Thursday, January 17, 2008

Back with New Year's Resolutions

I took a bit of a break from chess and blogging after Christmas, but I'm starting to get back into it. If you sent email in the last couple weeks, my apologies for the delay in responding.

I'm going to start the year with stating a few of my chess resolutions for the coming year, which will hopefully help keep me "on the wagon".

Things to do:

  • Play slow games against Fritz to practice consistency
  • Resume playing through master games. I'm thinking I should set a goal of at least one per day
  • Work on tactics, especially tougher problems. Key references: Encyclopaedia of Chess Middlegames, Anthology of Chess Middlegames, Van Perlo's Endgame Tactics, Tal's Winning Chess Combinations, and CT-ART
  • When tempted to play a few quick blitz games, substitute the above tactical exercises
  • Work my way through Muller's Rook Endgame DVD, entering all the examples into ChessBase and playing the positions versus the computer
  • Start analyzing my blitz games without a chess engine, and then check my analysis with a chess engine.
  • Play more rated tournaments

Things not to do:

  • Late-night ICC blitz benders

Now I need to get off my butt and get some instructive posts up. Thanks for your patience.


happyhippo said...

This sounds like an excellent plan you have there although I'm not sure of the benefits of analysing blitz games without chess engines.

Best of luck with your training program!

Grandpatzer said...

The idea is that you use your own brain to analyze the game as best you can, and then to check your analysis for accuracy. Compare this technique with, for example, that of Silman (playing through annotated master games move by move,without looking at the annotations first, trying to come up with the best candidate move, and then checking the annotations).

Some errors are clear oversights, which you'll spot in your analysis. Even in cases of blunders, however, it's not always clear what the best moves are, and there's still a need for analysis. In other cases, you have a situation where there are several candidates to consider but one is best. Here, using the computer to find the answer and then trying to rationalize the computer's choice is less effective, in my opinion, than trying to figure out the answer first and then comparing your analysis with the computer's.

Finally, passing the game through such a "double filter" separates the inaccuracies where you knew better from the inaccuracies that demonstrate your level of weakness. If you didn't spot an error with unlimited time at your disposal, and the computer immediately sees it, you should spend time on that position and work at trying to discover the "truth" for yourself. I say "truth" because you should always take computer evaluations with a grain of salt.

transformation said...

exactly! thank you.