11. … Nb4 is also a serious possibility, and also I thought about 11. … f5, although I concluded that it accomplishes nothing after 12. Qh5. Also, I knew that I had played one tournament game before against 11. h3 and was trying to remember how that game went, but I couldn’t do it.

12. de …

Why not? In this position a lot of humans will start having attacks of conscience and think about their development instead of eating more material. Unlike us, the computer has no conscience. It seems as if some of the computer’s materialistic philosophy has entered Rogalski’s game as well. In effect, he asks me: “Whaddaya got?”

Black to move.

Now I played my most spectacular move of the tournament. I’m not sure if it was a good move, but it was certainly spectacular!

12. … Ne3!?

Tim said that when I picked up my knight he was sure that I was going to move it to f4, and when it went to e3 he thought I had accidentally put it on the wrong square!

It’s possible that 12. … Bc5 or even 12. … Nf4 are better, but this move was an attempt to profit from the weakening of White’s position that resulted from 11. h3. Notice that the knight cannot be taken with the f-pawn because of … Qxg3+ — a possibility that would not have existed if the pawn were still on h2.

13. de …

Also an important and unforced decision. White might do better to decline the knight sac with 13. Qb3.

13. … Rxd1+ 14. Kxd1 Bb4 15. a3! …

After the game Tim thought this was a blunder. I thought so too. But after I destroyed one variation after another in our postmortem analysis, we came to realize that it has a subtle point. By playing 15. a3 here (or on the next move) White forces Black to trade a pair of pieces and thereby frees up the a1 rook for defense. If the bishop on b4 goes unchallenged, then Black simply doubles on the d-file and wins.

The reason this move appears to be a blunder is that it weakens the light squares on the queenside. But Black is not able to exploit this: When his queen comes to a4, White will play b3 and chase it away.

15. … Rd8+ 16. Bd2 Bxd2 17. Nxd2 Qe7

As noted above, 17. … Qa4+ 18. Kc1 h4 19. b3! really doesn’t get Black anywhere.

18. Rc1 …

Now the rook is available to help White defend. What should Black do next?

Black to move.

Here I saw that I could win the exchange with 18. … Qd7 19. Rc2 h4 20. Ne2 Be4. (By the way, I should interject here that Fritz finds a marvelous trap in case White should play 20. Be2 — a move I didn’t even consider. The answer is to allow White to go ahead and pin Black’s queen: 20. … hg! 21. Bg4 Qxg4 22. hg Bxg2 and White must give up the rook with 23. fg Bxh1, because otherwise Black will take on f2 and win. Wow! But this is an example of computer chess — I doubt that this would have happened in a human game.)

After 20. … Be4 21. Nd4 Bxc2+ 22. Kxc2 White has three pieces for the queen, and he can consolidate his position with Be2 and N2f3 (or, if necessary, N4f3). White has no weaknesses that Black can attack, aside from possibly the pawn on e5 (but that is balanced by Black’s weak pawns on b5 and h4). I did not think Black would have any winning chances. Nevertheless, the computer, which has been skeptical all along, thinks that this is Black’s best way to minimize his disadvantage. Naturally, that is not exactly the way I was thinking about the position.

Tim said the move he was worried about was 18. … b4, which looked like a killer in our postgame analysis. But Fritz calmly plays 19. Nf5! and it looks as if Black’s attack runs out of steam. What do you think? I really don’t know.

Finally, the move I played was 18. … Qxe5, which I thought was a useful move because it defends my weak points, centralizes the queen, and gets rid of an annoying pawn. But Tim said he was glad to see me do this. The game continued:

19. Rc3 h4 20. Ne2 Kb8 (A sad necessity. The advance … c5 does not have any force until Black unpins this pawn. But the tempo lost by this move allows White to consolidate his position.) 21. Nf4 Be4 22. Kc1 Qd6 23. Nd3 …

Black to move.

And now a catastrophe happened. I played 23. … c5??, forgetting that my bishop on e4 is now en prise. Of course Tim played 24. Nxe4, and I resigned a few moves later.

It’s a pity the game had to end this way. After the game we looked for a while at 23. … Bh7 without reaching any definite conclusion. It seems, though, as if White can consolidate his position pretty well with 24. Be2 followed by 25. Rd1. Although Black has a bit of a bind, it is hard to point to any definite threats. If Black can’t make anything happen, White’s material advantage of R + 2N versus Q is going to start making itself felt more and more strongly. The computer, of course, gives White a big 2-pawn advantage. While I wouldn’t go that far, I think that my attack has run out of steam to some extent. So perhaps I should have gone into the endgame of B + 2N versus Q when I had the chance.

In spite of the disastrous blunder at the end, I think that this game was a really good test for the 9. … Bb7 variation. I don’t see how I could have played the attack any better, and so I think we have to concede that Black’s superficially attractive move 12. … Ne3 really didn’t lead to any advantage. I’ll have to look at the alternatives to see if there was anything better.

And the position at move 18 definitely offers a moral: If you play chess in an “all or nothing” style, you may end up with nothing. There are some times when you may have to concede that your attack just isn’t working out the way that you thought it would. You may have to accept a smaller advantage than you expected, or you may even have to settle for equality or a slight disadvantage. However, this may be better than bull-headedly playing for your attack until the end. In retrospect, I probably should have settled for that slightly inferior queen versus three minor piece endgame. I just had to be objective.

Some tough lessons! But they were worth learning.

P.S. The plane landed, and I finished this post on my regular home computer.