Sunday, January 8, 2012

The Karate Kid

When he made Daniel Russo wax his cars, paint his fence, paint his house, and sand his floors, Mr. Miyagi was secretly training him, not to be a slave, but to have strength and muscle memory to be able to block all kinds of attacks. This week, I overheard a martial arts instructor talk about the stages of training the other day. She highlighted these four:
  • Unconscious Incompetence - You have no idea what you're doing right or wrong.
  • Conscious Incompetence - You know what you're doing wrong but can't fix it.
  • Conscious Competence - You are able to do things right if you think about it.
  • Unconscious Competence - You do things right without thinking about it.

A lot chess training is pattern recognition and I agree it is important. If you know the beginning landmarks of all the checkmate patterns, then your analysis tree doesn't have to re-invent those checkmates every time you see them, especially when those set-ups show up at the end of your own analysis horizons. Wheels need not be reinvented every time. But when you get to the state of unconscious competence, it's hard to trust that stranger upstairs when you don't even know his name or party affiliation.

Perhaps this is a totally different part of chess competence, but when I analyzed a postmortem with future world champion Steven Zierk, I was dumbfounded at how quickly he could snap through variations to get to the truth about a line. I have difficulty buying into the oft-held forth de Groot assertion that around the expert-master level, the depth of analysis is not different, but the intuition and experience of evaluating different positions is where the difference lies. Arthur C. Clarke said "Any sufficiently advanced technology is indistinguishable from magic." Well, Zierk was so advanced, I thought I was watching wizardry. He seemed to be able to find a 10-ply refutation with ease. That kind of concrete ability I think is at least as valuable as a 9-ply calculating ability with a 1-ply correct intuition.

I listened to the entirety of Malcolm Gladwell's audiobook Blink. Here's an article about intuition pertaining to emergency room physicians that briefly mentions de Groot and some research gender-typing chickens. They used the term "sexing chickens", but I prefer gender-typing as a safer term. :)

Maybe I'm stubborn and mistrustful of the non-concrete. Or maybe de Groot's analysis doesn't even apply if I'm a Class A player masquerading as an Expert. Until I'm able to correctly verbalize or at least demonstrate deep-ply understanding of a complex tactic or position, I won't be focusing on my intuition to get me to the truth. I guess I'll be stuck at training stage 3 for a while trying for conscious depth and accuracy.


frenez said...

there's a real difference between masters and experts/a players. i think experts and a's are analyzing the same position in the same way, doing 1 move at a time, saying, if i go there, he goes there, then i go there and i'm good or that's bad etc.

the master looks at the same position, not move by move, but as a whole, like reading a clock and 'instantly' 'sees' the same position. they don't get there move by move the way we do, they just intuitively feel it, see it, whatever you call it, it's wizardry.

unfortunately, you can't train for wizardry, you have to already be a wizard!

Soapstone said...

Hi frenez! That's an interesting idea about the way masters do analysis. Is that observation, theory, or proven fact? I know that de Groot and others talk about how masters are good at "chunking" and have improved board memorization skills. Perhaps I have been too focused on individual piece moves and can't see the whole board in my ply. Losing the forest for the trees so to speak. A long time ago I was thinking about encoding chess notation (like PGN) for a database. I could store the entire position in each database line or just have a piece move from a given point. In image processing they might talk of rasters versus vectors. Maybe masters think like rasters, while the rest of us think like vectors. As I often lament the small, but perceptible slow-downs of my mind, I'm inclined to agree with you about not training for wizardry. But maybe next time I do tactics problems, I'll try to see the whole board more often and try to rasterize my board vision. Maybe solve them blindfold after I've committed them to memory.

frenez said...

that's completely observation, both from analyzing with them and listening to their commentary. it seems they see the end of the analysis 1st and then fill in the moves later.

the master will make a comment about a variation and you're (me) head is spinning trying to figure it out and understand why it's losing or winning. then, you have to press the master to explain it and you can see their brains backing up to provide the moves ... if that makes any sense at all!