Analysis I

by zenquaker

So, one of the things everyone tells you to do when you’re studying Chess is to analyze your games. Or as I like to say: The unexamined Chess game is not worth playing. A Turkish player on one of the online asked how do you do that? The obvious solution is to plug it into a computer program and see what it says, but people also generally tell you not to use a computer to analyze your games. I thought, if that’s the negative response, what’s the positive one? What do strong Chess players say you should do to analyze your Chess games? So I did some research.

I did a web search and skimmed through the results, and came up with three articles on how to analyze your Chess games: one by Natalie Pagonina, one by “Roman” (Chessblogger), and one by Steve Lopez. I distilled all of this into my ten point plan for analyzing my games.

  1. Analyze all my games. Well, not all of them, I don’t have time. But generally Chess players say you learn more from your losses that your wins. I think there is much to learn from other games. You could have drawn a won game. You could have won a game because you had a great position, but you don’t know how you got there.
  2. During the game, mark the time point of each move. The moves you take a lot of time on are critical points worth analyzing. As I found out, sometimes so are the ones you don’t take any time on.
  3. Right after the game is over, write down what you were thinking during the game. You may not remember what you were thinking when you finally get around to analyzing things. If you don’t remember what you were thinking you can’t correct errors in your though process.
  4. Later on, scan the game without a computer. Move by move try to find tactical refutations, positional errors, and incorrect plans. Annotate the game again, perhaps with a different color.
  5. Now run it through a chess engine overnight, searching for moves better by 30 to 100 centipawns, for both sides (apparently 30 centipawns because unnamed grandmasters think the tempo is worth 1/3 of a pawn). Look for mistakes both in what you played and in what you analyzed in step 4. Make sure every analysis you bring in from an engine has verbal analysis attached to it, not just variations.
  6. Identify critical points in the game. These can be from the timing in step 2 or from the computer evaluation in step 5. They can even be from the notes in step 3, if you noted a change in attitude that affected your play. When looking at the computer evaluation, see where the evaluation of the position stopped hovering around even and went to one player or the other.
  7. Analyze the opening. Evaluate the position after the opening and decide whether or not your opening plan needs repair.
  8. Give a summary of the game. Why did it end the way it did? Which phase of the game was it decided in? What general problems did you have in the game?
  9. Go back and review the principles associated with the critical points in the game and the reasons for the game’s result. Figuring out what the problems were does you no good if you can’t fix the problems.
  10. Over time (once you have 20 to 25 games analyzed) look for trends in your games by rereading the summaries from step 8. If there are problems that crop up more than once or twice, make sure to work on them during study time.

Yesterday I analyzed my first game using this method, one I played against Shredder on Saturday. It was an incredibly useful process. For once I got a really clear idea of what my problem was in the game. Here’s the summary from that game:

I lost because I screwed up in an opening I didn’t really know. I could have recovered it, but I made a number of other errors. I rushed through plans I had made without stopping to consider after each move, I failed to consider his tactical opportunities, and I miscalculated on one of mine, I knew I had king safety issue but failed to take that into consideration when making moves, and I failed to look for basic game features I learned from Nimsowitsch and Silman.

Obviously the first issue is the opening, so today I went and tried to study up on that opening. Unfortunately there aren’t any good web resources for the King’s Indian Attack, perhaps because it’s not used a lot at the higher levels these days. The rest of the problems are really thought process problems. I’m studying a lot about Chess, but none of that is going to do me any good unless I can apply it to my games. I need a thought process: a plan as to how I am going to think after each move. I have some ideas in this direction, but I want to do some research. Perhaps I can find some articles as useful as the ones I found about analyzing games.