In just the last two days, two different rates of terrorist recidivism have been reported — 7% and 14%. These are only the latest in an ongoing, ever-changing, never-twice-the-same stream of recidivism rates since the U.S. began detaining so-called enemy combatants after 9/11.

Statistically, there should be an average number somewhere that holds relatively steady, but apparently no one can come up with one. At least, not one that is politically convenient. The rate keeps changing, depending on whom you ask, and when, and in what context. Basically, then, the much-discussed recidivism rate is meaningless.

How would such a number be calculated in the first place? Let’s assume for the moment that all detainees at Guantánamo are terrorists. If you want to know the rate of recidivism among them, you first have to release them. Then you have to have one or more people follow them, tracking their every move from that day forward, to see if they return to some form of terrorist activity. Huge expenditure of manpower, and a total waste if the individuals don’t revert to terrorism. Still, you have to keep track of them if you’re going to compile accurate numbers.

Or, in the alternative, you release them and if they get captured again while participating in terrorist activities, you count them as recidivist. That’s all well and good, but it doesn’t tell you what became of all the others you released. You don’t know if they’ve returned to terrorism or to the family farm.

So how do you determine a recidivsim rate? Pick a number. Any number. Throw it against the wall and see if it sticks. Your result will be at least as valid as any we’ve seen so far.