Originally posted by tomtom232
I don't know but I think you play so many weaker players in blitz, not any fault of yours, that you have become comfy on the attack and thus very uncomfy on the defense. I think you overreacted to the threats and actually the a5 move that the comp shows seems perfectly rational and not like a box move at all.
First, white doesn't want to trade pieces an ...[text shortened]... point though... Good sacrifices are probably overlooked quite often in the games of amateurs.
I agree, that I tend to make more mistakes on the defensive. It seems like I can work out the right attacks, but on defense, I see all these attacking ideas and get lost choosing the correct path. Some of it is those darn ghosts.
That a5 move is still sort of a baffler to me. I saw some lines where black exploits a pin on the a1 rook, but still, it's hard to believe. Everything seems to indicate that the two pins will be super strong. White can take on b7 or g7 and leave the king with nowhere to go. The center is wide open and my chess "upbringing" tells me the king must get out of the center as soon as possible. Be7, Rg8, and c5 seems to be the best line for black, but again, black must play very accurately.
Even in some of the best lines I looked at, black still only held an advantage from slight to around a pawn. That's really something, to be a whole piece down without a concrete win, and still only be down a pawn or less! I even looked at a few endings, where white had a c pawn, rooks, and a queen versus a knight, rooks, and a queen. The evaluation was still only a pawn and a half! Another variation, showed white pinning and winning the d7 knight, but back rank mate tricks prevented white from every making the capture! It's a lively struggle even in the simpler looking positions.
I hate to repeat myself, but I wonder how many variations like this we never see.
I bet a computer sees lines like these and rejects them for something slightly better all the time. I wish a computer could actually be programmed to look for equal to slight advantage/disadvantage piece sacrifices as opposed to always finding the absolute best move.
Heck, I'd settle for an opening book with about 20 or so piece sacrifices like this, to catch the unwary. Until that happens, I'll just have to settle for the games of Mikhail Tal or Greenpawn.