In the poker game Hold ‘Em, the best cards a player can start with are two aces. The game is structured so that a player eventually makes his best hand out of two initial private cards and five subsequent community cards that everyone can see. Betting takes place in four rounds, with increasing levels of information becoming available as more community cards appear on the table.

A good player who begins with two aces will place whatever initial bet he believes is most likely to maximize the amount of money other players bet, since he is certain he is in the strongest starting position. But as the hand plays out things become more complicated. As the community cards appear, they may do anything from making it clear to a player that he can’t lose, to saddling him what he now knows is almost certainly a losing hand. A particularly interesting aspect of the game is that a hand that looked nearly unbeatable can suddenly become an almost certain loser when the final card is dealt.

In this circumstance, a good player almost always folds. The fact that he (quite correctly) bet heavily earlier in the hand, and that therefore a lot of what was his money is now in the pot, is irrelevant to the present situation. The chips he invested earlier in the hand are what economists call “sunk costs:” that money has been spent, and trying to get it back now is just throwing good money after bad. (Things aren’t quite this simple _ nothing in poker is simple _ because a smart player will occasionally bluff, by making a heavy bet while holding a weak hand in an attempt to get other players to fold. But a player who bluffs a lot against competent opponents will soon be broke.)

It’s often noted that politicians hate to admit mistakes. In particular, presidents seem to find it almost impossible to do so, perhaps because admitting error seems to undermine the legitimacy of the enormous power they wield. But an even more difficult thing for a president to do is to argue that a decision which was correct at the time should now be reversed. This is likely to get him labeled a “flip-flopper,” as opposed to someone who changes his mind as more cards appear on the table.

For example, it’s still a plausible argument that, given the information available at the time, invading Iraq with the aim of overthrowing Saddam Hussein and turning it into a stable democracy was the right thing to do, for both moral and strategic reasons. Opponents of the war who deny this are claiming an omniscience no one has.

On the other hand, supporters of the war indulge in a dangerous fallacy when they argue that because invading Iraq was the right decision at the time, we must now “stay the course.” This is the equivalent of the poker player who, after playing an apparently strong hand correctly all along, refuses to fold when the final card ruins his hand, on the basis of the logic that his earlier bets would then have been in vain.

It may well be that going to war to overthrow Hussein was the right decision at the time, but that continuing to shove chips into the pot in the hope that Iraq will eventually _ five, ten, twenty years from now? _ become a stable democracy is an example of what poker players call going “on tilt.”

A player goes on tilt when he plays a strong hand correctly, loses anyway, then starts making bad bets because he lets his emotions get the better of him. Hawks and doves should consider the possibility that this describes the Bush administration’s current Iraq policy. But since that would require doves to admit they might have been wrong about the decision to invade Iraq, and hawks to admit they might be wrong about the decision to stay, I wouldn’t bet on it happening.

(Paul Campos is a law professor at the University of Colorado and can be reached at Paul.Campos(at)