Walter Mondale's defeat in the Milwaukee straw- balloting last weekend could be liberating for him. The setback has been minimized by his people, described not inappropriately as the political equivalent of baseball's exhibition season. But now Mondale, whose support has been compared unfavorably with that of Edmund Muskie, a failed front-runner, as being a mile wide and an inch deep, has received a semiofficial report card that his candidacy is in trouble. Muskie did not get his evidence until the New Hampshire and Florida primaries in 1972, by which time it was too late.

Mondale lost in Wisconsin, and one underlying premise of his campaign--the inevitability of his nomination--was a casualty. Those Democrats who were concerned that Mondale would triumphantly waltz, on the strength of endorsements and bushels of smart money, to a coronation at next summer's San Francisco convention, need no longer worry. A Democratic brawl is guaranteed. The shape of that fight will be determined in large part by decisions Mondale makes in the next few weeks.

In making those decisions, he and his aides might want to consult a new book by Mark Bisnow, who appeared in the 1980 campaign as the traveling amanuensis of John Anderson. Drawing on his experiences in Anderson's bid for the presidency, Bisnow has written "Diary of a Darkhorse," which has an interesting theory about what makes a candicacy go.

Bisnow writes that the Anderson campaign, which in late December 1979 had to float a loan to meet its modest payroll, was by late February 1980 at least dog-paddling in money. Only two major events had intervened to explain the change. At a GOP candidates' debate in January, Anderson had been outspoken in debunking his party's orthodoxy, which called for simultaneous tax cuts, balanced budgets and defense increases. In February, the nervy Anderson was televised confronting some growling New Hampshire gun owners who looked like extras from the movie "Deliverance."

The conclusion: every political campaign is about differences. Differences in values, in vision and in alleged virtue. In the long run, Anderson's campaign was essentially derivative; his principal recommendation, according to his supporters, was that he was not either Jimmy Carter or Ronald Reagan. That argument did not prove to be persuasive for 93 percent of the voters.

What is the Mondale difference? What is unique about the Mondale candidacy? Why should citizens, other than those fair-weather pragmatists, care that Fritz Mondale rather than Fritz Hollings is the next Democratic president? Who are the villains a President Mondale would bring first to the bar of justice? Who are the victims he would first aid?

The first post-Wisconsin report that he considered the nuclear freeze, with which Wisconsin winner Alan Cranston has been most closely identified, the most important campaign issue did not help Mondale. There was unfortunately, if unfairly, a whiff of "me, too" to the statement. But now, it's up to Mondale to tell the nation what kind of president he wants to be, who he really is. It may be his last, best chance.