Last Friday, on Lateline, party heavyweights Rod Cameron (Labor) and Michael Kroger (Liberal) agreed on one thing: that Labor’s huge opinion poll leads are not believable, and it’s really only “about 54 to 46”.

You can see what they were getting at. Sort of. But if polls show Labor ahead by 10 to 20 points, then that’s what they show. There’s little point putting another number on it.

Current consensus has it that Rudd’s lead is “soft”, and “soft” support can be swayed back to the Government. True. But all opinion poll positions are “soft”. There might be “soft” votes on the other side as well. What people are really saying is, again, that they don’t believe the numbers.

Should we believe them? Newspoll’s 60-40 yesterday seemed out of this world; is there any historical precedent at the ballot box? Yes, the biggest results in Australian federal history were about 61 to 39 (against Labor) in 1931, and 59 to 41 (to Labor) in 1943. (Adam Carr’s estimates.) The first saw a change of government in the Great Depression, the second, in World War II, was won by the incumbent.

But in sunny, plasma-telly Australia in 2007, it’s difficult to believe a government could get such a thumping. On the other hand, there’s a first time for everything.

Everyone also agrees Wayne Goss’s famous 1995 “baseball bats” analogy doesn’t apply today, because people don’t loathe John Howard as they did Paul Keating. Maybe not. But if there’s a big election result this year, we’ll decide they did after all. (And the 1996 baseball bats were really only a Queensland thing; Victoria for example gave the Keating government a small two-party-preferred majority.)

Yes, the poll numbers will almost certainly narrow substantially between now and the election. The economy in general, and IR and unions in particular, will be tricky terrain for Rudd, particularly during the campaign.

But there’s a chance the polls won’t change much and we find ourselves delivering the most one-sided result in decades.

As a deep, thoughtful and recent US defence secretary once noted: Stuff happens.

And another thing: Some folks have wondered: why so many Newspolls over the last few weeks? There were two extra ones last week, on top of the regular fortnightly ones. I asked The Australian’s editor-in-chief Chris Mitchell, and his answer won’t set your pulse racing: the paper asked Newspoll for extra questions on IR and who voters think will win the election, and the pollster suggested running extra polls rather than tacking the questions onto the end of the next regular one. Mitchell agreed.