Originally I entitled this post "Event Horizon" because of the difficulties in calculating long variations. I’ve never seen the Sci-Fi/Slasher named
“Event Horizon” so I decided to change to a movie I had seen, Disney's 1979
The Black Hole. In the movie, there are two main robots: a good R2D2-like robot named Vincent and a 7-foot tall frankenstinian evil robot named Maximillian.
Here’s a diagram from my game. Black to move.
One of my favorite chess sayings is the Indian proverb “Chess is a sea in which a gnat can drink and an elephant can bathe.” Well, sometimes it’s a frickin’ black hole that sucks up all your creativity, ambition, energy, and time.
The reason why I’m so annoyed is that I thought I had this position 90% figured out, but it turns out I knew only 10% of what was happening. I was playing Black and I had just wrested the initiative from White. For this blog, I was all set to moralize on how White neglected his development, fell behind, lost his f2 pawn to a deep combination, and couldn’t find a safe place to put his king. But my beautiful combination was torn to shreds by an evil robot named Fritz.
This is what I thought in the above position: I’ve piled up my knight and rook on the soft spot of White’s position at f2. The only defender available is the queen and even though I’m a pawn down, I’d “buy” White’s queen and f2 pawn for my rook and knight, so I’m definitely going to capture on f2. Now which piece to capture with? If I take with the knight, White can legally castle short and take advantage of the fact that we both have open f-files. I might even be behind in development because of my undeveloped c8 bishop. What if I take with the rook? It’s pretty risky since the only thing guarding it is the g5 knight. If after Rxf2, Bxg5 is definitely out because of Rxd2. What about Rxf2, h3, removing the guard? That’s trouble, but oh wait, I have Qh4. Discovered check is a very strong threat. But then hxg4 threatens my queen which is supposed to deliver the discovered check, so I might have to use the double check to fight my way out. Let me first check out Qxh1+, then Kxf2 and I come out an exchange ahead with Qxa1. My queen is offside, but I should be able to get out with Bd7 and Rf8+. Going back to hxg4, does Rxe2 dbl. ch. work? Kxe2 is pretty much forced, but this connects the two rooks and I can’t capture Qxh1 any more. What about Bxg4+? This almost forces him into playing Kf1 cutting off the rooks’ mutual defense, but he has the resource of Nf3. Then I get exf3+, but White gets gxf3 and my bishop and queen are both attacked. I gave up calculating Bxf3+ Kxf3 Rf8+ because it was beyond my “event horizon”. 16…Rxf2 17.h3 Qh4 18.hxg4 Rxe2+ 19.Kxe2 Bxg4+ 20.Nf3 exf3+ 21.gxf3 Bxf3+ 22.Kxf3 Rf8+. Thirteen ply make my head hurt. At least, I have the other variation 16…Rxf2 17.h3 Qh4 18.hxg4 Qxh1+ 19.Kxf2 Qxa1 and I’m ahead in material. So I made the move Rxf2.
It turns out the variation with the double check 18…Rxe2+ is better because after, 22...Rf8+ follows 23.Kg2 Qg5+ 24.Kh2 Rf5! 25.Rhe1 Qh5+ 26.Kg1 Rg5+ 27.Kf1 Qf7+ 28.Qf2 Rf5 29.Qxf5 Qxf5. Black has more pawns and a safer king in the duel between the Black Queen and the White Rooks. And I only had to calculate 27 ply to get here! The variation I settled on at the end of the last paragraph is bad because of 20.Nb3 Qb1 21.Qd8+ and White gets at least a perpetual check. In the end, Rxf2 is a good move and White’s best answer is probably h3, drawing Black’s queen to h4. Then White castles long and uses the time Black takes to unwind his pieces to launch a counterattack.
I guess since Rxf2 ended up being sound, I can still moralize that White lost because he fell behind on development. Sometimes this is the advantage of playing Black: White plays too loose, overestimating his opening advantage. I have been guilty of doing the same thing and accentuating the mistake by opening up the position as well.
Why should I be complaining about my third win of 2008? Winning with luck and serendipity seems cheap. Winning like I know what I'm doing is where I have to be to get to the next level. Check out
Dana Mackenzie's analysis of IM Josh Friedel's win for an example of what I'm talking about. Interestingly enough, Dana also has a more recent post talking about
good and bad aspects of computer chess.