e, filling
up the tape in his minicorder with the same old answers that others had
gotten.

He left, giving the general a brisk salute and turning before the
general had time to return it.

Back in his office, he filed the tape dutifully and started on Item Two
of the duty list: _Strategy Analysis of Battle Reports_.

Strategy analysis always irritated and upset him. He knew that if he'd
just go about it in the approved way, there would be no
irritation--only boredom. But he was constitutionally incapable of
working that way. In spite of himself, he always played a little game
with himself and with the General Strategy Computer.

The only battle of significance in the past week had been the defense
of an Earth outpost called Bennington IV. Theoretically, MacMaine was
supposed to check over the entire report, find out where the losing
side had erred, and feed correctional information into the Computer.
But he couldn't resist stopping after he had read the first section:
_Information Known to Earth Commander at Moment of Initial Contact_.

Then he would stop and consider how he, personally, would have handled
the situation if he had been the Earth commander. So many ships in
such-and-such places. Enemy fleet approaching at such-and-such
velocities. Battle array of enemy thus-and-so.

Now what?

MacMaine thought over the information on the defense of Bennington IV
and devised a battle plan. There was a weak point in the enemy's
attack, but it was rather obvious. MacMaine searched until he found
another weak point, much less obvious than the first. He knew it would
be there. It was.

Then he proceeded to ignore both weak points and concentrate on what he
would do if he were the enemy commander. The weak points were traps;
the computer could see them and avoid them. Which was just exactly what
was wrong with the computer's logic. In avoiding the traps, it also
avoided the best way to hit the enemy. A weak point _is_ weak, no
matter how well it may be booby-trapped. In baiting a rat trap, you
have to use real cheese because an imitation won't work.

_Of course_, MacMaine thought to himself, _you can always poison the
cheese, but let's not carry the analogy too far._

All right, then. How to hit the traps?

* * * * *

It took him half an hour to devise a completely wacky and unorthodox
way of hitting the holes in the enemy advance. He checked the time
carefully, because