lichess.org
Donate

Calculation vs Pattern Recognition

@petri999 said in #30:
> Rely is wrong word. With more pattern/chunks/whatnot top player can zoom in into interesting moves. But they do no play guessing games. some sort concrete analysis is key verifying intuitive findings.
> en.wikipedia.org/wiki/Adriaan_de_Groot
> made the most famous study of subject. I think interesting finding that good player spent quite bit of time after choosing the move trying to prove it to be wrong.

That makes sense, intuition as generator, and logic as filter. That's how I think, even at math or modelling level...
Calculation being the laboratory of the intuition hypotheses...(could be viewed as intuition pruning, but I think not in my case, my intuition is generative, and often wrong, but less wrong with time i hope, as in too wide, and getting biased better upon experience, which could be view as pruning if looking at all legal candidates, yes, but intuition also looks at positions in some possible depths, skipping turns if necessary, etc....).

I will look at the link.

Note: i am interested in the psychology for the progressing newbie (which we are forever) ... I think a lot of psychology has been spent on the destination... but that is what some are talking about here....
It is natural that psychology would look at the contrast between naive subject and expert subject first.. professional musician, expest this or that... expert language practitioner (words can be counted and shared, makes for insights into thinking, and chunking that can't be enumerated).

but there is the connexionist view, more recent psychology points of view. Visual thinking that may not be of linguistic nature (although language is the bridge).

De groot, though having feet in both fields would have generated great insights. (and must have). probably what i think is known about chess psychology must have originated in his work.

From wikipedia link.

> De Groot concurred with Alfred Binet that visual memory and visual perception are important attributors and that problem-solving ability is of paramount importance. Memory is particularly important, according to de Groot (1965), in that there are no ‘new’ moves in chess, so those from personal experience (or from the experience of others) can be committed to memory.

To the learner though, there are new moves. And to think of it, isn't chess about an infinite source of new positions, where a "move" is by definition new (the move is not the piece changing square, but the resulting successor position).
i meant if positions can be found different even for experts having been at the top for long enough ("infinite" source), then these constitute new moves... The question then is about pattern generalization: which pattern of many or whatever internally digested objects is likely to apply in planning/decision making from current new position. I don't think linguistics helpful enumeration properties, are the complete story. From the little semantic scholar strolling I did, not exhaustive, I was under the impression that the estimations of how many patterns (and I guess the notion of pattern itself) was transposed from known linguistic cognitive psychology then, based on how many words a native or expert practitioner of a given language could hold (and possibly judiciously use in the correct text context?). I may be wrong, and might need to dig further.

There might be finite human capacity for patterns, but how they relate to the new position experience (yes i am assuming some candor left in even the most expert chess players left) might be some that needs also some adjustment and hence part of the learning psychology we might be concerned with. But I won't insist about what expert do, I only need to suggest my point for not yet expert players.... Also, maybe that cited paragraph is tangential to what De Groot brought forward, or out of context, which i would not know.

This topic has been archived and can no longer be replied to.