Against an opponent who plays perfectly, your strategy may be just as good as the one I suggest. When the game turns, a perfect opponent will be sure to get his cube in while you are still taking. He will, that is, unless he goes from "not good enough" to "double-pass" in one roll.

Against real opponents, there is also the possibility of a cube error. By deferring your double (while being careful not to lose your market), you will still win with the cube at 4. When you lose, however, you may find that you are losing only 2 points when the game turns, and your opponent doubles too late.

It is also possible that a real opponent will drop a cube he should be taking, and thereby cede equity to you (even though his pass would mean that you only win 2 points). When you double early, this sort of error is highly unlikely.

Incidentally, the strategy I describe is used by some expert and world class players at 2-away, 2-away.

Come to think of it, I am confused, too! I guess I was thinking about the 2-away, 2-away where both players have access to the cube.

I can't really help you here. Volatility and market losers mean that every position must be evaluated on its own. In a smoothly changing, non-volatile situation, I would be aiming for 72% to 73%.