You create them. You get a certain number of people who have never played go. You give them a rule book, teach them go problems that get them to a certain level of problem solving competence, and then you use different problems with similar themes, or slightly more complexe (3 move sequences, rather than 2 move answers) sequences to solve in the study itself.

But, I shouldn't be here. I already hijacked the thread once apparently.

CSam

CSam: you might be able to get decent results (though you're talking about a huge and expensive experiment) but there'd be a problem with ecological validity. Normal go players will have different motivational sets and talents than people selected at random, or via any procedure other than looking at people who are motivated to play go. Facts about how they learn may not match how go players learn.

_________________
Occupy Babel!

No study is perfect. I think the self-selection-in-study-participants dilemma is practically (literally?) a field of study in itself.

Here, I don't think it'd be a show-stopper. The study would still provide interesting information.

here is what an actual pro had to say about this (just before becoming a pro)

Jochen: Do you sometimes study the "Lessons in the Fundamentals of Go?"

Hans: Sure. Also, I do two hours of tsume go every day. No difficult problems, more easy ones, but a large number of them. I put them into a file and I use them to practice recognizing a shape at first glance without thinking much, which means to internalize the shape. Of course I do some difficult ones too.


he was surely not a kyu player at the time . i'm just posting this as it might force us to think twice about how strong players think/train, how their brains work.

the whole interview is here (and very interesting imho)

http://361points.com/hans/

hope this helps