I was going to predict Trump 30, Cruz 28, Rubio 19, but before I started writing this Ross Douthat floated those exact numbers and now I’ll be accused of copycatting if I co-sign them. So let’s shave two off of Rubio’s take for my best guess. I think I like that better anyway. We’ve had so many “Marco moments!” from the media during the campaign which never quite panned out that it’s natural to think his late surge will be weaker than expected too.

Here’s your thread to offer your own predictions, although the race is so fluid and Trump’s ability to turn out his voters is such a black box that I think each forecast tells you more about the psychology of the forecaster than about his or her grip on what’s happening. In my case, when in doubt I revert to pessimism: You’ll rarely be wrong in expecting the worst outcome, and even when you are the results will come as a pleasant surprise rather than a disappointment. Ergo, Trump wins. I feel pulled the other way, though, by the fact that Cruz is the one with all the hard evidence of impressive turnout on his side. If you think Cruz is going to win, it’s because you know what percentage of evangelicals typically turn out for the caucuses, you know how many people it takes to staff an effective phone bank, you’re sold on the granular microtargeting that Cruz’s data team is using to reach likely voters, and so on. If you think Trump is going to win, it’s because Trumpmania is amazing and Trump is a winner by definition and there ain’t no stopping this train, baby. It’s the difference between sabermetrics and intuition. To expect a Trump win despite Cruz’s multitude of traditional advantages — endorsements, ground game, analytics, grassroots cred, vocally Christian — is, ironically, to take a leap of faith. That’s a weird thing for an atheist to do.

It’s not all a matter of faith, though. The polls are what they are: Cruz hasn’t led in any survey taken in Iowa over the past 10 days, and various major pollsters show him trailing Trump by exactly, or almost exactly, the same amount. The final surveys from Monmouth, Quinnipiac, and WSJ/NBC all have Trump up seven points. PPP’s final survey had Trump up eight. The big Selzer poll this weekend had Trump by five. If they’re missing on turnout, they’re all missing in a remarkably consistent way. The trendline for Cruz, meanwhile, has been reliably downward. On January 10th, he was above 30 percent in RCP’s average. Today, after weeks of attacks on his eligibility, his loan from Goldman Sachs, his tithing habits, his opposition to ethanol subsidies, and so on, he’s at 23.9. Given that data, it’s arguably as much a leap of faith to think Cruz wins tonight as it is to think Trump does. And as far as his organization goes, this quote from Cruz’s Iowa state director describing their ground-game advantage over Trump keeps coming back to me:

“I’ve seen no evidence of and heard no supporting evidence to suggest they’re running anything like this [phone bank],” Mr. English said, nodding toward the din of dozens of simultaneous phone conversations. “If this is what it takes to get regular caucus voters to go to caucus, it would take this plus something in order to get people who don’t.”

Well, no. Or rather, yes — for a garden-variety politician. The core lesson of the past eight months is that Trump is anything but garden variety. It’s not a leap of faith to think that a guy who spent the better part of a year breaking various rules of primary politics, from his “gaffe” about McCain’s POW status to his barely-even-trying pandering to evangelicals to his willingness to be outspent by other candidates while he relies on earned media, can break one more by turning out droves of first-time caucusgoers without an impressive ground game. Trump’s support, as Harry Enten wrote this morning, is narrow but deep; only a third or so of the party prefers him, but that third is highly motivated. Some Trumpers are part of a cult of personality, others see him as the heaviest blunt object available to throw at the political class, others are devoted because he’s the only candidate focused on protectionism for American workers. The question tonight is simply whether there are, say, 50,000 people in Iowa, Democrats and independents included, willing to show up for an hour to make their devotion plain without a team of campaign workers begging them to do so. To go back to the excerpt, why would it necessarily take more cajoling rather than less for a guy with his own cult of personality to convince his followers to turn out, even if they haven’t caucused before? Is the idea here that Trump fans simply refuse to be proactive in showing up for the one guy running who’s going to Make America Great Again?

Per the latest numbers released from the Iowa secretary of state, there are 616,000 or so registered Republicans in Iowa as of today, although of course more can register at the caucus tonight. The record for turnout in a GOP caucus was 122,000 in 2012, meaning that “good” turnout in the past typically involved 20 percent or so of registered Republicans showing up. By Team Cruz’s own admission, turnout of 175,000 tonight would strongly favor a Trump win. In other words, can Trump mobilize an extra 50,000 of the roughly 500,000 Iowa Republicans who wouldn’t normally caucus to show up tonight, after month upon month of heavy media saturation? Before you answer, bear in mind that there are also 587,000 registered Democrats in Iowa, many of whom are being targeted by Bernie Sanders but some of whom may prefer Trump’s nationalist, Jacksonian brand of protectionism. The real task tonight, then is, to mobilize 50,000 new caucusgoers from the pool of 500,000 leftover Republicans plus the pool of hundreds of thousands of Democrats who may not feel motivated to caucus for Bernie or Hillary for whatever reason. That’s challenging, but not as challenging as Team Cruz makes it sound, I think. I expect Trump will get it done, but that his margin of victory will be less than the pollsters expect precisely because Cruz’s ground game is so good. In fact, here’s a smart point from a savvy friend about an early sign to watch for tonight that things are going Trump’s way:

One other question for the Trump naysayers: If his turnout tonight is destined to disappoint, why are so many of his fans telling pollsters that they’re likely to caucus? What’s the incentive to lie about that? A few months ago, a leading theory among the commentariat was that Trump tended to do better in online polls because online polls afford a degree of anonymity that a phone poll doesn’t. The idea was that there’s a segment of the population that prefers Trump but is embarrassed to say so to another person. Yet in poll after poll lately, including the highly respected Selzer phone poll, Trump’s supporters have not only been willing to admit their support but to say that they’re likely to caucus for him. How do we square that? Are they BSing the pollsters about turning out to make themselves seem more civically minded? Are they telling the truth but are somehow too stupid to find basic instructions on how to caucus today even though every newspaper in Iowa and various websites, including Trump’s, have guides? One of the questions tomorrow if Cruz ends up winning is why people claimed that they intended to caucus for Trump and then ended up not doing so. Answer that question now. If all of the polls are wrong, why were they wrong? Who lied to the pollster and why?

The thread is open. Predict away. For your viewing pleasure, here’s Sarah Palin, former tea-party star, insisting that it wasn’t a hard choice at all between Donald Trump and Mr. Tea Party Ted Cruz. Exit question: How much would Trump have to win by to instigate a real panic among the two-thirds of the party that doesn’t prefer him? A two-point win over Cruz is “survivable,” I think, in that it’ll be spun as the product of a big field. As the field winnows in the next few weeks, Trump will start losing. (He trailed Cruz by double digits head-to-head in the Selzer poll.) If he blows Cruz out, though, winning by 10 points or more, then I think you’ll start seeing “in case of emergency, break glass” reactions by anti-Trumpers. Pressure on Jeb to get out and back Rubio will be intense if Rubio does well tonight. You may hear Romney’s name thrown around again, and maybe whispers about a third-party challenge mounted by conservatives in the new expectation that Trump will be the nominee. The rest of the party won’t simply acquiesce, though, when Trump’s national favorable rating looks as dismal as this. We’ll see.