NOW THAT the nominating conventions are about to start, the time is ripe for revealing the Brown system for Predicting Eventual Election Results (or PEER for short). Its accuracy is not, of course, any more guaranteed than the Harris, Gallup or Washington Post polls, but it is a lot cheaper to undertake and has the added advantage that you can conduct it yourself. In a do-it-yourself age, this is no small asset. y

Of its accuracy, there is much to be said. Only once in more than 30 years have I had doubts concerning it, and the extenuating circumstances then were easily correctible. Can Gallup say as much?

Putting PEER to work is simplicity itself. You simply tune in your TV when one of the principals in the current presidential sweepstakes is front and center, turn off the sound, and concentrate on what you see. No listening is necessary. In fact, not listening is the secret ingredient.

Contrary to traditionally held views, you will be missing very little. Thus far, no one except John Anderson is saying much of anything worth hearing anyway; and what with the almost universal tendency to take positions that offend no one except the Russians, you will be spared a lot of nonsense about inflation, gun controls, balanced budgets and abortions. This is why PEER is so effective. It simply tunes out the irrelevant and immaterial and allows one to concentrate on what is really important: how people look while they're saying it .

If you majored in political science in college or are a member of the League of Women Voters and, therefore, troubled by the implications of all this, there is, I would remind you, solid theory behind the PEER approach. The great majority of Americans are paying less and less attention to what people in high places are saying these days; and, anyway, issues are getting so complex that you have to work for a research corporation or write editorials to understand them. So, why bother?

After you have studied each candidate several times under varying circumstances (speeches, interviews, debates, airport arrivals, and commercials) without the audio, you will probably have acquired a sufficiently large data bank to make comparisons. Americans traditionally vote against rather than for, and you will do well to look for negatives and to evaluate them. It is not hard to do. You can even invite friends to help you, including the children, whether 18 or not.

My discovery of the PEER method in 1948 was serendipity itself. During the Truman-Dewey campaign, the sound on my TV set went on the blink and I was left with the alternative of either turning the thing off or making do with the visual only. I chose the latter.

Truman, you will remember if you are over 50, was singularly ungifted as a speaker, while Dewey had, or was reputed to have, considerable skill as one. But with the audio off Truman clearly had the advantage. His image was a vibrant one -- he came through human, fighting and convincing. Dewey, on the other hand, looked smug, arrogant and overconfident. I remember telling friends that if the Democrats could arrange to turn off the sound on half the TV sets in America, Truman would win easily. Maybe that was what happened. Or maybe, as Mencken might have put it, the "great rancid American public" really got the word and voted what it saw. A week or so later, when a repairman finally got around to changing three or four of our tubes, I found myself reverting to the "sound off" process I had inadvertently developed. Thus, PEER was born.

I have tested the system anew in each succeeding election. The Eisenhower-Stevenson campaigns of 1952 and 1956 were not contest. It was the masculine, no-nonsense leader against someone who without sound came through as weak and at times effeminate. No matter the quality of his mind or the wit and erudition of his speeches, Adlai muffed the visuals -- which is, of course, what both TV and elections are all about. The Kennedy-Nixon debates, soundless, sounded the doom of the Nixon candidacy. Even though he seemed less in need of a shave as the debates went on, Nixon's manner did him in. Kennedy was the warm, bright American boy who could easily be worthy of the White House.

All surveys have their weakness, whether one uses a telephone book or "The Statistical Abstract" to make one's sample, and the Johnson-Goldwater campaign of 1964 tested PEER to its fullest. Goldwater as a man came through more strongly, but the system should not entirely be faulted. At least, without sound one was spared both the Johnson accent and the wild promises, including the one about Social Security, which Goldwater was making. I ultimately picked Johnson as the winner because PEER revealed something in the places he visited. The cameras faithfully recorded the crush of people around him, the eagerness with which they sought to get close to him, their enthusiasm for him. In the end, I gave him the nod.

The second time Nixon ran for the presidency, he (or his advisers) had learned something about the medium. He was, for one thing, clean shaven, or at least cleanly powered, and less shifty-looking. (Those things count.) Hubert Humphrey, whatever his other virtues, came through jerkily, like a character in a movie from the 1920s. I knew it would be close, but PEER's prediction was Nixon. Nixon against McGovern in 1972 was no contest. Whatever else he had going for him, McGovern (like Joe Clark, the recently deposed Canadian prime minister) did not look like a president.

The Ford-Carter campaign was again a close one. Both are All-American-looking men who smile readily, and the tube showed them that way. The key event, in my judgment, was that night the sound track broke down during one of the debates. I was listening soundless at the time and was confused, to say the least, by the network's accommodation. Both men seemed to stand there endlessly, waiting for the cables to be re-spliced. Carter did much the better in pantomine that night. Ford appeared distinctly uneasy. After all, he was president and presidents are supposed to act decisively even when TV cables come apart. He didn't, and it hurt him.

What will PEER reveal in 1980? Here are some observations I have already made. The president, for one thing, has aged appreciably. This can be proof that he is really working at the job; or it can mean that it is grinding him down. Nor does he look as happy as he once did. I suppose it is a little much to ask someone to look elated with inflation what it is or the ayatollah what he is, but Americans want their chief executive to show confidence. Rosalynn may be doing that but Jimmy isn't.

The Reagan problem, even if his hair looks as youthful and well groomed as Nancy's, is that matter of the aging throat. His collars, expensive as they must be, don't quite cover what the cameras have revealed of an old man's neck. Pleasant as he comes through, he doesn't moves as sprily as he might.

Ted Kennedy with the sound off varies between looking (a) the peerless leader and (b) being uncertain of himself and bored with what he has to do. The contrast leaves one perplexed. We will be voting for a full-time president, not one who is good only on Mondays, Wednesdays and Fridays.

John Anderson has some important Trumanesque mannerisms about him. Also, we haven't had a white-haired president in well over 100 years, and this might be in his favor, as being a tall blonde from the Southwest helps out in a Miss America contest in the year after a redhead from Minnesota won it all.

CBS's Voter Profile Analysis predicts winners by testing those who have voted by 11 o'clock in the morning on election day. This is a good stunt and has seemed to work so far. To my way of thinking, however, PEER is more reliable in the long run, less expensive, and a lot less trouble. Try it yourself and then decide who's going to take it all. You'll probably learn more from it than you have about what is going on in Tehran.