>A great deal of psychological data can be gleaned from outside observation, so it would not be wise for top political figure like President Putin, President Bush, Condi Rise, or Henry Kissinger, for example, to play chess in public, where they could be observed by potential political foes. Psychologists, and other think-tank types, can learn things like whether the individual prefers playing offense or defense, the degree of risk the individual is willing to take, the individual's aggression level, how easily the individual can be made to modify his strategy when confronted with a threat, how patient the individual is in laying out his plans, and even IQ, just from watching a person play a few games of chess.

These can be faked, with or without coaching. I was once tested for a quiz (back in the 80s when it wasn't all trivia), and one battery of questions was an obvious personality assessment. I just got myself into a mood of a person I imagined would like to have fun with them, or on their account, and started answering as that person. Yeah, I got through :).

Or the classical "watch what he's drawing on the margins" - knowing that pretty much everyone has read something of the kind, I've deliberately drawn scaffolds, nooses, knives and skulls while sitting at the staff meetings in the school where I was teaching.

IOW, small sample is a small sample - which still doesn't invalidate the analysis of GW2B. We got a decent sample of him.