Collecting data
At various times in my chess progress I have felt as if I was at a wall. I was beating higher-rated players, yet sometimes losing to players that I should beat consistently.
I've taken a more analytical approach to discovering my weaknesses and changed a bit of my game. Training directly in your weakest aspect of your game really shores up your weak links.
The newest thing I am going to do is compile some statistics about the reasons I lost
tactically in 5 0 games. Many people say speed chess is 'bad' for your chess
however I do not think this has much foundation. Many strong players play
speed chess regularly and ICC is full of very strong players who play all day.
In my opinion, speed chess is just one aspect of improving your game. What speed chess
CAN help you in is forcing you to move as quickly as possible, and that carries
over directly to rated uscf games. Ideally, a 5 second increment would be on
all games, including speed games, in order to match tournament conditions, but we all know 5 0
is the most popular format
for speed games online.
I have a nice program called 'chess assistant 9' by convekta. It is a competitor
to the chessbase line of products and has some interesting features. The one
feature that I think is a killer is that you can have it analyze your ICC games
*while you are playing them* and then exactly when the game is finished
it will paste a bunch of analysis to your game board. This is amazing because it
means you can find out if your tactics were sound or what you should have done
instead of blunder. Of course it is not going to give you insightful positional
analysis in the 2 seconds it spent analyzing a move, or even longer, but tactically
it could do wonders for your game.
My plan is to play alot of 5 minute games on iCC (actually I do that now), set
the program to only annotate when I make a blunder or miss a nice tactical shot
that would change the program's evaluation by half a pawn or more.
Each time I lose a game because of a knight fork, a pin, a skewer, or something
more complicated, I will file it away under a 'theme'. For instance, the theme
of 'not looking where the guy just moved his knight and can now fork you with'
is a common one that I miss when I am rusty. I will then add up all the results
of the themes and see where my tactical game is most lacking. Also I will be
saving the position of each game so I can see exactly how it played out.
This may seem like alot of work for little reward, but if it identifies a mistake
that I make over and over again, that alone represents a big jump in actual strength
once I become aware of this mistake (and hopefully eliminate it).
Stay tuned for the results of this endeavor (i plan to upload some data tonight )
Update: My first game and I already learned alot using this method! The game
was a sicilian defense in which he made a mistake according to MCO and played
g6 on move 2 (after Nc3). I have played this many times and love the positions
I get out of this line after d4.
Then he played Qa5, and attacked my e5 pawn. d6 was to come and he would
get some intense pressure against e5, which could become worse after Nc6.
When I saw Qa5 I immediately thought defensively, and started deciding how
to defend my pawn on e5. Instead, as the program showed me immediately
after the game, I could have played Qd5, a nice counter-attack which totally
refutes his last move.
It threatens f7, threatens the queen on a5, and if black
trades queens, white's knight gets a nice position on d5 in which it is threatening
unstoppable forks on c7 and b6.
I'm not sure what to file this tactic under but I guess "forced trades" would be
a nice category, in order to indicate that I force a trade that he has to accept,
except the piece I recapture with gets a dominating position.
The game continued and again, I could have played Qd5. I never considered
this at all. To be true it was a 5 0 game, yet it's such an obvious tactic that
the blindness of it in my head is something I can learn from.
Here is the full game: