The magnitude of the Republican victory this week and the defeat of Democrats in the Senate lay beyond the predictive abilities of the press.

It is all well and good to say that journalism is not a mechanism for predictions, but there is no avoiding the fact that political coverage and the polls indicated a close race that never happened. Even late forecasts of a Reagan win did not anticipate the thundering Republican landslide.

In the aftermath, many thoughtful editors are puzzling over what went wrong. should they have seen it more clearly? are their measuring devices good enough? should they trust political instincts more than the polls?

Richard Leonard, editor of The Milwaukee Journal, is preplexed. "Late decisions by voters may have thrown off normal perceptions, but this is not one of the glorious moments for polling. it is an art that leaves much to be desired."

For Eugene Patterson, editor of The St. Petersburg (fla.) Times, the press was too conservative. "Both the press and the polls were more cautious than the public this time around," Patterson says.

Harry Rosenfeld, editor of Albany's Times-Union, believes there were too many undecided voters and the national polls weren't watching the Senate races. But he defends polling techniques while saying they failed to detect that voters were not as committed to their choices as pollsters believed.

Claude Sitton of the Raleigh (N.C.) News and Observer, said pollsters will take great comfort in their margins for error. He believes there are no hard and fast rules for political decision and finds it reassuring that voters can change their minds outside of efforts to quantify them. "I think many voters were simply telling the pollsters that it was none of their business."

In The Washington Post, the stress between poll results and political coverage showed up in two stories that ran side by side on Sunday. The polls showed a narrowly divided electorate, with Carter in a slight lead that was within the margin for error. Voters were described as unusually volatile, with 10 million undecided. The findings agreed with a Gallup poll.

The articles on polls reported that it was impossible to say whether Carter or Reagan "holds a lead in this extraordinarily close election."

Contrasted with that, The Post's leading political article of the day examined the electoral vote map found Reagan in the driver's seat and said that Carter's only hope lay in a return of the American hostages in Iran or some other event of similar dimension.

The political article went on to say that there was a possibility of a Republican landslide and found evidence of "brightening Republican prospects"in Senate races.

Barry Sussman, quarterback of The Post's polls, is philosophical about the discrepancies. "We explain better than we predict," he said.

The most telling comments come from Michael Gartner, editor of the Des Moines Register. "We failed our readers," he says. "Maybe a large number of voters changed their minds late. Maybe the pollsters wearied. Maybe the polls have outlined their usefulness. We were extremely cautious about how we interpreted our polling data. We wrote in a lot of disclaimers. We should have been able to report more accurately. Something was terribly wrong about how we handled the preelection story. Nobody expected what happened."

It is clear that the decisions made in the privacy of voting booths were too intimate for the blunt instruments of journalism to detect in advance. A powerful political rhythm was indecipherable until, in its own time, it produced its own evidence. The election was conspicuously more interesting and dramatic than the press discovered.

It will be even more fascinating to see if the experience to 1980 will lead the press to turn its back on polls and take risks in predicting political decisions.