If you ask a political buff in this central Illinois city who's going to capture the House seat Nov. 7 here in Illinois' 22nd Congressional district, you'll probably be told that Republican Dan Crane is the likely winner.

Crane, after all, has a poll to prove it.

But if you drive 120 miles down State Rte. 1 and ask the same question of a politician in the southern sector of the sprawling district, you'll probably hear that Crane's Democratic opponent, Terry Bruce, is going to win.

Bruce, after all, has a poll to prove it.

If you take the question to TV station WCIA, in Champaign, the news staff will refer you to the WCIA poll, which showed Crane leading by 10 points two weeks ago. But if you ask again at radio station WSHY, in Sheibyville, folks there will tell you that their poll shows the two candidates running just about even.

Although weeks of work and thousands of dollars have been poured into various polls to predict the winner in the 22nd, a rural switch or east central Illinois, nobody can say with much confidence today which candidate is likely to prevail.

A nonpolitician with a little patience might be tempted to ask, why bother? If the candidates could just wait three weeks until election day, their curiousity would be stated without the trouble and expense of taking a poll.

The answer is that curiosity about who will is only one reason - a minor one, at that - for taking a poll. Contrary to the classic cliche that "the only poll that counts is the one on election day," opinion polls have been important tools for both candidates in the 22nd, contributing to decisions about strategy, scheduling, and stands on issues.

The news garnered through polling, however helpful, is not always welcome.

Several months ago, for example, Crane solicited the support of former agriculture secretary Earl Butz. To Dan's delight, Butz agreed to help, and he has traveled with Crane on campaign swings through the district's farm towns.

But when Crane threw a question about Butz into one of his polls, the candidate learned to his chagrin that Butz was actually a negative factor, driving away more votes than he attracted. As tactfully as possible, Crane's managers made it clear to Butz that his services would no longer be needed.

Crane's most recent poll carried a similar message about Don Watson, the prominent local Democrat who lost to Terry Bruce in the Democratic congressional primary last spring. Watson had electrified the district in September by announcing his support for Crane, the Republican, in the general election.

But Crane's poll showed that Watson endorsement would have little impact on independent voters, and might stir sympathy for Bruce among Democrats. So the Crane camp decided to play down its new Washington connection.

Such useful information does not come cheap. The Republican Party has put up nearly $10,000 to purchase two polls for the Crane campaign from Lance Tarrance and Associates, a Houston firm that does business with many conservative candidates. The Bruce campaign has paid somewhat less for less ambitious polling by Cliff Worth, a Carbondale, Ill., pulsetaker who has worked for Terry's political mentor, Rep. Paul Simon (D-Ill.). Both candidates are scratching now find money for more polling in the campaign's final weeks.

The polls tell the candidates which geographic areas, which issues, and which personality traits they should emphasize in their campaigning.

All the pollls to date agree that Terry Bruce is leading in the southern half of the district, a region that he has represented for eight years in the state Senate. Crane, a dentist and prominent Reaganite from Danville, is doing better in the more populous northern counties near his hometown.

But Crane's latest poll showed that he made significant inroads in Bruce's southern turf this fall. His managers quickly began shifting the schedule to get Dan into the southern counties more often, since they were no longer considered out of his reach.

Crane's expensive polls provide elaborate breakdowns indicating which issues matter most to each bloc of voters - classified by age, sex, party, hometown. Thus when Dan sets off to give a speech to, say, the 4-H club in Flora, he can check the poll to find out what young people in that area want to hear.

In addition, of course, the polls tell the candidates who is ahead. But the various polls do not agree on that.

Crane's first poll, taken in May, showed that Terry Bruce had a lead of six percentage points, with a third of the voters undecided. A poll commissioned by Bruce at the end of summer came up with roughly the same result.

A second Crane poll, conducted at the end of September, has just been tabulated, and it provides two answers.

One question, in which the respondent was asked flat-out "Who are you voting for this fall?", showed this result:

Bruce:41

Crane:40

Undecided:19

But when the poll takers pushed harder, pressing the undecided respondents with the question "Whom are you leaning toward?", the result was reversed:

Crane:43

Bruce:42

Undecided;15

Despite its ambiguity, the poll was a forceful spark for Crane and his campaign team. For the first time, the whole staff was convinced that Dan Crane was a winner. In the belief that others would think so as well, Crane quickly dispatched stacks of letters citing the poll as a new reason for contributors to send the Crane effort a little more money.

The poll had one additional effect: it inserted a new note of caution into the Crane campaign. If Dan was in fact leading, his manners said, he'd better be careful what he said and did during the last few weeks before the election. That advice plopped Dan Crane into an agonizing philosophical dilemma.

NEXT: A Morale Crisis