HERE'S A PROPOSITION for anyone looking to make a little easy money between now and 1988: If someone is offering odds, bet against Gary Hart or whoever it is that the next Gallup Poll shows to be the Democratic front- runner. The odds -- if recent history is any guide -- argue strongly that whomever the polls show the early front-runner has no better than a 50 percent chance of being the Democratic Party nominee.

Since 1972, the Gallup Poll's early front-runner of the non-incumbent party has fallen by the wayside before the finish line in half the races. In November 1970, Sen. Edmund S. Muskie (D-Me.) was two points up on Sen. Edward M. Kennedy (D-Mass.). Sen. George S. McGovern (D-S.D.), who eventually was nominated, had only 2 percent support from the Democrats polled then.

In late February 1974 Jimmy Carter didn't even show up on the Gallup Poll. Kennedy was the leader among Democrats with 44 percent, followed by George Wallace with 17 percent. Little had changed by September 1974. When Gallup did a laundry list of possible candidates (Hugh Carey, Ella Grasso, Milton Shapp, Julian Bond and Sargent Shriver to name only a few of the less prominent) in November 1974, Carter didn't appear on that list, either. Carter's debut on the Gallup Poll came in March 1975 -- with 1 percent of the support -- still behind Julian Bond (4 percent) and John V. Lindsay III (2 percent).

In 1980, the early polls showed Ronald Reagan leading the Republicans, but when Gallup polled Democrats for their choice starting in 1979, Kennedy was well ahead of Carter (53 to 21 percent).

If the polls are no better than a coin toss for predicting the winner, what is the point of this exercise? I posit that political man suffers from some primordal, cosmic anxiety that requires him to know -- far in advance, apparently -- who is going to be president. So we have this superstitious ritual every four years where we try to guess -- that's all it is really -- who the nominees are going to be.

For most of us, it couldn't make the slightest difference. Most of us, after all, don't feel an intimate connection with the executive branch of the federal government, unless of course one of us, or a loved one, is drafted, or our taxes are raised.

Some people, however, do have that intimate connection to the executive branch. Among these are people called political writers. They make their living this way, and obviously, they have to have something to write about. So first of all, they need a competition, a contest, a race -- a sporting event, if you get my drift.

The media, of course, love to have a front-runner. That way, they have something to focus on -- perhaps zero-in is a better phrase. The political writers then can try to generate early excitement by assessing the quality of the candidate's organization (the equivalent of the offensive line in football).

When people get tired of reading about how overwhelming the organization is, political writers can construct a David to take on Goliath. It's usually no problem to find a suitable candidate because he or she doesn't need to do anything. Say yes, you're interested, and you're in the race. Who cares if you're 40 points behind -- you're David. Say no, you're not a candidate, and you're just being coy. If you keep on saying no, eventually the political writers will nominate someone else to be their David.

Now, here's another problem. What if no one is a clear front-runner early on? Then what? Simple. Political writers appoint one. My favorite is the Time Magazine cover of Feb. 17, 1975 picturing Sen. Henry Jackson (D-Wash.) with the caption "Scoop Out Front." Based on what? The Gallup Poll had George Wallace on top then with 22 percent, followed by Sen. Hubert Humphrey (D-Minn.) with 16 percent and then Jackson with 13 percent. Wallace was unacceptable as the front-runner. Nobody would read stories about Humphrey anymore, so give Time credit for trying to get something going, even if Jackson eventually went absolutely nowhere.

The whole focus of political reporting is based on possibilities and probabilities -- can he win? is the key question. Should he win is a question that political writers don't ask. Even if a candidate is privately thought to be "good," political writers won't treat the candidate seriously unless he or she has the money, organization and a certain je ne sais quoi that labels the candidate a possible nominee. Now if the candidate lacks money and organization but still manages to defy all the odds and demonstrate that he has appeal -- that certain je ne sais quoi -- then of course, it's a whole new ballgame (e.g., Gary Hart in 1984).

Why do political writers feel compelled to tell their readers -- who presumably constitute the people making the choice in the first place -- whom they seem to favor when it's clear that the voters aren't sure themselves? Why do political writers keep trying to predict the outcome of elections when they have been demonstrably wrong so often?

Could anyone have foreseen that Eugene McCarthy's 1968 effort would persuade Lyndon Johnson not to run again? That McGovern would capitalize on the nominating rules procedure changes he had presided over to put himself in the driver's seat? That Carter would tap a deep-seated urge Americans felt to trust their president? That Walter Mondale's grand organization would cloak a man-sized Wizard of Oz who was only too human -- and dull to boot?

The answer is probably not.

As for myself, as in the past, I will resist the urge to pull aside one of my political writer colleagues and ask him furtively to tell me who is going to win because I can't stand the suspense. The truth is, I can.