TAKEN AS A WHOLE the midterm elections could be either a boon or a disaster for pundits. It all depends on how inventive, not to say shameless, the individual pundit is in dealing with the kind of election returns that can be used to prove either everything or nothing on a national, big trend, whither-the-two-political-parties scale. We reached this conclusion the hard way: by trying to extract a few interstate truths and transcontinental theories from the returns (beyond the self-evident fact that an anti-tax politics is prominent in the land) and observing that none of our big ideas could survive scrutiny.

Try this one: that the voters, nationwide, were registering their fed-upness with things as they are and with the used-goods politicians who have been hanging around for years symbolizing the status quo. The triumph of Harry Hughes in Maryland, the upending of the whole Minnesota top Democratic leadership, the dumping of both conservative Republican Gov. Meldrim Thomson and liberal Democratic Sen. Thomas Mcintyre in New Hampshire, the defeat of Republican Sen. Edward Brooke in Massachusetts - you can accumulate a lot of evidence for this conclusion. But what are you going to do with the continuing success of Sen. Strom Thurmond in South Carolina, or the ease with which voters returned to office two notorious congressional establishment stalwarts, the indicted Rep. Daniel Flood of Pennsylvania and the convicted Rep. Charles Diggs of Michigan? Yes, it's true that Republican veteran Sen. Charles Percy of Illinois nearly lost, but he pulled out of that noisy near-fatal dive - admittedly only some 15 feet above ground, but he did it. All wasn't ashes for the longtime "ins," and in fact, in the far West, the political climate seemed to favor incumbents.

Similarly, you can show it was a fine day for left-of-center Democrats or a mini-rout. In New Jersey, Bill Bradley prevailed over the articulate and tough conservative Jeffrey Bell, whereas , in Iowa, Dick Clark got beat by the very-far-to-the-right Republican challenger, Roger Jepsen. What all this strongly suggests in 1978 is what it almost always strongly suggests - that individual and even idiosyncratic political circumstances tend more than great whooshing national trends to determine who wins and who loses around the country in a midterm election.

The Democratic debacle in Minnesota is a good example, because, among other things, it had been building for years and for reasons that predate by at least a decade the current political atmosphere. Minnesota Democratic politics is, in fact, the story of a success-ruined enterprise, a system that produced too many winners and too much victory and too much leadership. The leaders began to compete among themselves; the winners closed down the channels to winning for younger and equally ambitious contenders; the victories bred, ad they often do, complacency and, in the case of national victories, created leaders like Hubert Humphrey and Walter Mondale who were obliged to represent government policy and values to a constituency that was by its nature and history critical, independent, unaccustomed to accepting an establishment "line" - even its own.

The defeats of Sen. Robert Griffin in Michigan and Sen. Floyd Haskell in Colorado are additional examples of the kind of special-circumstance outcomes we have in mind, and the phenomenon could be multiplied many times over. Some really terrific people won - and lost. It's the American way.