Move Selection

[RobertJasiek]

how are you picking interesting moves to examine?

Simply speaking:

1) Positional analysis. This includes functional partition(s) of the board position, territory, influence, options, other strategic concepts etc. I support each analysis question by reading if necessary, e.g., quiescience reading for determining the territory count.

2) Aims. In the context of (1), I set possibly interesting aims. In particular, for each local part in the partition(s) of the board position, I consider aims.

3) For each local part together with an associated aim, I do a Local Move Selection to find the best move(s) (or sequence(s)) in the local part fitting the aim.

4) Global strategic planning. This brings (1)..(3) together. I do a sort of global reading while concentrating the previously determined moves / sequences and joining them while following consistent aims. In particular, a first move candidate appears.

5) I verify (4) with more tactical reading.

[hyperpape]

Other than the formality of your description, I see nothing that I would not assume any good player does. Does that seem inaccurate to you?

P.S. I do not think that thread splitting was necessary in this case.

[RobertJasiek]

Other approaches or variations of my approach to move selection can be imagined, so I would not be sure that every good player uses about my description. (E.g., one can imagine a somewhat strict strategy at the top - tactics at the bottom approach or vice versa.) Different good players do positional analysis with different frequency or detail.

It might, however, be possible to say that all "good" players do all of the following: analysis, aims, strategy, tactics, comparison, verification. Some players might do some of these only every few moves, other players might be aware of reasonably accurate analysis all the time. In games with sufficient thinking time, I prefer an all the time approach and use incremental updates to achieve this, but recalculate everything every couple of moves of the game.