[mnzboovz] Developing opening theory

Optimistically, one can describe the development of chess opening theory as a vast distributed computing project spanning centuries, seeking ground truth.  Computers can continue this computation, and they do it well.

Pessimistically, and realistically, theory develops in the direction of offering the best "chances" against an opponent unfamiliar with the position, while also minimizing risks if the opponent is familiar with the position.  It is a psychological struggle about the familiar and unfamiliar, human fallibility, a risk-benefit analysis with probabilities.  This is very different from how computers play the game, at least currently.

The original idea was, chess at the highest level, with paid competitors and widely published and studied games, should pursue the noble optimistic goal above, advancing the state of knowledge about ground truth.  The game at the highest level should be modified to best further the search for truth: probably incorporate computers and databases into at least the opening phase of the game, e.g., Advanced Chess.

However, chess at the highest level (and every other level) is actually mostly played according to the pessimistic interpretation above: how can I trick my opponent, for one game only?  I mostly cannot see how that can be changed, except maybe for matches.

Incidentally, correspondence chess, despite computers, is also closer to the pessimistic interpretation: how can I avoid the weaknesses in my chess engines and exploit the weaknesses in my opponent's engines?

Actual ground truth in chess is probably that it is a draw.