Saturday, January 09, 2010

Finished level 30

I actually finished level 30 about a week ago. But getting back on track at work from the holidays has taken most of my time. First the results (then the commentary)


>-------------------<
| Level 10 100% |
>-------------------<
| Level 20 97% |
>-------------------<
| Level 30 91% |
>-------------------<



The last time I did CT-Art, it took me until my 4th repetition of level 30 to get from this high of a percentage. So that indicates that I have made some tactical progress. On the other hand, my 5th (and final) rep of level 30 I did at 97%. So I have slipped from that level. And of course, this time through was much slower than what I was doing before.

Some of the level 30 problems are really quite difficult and require seeing a lot of variations. The fact that I was able to do them so quickly in the past raises an old question: Does repetition just teach us the solutions to particular problems? This is an important question, because our real goal is to do better in new positions. At the moment, my answer the question is a resounding "I don't know."

But, I'm inclined to just keep working instead of trying to optimize the process. My results on the most recent run through level 30 are significantly better than they were the first 3 times I did these problems. Since that was 2 years ago, the effect of memorizing the solutions is probably not that great (and I did not feel while solving the problems that I just knew the solution, I could really calculate the problems better).

2 Comments:

At 8:18 AM, Blogger From the patzer said...

By doing the same set of problems over and over again i could do them in a small amount of time with a 100% score. Nothing impressive, i just had the solution of each exercise in my head because i had done then two to three times already.

Which is offcourse totally wrong since one must not memorize the solution but the characteristics of the position.

After recognizing the above fact i stopped doing repetitions since it didn't learn me anything.

 
At 1:29 PM, Blogger Blue Devil Knight said...

This is an interesting, perennial, debate about the Circles.

My first attempt at answering it was back here. I still largely agree with what I said.

It isn't clear at all that memorizing particular problems is bad, or that the brain works by remembering particular problems rather than forming gravitational wells for each problem that will suck in any nearby pattern in feature space. Indeed, in all cases of human memory studies so far, it seems the latter, so why would we expect it to be different in chess? (E.g., face recognition).

That said, I am not convinced the circles are all that great. And I'm not sure why. They certainly helped me a lot, brought me from complete patzer to someone whose mistakes didn't involve dropping major pieces every game.

Within the circles genre, I think it might be better to have a group of similar, but not identical problems, just to foster the construction of a wider gravitational well. And of course when our pattern recognition kicks in, we have to be careful in chess--one piece on a different square can mean the typical pattern doesn't work). Recognition of a "pattern" should usually be taken as a suggestion from our brain, not an edict. This is one sense in which Rowson is right to talk about forgetting what you know.

 

Post a Comment

<< Home