The traditional way to evaluate a position is to determine whether it favors White or Black — that is, who has the better chances of winning, provided both players make the best moves.

Computers have refined this to a high degree and can tell you, for example, whether White is ahead by the equivalent of 0.39 of a pawn in terms of various esoteric positional factors.

But masters often ask a more practical question: Who has an easier time finding good moves?

In this week’s game from the recent Olympiad team tournament, Black appears to have a solid game, with an extra pawn and excellent winning chances after he tries to drive the enemy rook back with 28 . . . Nb8.

Machines tend to dismiss White’s reply, the sacrifice 29 Rxg6!, because its consequences can’t be foreseen. But White’s position is much easier to play after that.

In a game between two computers, Black was the one to bet on — beginning at least by move 20. But when it’s up to humans to find exact replies, the attacker is usually the heavy favorite, and Alexey Shirov (White) finished off with another sacrifice, 36 Rc6+!, after Black’s went astray in defense.