In Sunday's Washington Post, a group of historians tried to predict what history will ultimately say about George W. Bush's presidency. One said that he is the worst president, ever; a second agreed that he was pretty bad, but still might redeem himself in his last two years; and another said that only time will tell, noting that our views of presidents often change with the perspective of time.

Historians have been playing this game for many years. It makes them feel relevant. However, the methodology of such efforts never gets above that of a simple popularity poll. A historian will survey a group of his friends, and they are asked to rank the presidents on whether they are great, near-great, average, below average or failures.

Obviously, this method is fraught with problems. For one thing, the historians chosen to participate are not picked randomly and therefore are not necessarily representative of all historians. Also, they have different specialties and may know a lot about some presidents but very little about others. The historians are overwhelmingly based at elite universities and thus tend to be much more liberal politically than the average American. And of course, they are well aware of previous rankings and seldom deviate from them except marginally.

The biggest problem I have always had with these presidential rankings, however, is that no one ever appears to use objective, measurable criteria for placing a president high or low on the list. The main criterion seems to be activity -- doing a lot while in office. This creates a strong bias in favor of presidents who served during times of crisis and against those who served during times of peace and prosperity.

To my mind, this sometimes gets the whole ranking system upside down. This is especially so when one considers that occasionally the crises that presidents have had to deal with were in fact their own fault. In effect, those who did their jobs well and avoided unnecessary wars, recessions or other avoidable woes get punished, while the screw-ups are sometimes rewarded for fixing their own mistakes.

Thus Calvin Coolidge almost always ranks low in the presidential popularity polls because he didn't do much of anything in office. But there wasn't much that needed doing. He kept the nation out of war, maintained prosperity and was not tempted to undertake a lot of unneeded "reforms" just to keep busy and raise his popularity rating among future historians. For my money, this makes Coolidge among our best presidents, not one of the worst.

At the other end of the scale, Franklin Roosevelt nearly always ranks high on the list because he did a lot of stuff and coped with major crises. But he caused some of the problems he is credited with fixing. In the view of economists, as opposed to historians, Roosevelt's economic policies mostly deepened and prolonged the Great Depression. Yet he gets credit for ending it simply because he stayed in office long enough for the depression to end on its own. If Roosevelt had left office after two terms, like every other president, perhaps Wendell Wilkie would instead be considered among our great presidents.

In other cases, presidents seem to benefit mainly from things they did outside of office. For example, Thomas Jefferson always ranks high on the list. But he really wasn't an outstanding president. His greatest accomplishment, writing the Declaration of Independence, took place a quarter of a century before he became president. Tellingly, Jefferson himself did not list his presidency as among his three greatest accomplishments.

I have always suspected that Woodrow Wilson benefits undeservedly from having been a professor of history at Princeton before becoming president. Historians are naturally biased in favor of one of their own. John F. Kennedy gets a similar boost from having employed one of the nation's best-known historians, Arthur Schlesinger Jr., as a close adviser.

I suggest that an objective criterion for future presidential rankings ought to be how many people their policies killed unnecessarily. On this basis, Wilson would be among the worst because, in my opinion, America had no vital interests at stake in World War I and never should have become involved in it. And Harry Truman probably didn't need to drop atomic bombs on both Hiroshima and Nagasaki.

To those who think this is a better way of ranking our presidents, one place to start is by going to this Webpage: www.opencrs.com/document/RL32492. There, you can download a document produced by the Congressional Research Service titled, "American War and Military Operations Casualties: Lists and Statistics." It reports the number of American military casualties from every war in history except the current one, which changes daily. Depending on how legitimate you believe a war was, you can do your own rankings of the presidents.