|
|
| From | Message | ||
|---|---|---|---|
|
 Ideas on how to analyze Chess gamesThe ideas mentioned here are of course not my own but have been derived from the writings of Lev Polugaevsky and Jonathan Rowson.....okay here goes... 1) Play over the game quickly, in 15 or 20 minutes, "to awaken your thoughts." 2) Play it over in the course of an hour to "synopsize” the key moments, (trends, turning points, emotional shifts). 3) Analyze the key moments in detail. a) Pending counterplay…you see your opponent’s idea even though it hasn’t happened yet. b) The prevailing trend seems to have stopped…there is no obvious way to continue the trend or increase the advantage naturally. c) An adundance of choice…you have lots of reasonable moves, but none that seem to be outstanding. d) Your opponent’s last move was somehow unusual or peculiar…there is no clear way to define ‘unusual’ in this context of course, but it will normally be a move that defies convention in some way. 4) Analyze the opening…investigate the roads not taken where the game could have gone down a different path altering pawn structures. 5) Play the game over once more and write annotations. |
||
|
wschmidt 16-Jun-13, 07:43 |
I think these are all good ideas and So, I'm trying out a new "surgical strike" approach to game analysis. Here are the steps: 1. Have Fritz analyze it, two minutes a move, on a setting that shows only the move score, not the recommended move or variations. This is often an overnight assignment for Fritz. 2. Play over the game quickly. In my case, this is 5 minutes or so. 3. Analyze the opening. I pretty much ignore Fritz at this point and rely on my books and database info. If it's an OTB game, chances are good I've varied from my repertoire although it may not be a blunder. If it's a GK game the first variation is usually an out-of-book move by my opponent. I spend some time on getting a sense of whether I could have improved play at this point. 4. Go over the game more slowly, keeping an eye on the score Fritz has assigned to the moves. When I reach a move where the score drops by about half a pawn or more, I pause and ask myself what a better move might be. I do this for all of my weaker moves and, if I'm interested and have the time and energy, for my opponents move as well. Only after I've come up with my candidate moves do I have Fritz re-analyze the position at that point and tell me what it's thinking. While going through the game, if I reach a point where something interests me, I'll analyze it even if the score hasn't fluctuated enough to trigger it. On the flip-side, to save time and frustration, if the score is so badly against me that I'm clearly losing, I don't spend time trying to figure out why my move was -6.50 and Fritz saw one that was -4.65. The key to this step is to do the work before seeing the Fritz recommendation. And to be flexible. Depending on my mood and the time available, sometimes my cut off level for analysis will be .30, other times an entire pawn. And sometimes it varies within a single game. 5. After completing Step 4, I usually play over the game quickly once again, revisiting the turning points. if it's an especially painful loss, I let myself skip this step. 6. Go on to something else. I realize this is pretty wordy, but in practice it's proved to be pretty streamlined. It seems to make efficient use of Fritz and, because I don't have the recommended variations in front of me right away, forces me to think about the key positions on my own first. I'd be interested in hearing any recommendations for refinement. |
||
|
bigpeta 16-Jun-13, 12:55 |
link to my ideas on this subject gameknot.com |
