I played an Internet blitz game yesterday where, as White, I used my favorite Torre Attack and I was advancing against my opponent’s K while he was expanding on the Q-side…normal procedure in the line we played. Right when it looked like my attack was going to crash through my opponent found a defensive move that brought the attack to a grinding halt. That’s when I realized, not only was I probably going to lose, but I had only a few seconds left. With no time to think I just threw some spite checks at him and that’s when he blundered giving me a winning attack…which I failed to see. The remaining moves were played at lightning speed by both of us and after Black played 41…Qb1 to pick off the B, I moused my N to play 42.Ng6+ just as time expired.
Out of curiosity I analyzed the game with Houdini 1.5 and was very surprised at the results! The position was more complicated that I thought but what amazed me was Houdini’s analysis. It analyzed one variation from move 44 to move 145. I don’t know how accurate all that is but it’s a pretty impressive feat.
Just as an aside, in view of the recent posts about engine use, I have to ask what if a titled player reached the position after Black’s 44…Rxc2 where it took Houdini a hundred moves to figure things out? Of course one would never rely on that string of analysis, so it would take even more investigating. No wonder these world championship CC tournaments can last 10 years.
It also leads you to believe that when analyzing with an engine there’s more to letting it think for a few seconds and then trusting its output.
No comments:
Post a Comment