Pattern-recognition, not data-crunching (or why my mother is as accomplished as Kasparov)

Practically everyone thinks their mother is the best cook in the world. Let me assert that all of them are wrong, because my mother is clearly the best!

The distinctive feature of my mother’s cooking is her inventiveness. She can literally concoct new dishes with a dizzying array of ingredients across different cuisines. Succulent Indian vegetables & spices would add just the right amount of tanginess to a continental dish; a mélange of Mexican ingredients would find themselves in the midst of Indian roti-curry. And these disparate ingredients would all somehow come together to make culinary magic!

Truly good cooking then is not so much about knowing a wide array of recipes; it is really about knowing how ingredients work together. It is an art of being able to work with rules or ‘good moves’. This article (http://www.slate.com/id/2219243/pagenum/all) I read recently also talks about the same- master chefs are not storehouses of a zillion recipes, they are artists who can compose something new by creative application of heuristics that they pick up over the course of time. Cooking is hence, a pattern-recognition and not a data-crunching problem.

I remember reading an article about chess players that pointed to a similar thinking process. Grandmasters and professional chess players are expert pattern recognizers, not people with either elephantine memories or faster computational ability.

2-3 observations indicated this: Chess pros were no better than laymen or amateurs in rearranging chess pieces on a board that they were shown, if the pieces were randomly arranged (making it an ‘illegal’ position in chess). This showed that superior memory is not the cause for the difference in chess ability.

Similarly, self-reporting by chess pros showed that they had not evaluated more moves than amateurs before deciding their move. So it wasn’t even faster data crunching that explained the difference.

The true difference was in the pattern recognition- chess pros evaluated less ‘bad’ moves than amateurs. It is almost as if they had developed (or were born with) a filter that left out bad moves, just as amateurs would not evaluate illegal moves.

It is this knowledge of heuristics that distinguishes good chess players- they probably would be thinking stuff like “never sacrifice a pawn after castling if you’ve lost a knight”, “use a rook to check, with a knight threatening the opponent’s queen” and so on, just as a great cook must be thinking “Use cardamom with sautéed vegetables only if either tofu or mushroom is part of the salad” etc.

In fact, this would likely be the case with all art- it is the application of a wide range of ‘rules’ in varied permutations. We probably would not be able to identify all the rules, neither would a true artist be able to articulate them. But subconsciously, they are working with patterns, and therein lies their wizardry