What a difference a few days — and more than a few dollars of attack ads from the Romney super-PAC — have made in the race for the Republican presidential nomination. Mitt himself piously and hypocritically denounces all super-PACs, even as his shreds Newt. The attacks have been amplified in the media echo chamber, reinforced by a relentless barrage from the GOP establishment, and validated by Gingrich's chronic and flap-mouthed tendency to verbal excesses that he can't or won't explain away. Witness his attempt to deal with his Freddie Mac windfall by boasting about how much he makes in an hour for speeches to special interest groups — more than most American families earn in a year. And now in Iowa, he has suffered the defection of religious right leaders briefly inclined to him; their followers simply declined to forgive his past personal transgressions, even if the Lord has.

So it turns out that Newt isn't teflon after all, but more like velcro. Ron Paul has grabbed a lead in the Hawkeye State, with Romney even edging ahead in one poll. Ironically, more of the Gingrich voters there name Romney as their second choice — all of which opens up the possibility that he could seize a longshot victory from the jaws of long-likely defeat, cement his comfortable lead in New Hampshire, ride the resulting momentum to a respectable finish in Mormon-averse South Carolina, then take Florida and almost certainly an early nomination. 

Ron Paul, the congressman and crank with the prejudice-and-poison newsletters, and a treasure trove of conspiracy theories, isolationist appeals, and calls to legalize drugs, is Romney's own second choice in Iowa. For Republicans, he's the unthinkable nominee and for Democrats, the ideal one. If he emerges as the alternative, the deal is sealed for Romney. 

After all the years, all the cash, all the campaigning, and the serial collapse of his challengers, will Romney win Iowa — and end up taking the whole game?

Despite a temptation to believe Gingrich could do it — in a contest between the conservative and the con man — the flaws in the former Speaker and the other former frontrunners have left the flip-flopper headed for an inevitable if reluctant coronation at next summer's convention in Tampa. As I wrote earlier, who the hell else they got? After an endless search for the un-Romney, Romney seems to be on a steady mid-20 percent path to success, precisely because he is the un-Gingrich, the un-Paul, the un-Cain, the un-Bachmann. As for Jon Huntsman, he needs a Romney humiliation in Iowa to move from modest liftoff into the New Hampshire stratosphere.

Any such humiliation appears to be a forlorn hope, but the far-fetched has happened in Iowa before. In 2004, Howard Dean was far ahead and John Kerry far behind just weeks before Kerry dominated the caucuses. In 2008, Mike Huckabee prevailed when McCain strategist Steve Schmidt shrewdly dispatched his candidate to New Hampshire at the last hour to siphon votes from Romney. McCain's fourth-place finish in Iowa was actually a win that propped up Huckabee through a three-way race in Florida which the soon-to-be nominee carried with a minority of the vote.

After all the years, all the cash, all the campaigning, and the serial collapse of his challengers, will Romney win Iowa — and end up taking the whole game? The lead-up to caucus night is rife with sound and fury, much of it signifying nothing more than political noise. So let me offer my guide to the caucuses I've known for over 30 years. 

First, Iowa has a different predicative value for Democrats and Republicans. Since 1980, in contested caucuses, the Democratic winner has gone on to the nomination 84% of the time. For Republicans, the number is 60%. Ronald Reagan lost on his way to the White House; so did George H. W. Bush. 

Second, and this should cheer Ron Paul and his kooky fans, Iowa Republicans don't mind giving a boost to preposterous candidates so long as they are ideologically pure. Thus the Reverend Pat Robertson, the one who heralds hurricanes as divine punishment for America's sins, finished ahead of the next president of the United States in 1988. The risible Alan Keyes, a perennial candidate whose politics were exotic then, and who would later become a "birther" who denounced President Obama as a "usurper," finished third in 2000. No such figure has ever done as well on the Democratic side. The earnest, entirely un-presidential Dennis Kucinich ran as the purist anti-war choice in 2004, but caucus goers dismissed him with about 1 percent of the vote. 

So there is hope for Paul — and maybe even a glimmer of possibility for Rick Santorum, the ex-senator from Pennsylvania who was thoroughly thrashed for re-election in 2006, but has just been endorsed by the evangelical activists who left Gingrich standing alone at the altar. They've embraced Santorum as the anti-gay, anti-women's rights, anti-separation of church and state candidate. That does appear to describe all of the GOP field in Iowa, but the activists calculate that maybe they can pull off an upset for the most fervent of the lot. (They've written off Michele Bachmann and urged her to drop out.) 

Paul or Santorum would be fine with Romney as long as the former Massachusetts governor finishes second. He would beat Paul like a drum, and Santorum comes across more as a sour little drummer boy than a credible president. 

Third, Iowa caucus goers shop late and then settle on their final preferences. I've already noted the Kerry surge in 2004 — when most of the press corp and political wise guys assumed until the last days or even the day of that the contest was down to Dean and Dick Gephardt. In his first try, in 1988, Gephardt was limping along at 2 to 4 percent before he went on television only weeks before the caucuses — and won. Reagan led for months before losing in 1980 — and the first Bush led then lost in 1988. So did Mitt — and Hillary — the last time out. 

As caucus day nears, Iowans ask a question, and then decide which candidate is the answer. For Democrats in 2004, the question was who was a plausible commander-in-chief, someone who might beat Bush. So they famously dated Dean and married Kerry. In 2008, the question was who stood most for change. Hillary Clinton ran on experience — and Obama rode a wave as he proclaimed the "fierce urgency of now." For Republicans that year — and Republicans in Iowa may be less pragmatic then Democrats — the question then was what it may be again this year: Who's the most conservative? That's where Romney faltered four years ago, and that's why he may benefit from a close-run, fragmented finish on January 3 that could let him win with decidedly less than 30 percent, or come in second to an opponent who is entirely expendable and massively underfunded in the ensuing contests. (If that's Paul, he may later exact revenge by launching a third party bid, all but ensuring a victory for Obama.) 

There is a wild card too — one that may be remotely in the offing. The Romney super- PAC has just targeted Perry as well as Gingrich for being insufficiently intolerant on the issue of immigration. The PAC's strategists are obviously seeing numbers. Either the Perry hit is designed to tar Newt with guilt by association, or Team Mitt is concerned that after turning on Gingrich and passing on Santorum, hardcore conservatives could revert to the big-buckled, big-haired, big-bloopered Perry. At least he'd have the resources to compete in the coming months — and as hard as it is to believe, Republicans could conceivably nominate him.

One thing in all this is clear: Romney is not conceding the conservative label and he's pursuing the heresies of others with no sense of shame over his own past apostasies.

Will it work? Well, fourth and finally, the best place to look before the caucuses is in The Des Moines Register. While the editorial board's endorsement doesn't mean much, the paper's Iowa Poll does.

Romney captured the endorsement, but it may be cold comfort to him that the Register's pick has prevailed only 25 percent of the time since 1988. Nate Silver concludes that "6 of the [chosen] 8 candidates" — even those that lost — did "beat their projections [based on polling], although it was by a small margin in most cases." Of course, given the tightness of the race, a mini-bounce may be enough for Romney. 

We shall see — and we will know more — when the Iowa Poll drops like the ball in Times Square on New Year's Eve. In the last three cycles, in both parties, this poll's findings virtually mirrored the final results. In 2000, the Register's numbers sounded doom for Bill Bradley against Al Gore — and by the 2-to-1 margin that was soon to be recorded at the caucuses. The same survey had George W. Bush at 43 percent, just two points more than he actually received. Four years later, the final Iowa Poll stunningly (and correctly) reported that Kerry had soared from single digits to first place, and that John Edwards was coming up fast. In 2008, Obama was seven points ahead of Clinton in the final Register poll — he bested her by nine — and Huckabee was six ahead of Romney, and actually beat him by nine.

This is a remarkable track record that stretches back to 1988. It was the Register's finding that showed Gephardt, in his first race, rising fast from the ashes of last place. 

There are countless polls out there. But this is the Iowa caucuses and on the record it's the Register's Iowa Poll that counts. It has an uncanny feel for the penultimate tides that so often sweep the state and decide the contest. 

Maybe this time, Iowa Republicans will improve their batting average and go with the eventual nominee. Maybe they will flee the far-right and far-out choices. And certainly, they're still making up their minds. Wait for the Register survey and then wait for caucus night — not just for the results, but to see which way Mitt Romney flip-flops. He always knew Iowa was critical, he will exult — or he'll muster some polite words to insist that it's just one stop on the journey to Tampa.