Joe Biden speech
Watch Biden's blistering rebuke of Trump
02:38 - Source: CNN
CNN  — 

First things first: The theme song of the week is the theme from The NFL on Fox.

Poll of the week: A new Monmouth University poll of likely Democratic Iowa caucus-goers finds that former Vice President Joe Biden leads the field with 28%, while Sen. Elizabeth Warren of Massachusetts at 19%, Sen. Kamala Harris of California at 11%, Sen. Bernie Sanders of Vermont at 9% and South Bend, Indiana, Mayor Pete Buttigieg at 8% round out the top five.

This poll is the latest to find that Biden is ahead in Iowa, though with less than 30% of the vote.

What’s the point: Biden’s candidacy is perplexing to analysts because he’s clearly a front-runner (i.e. he leads in every poll), but he’s not running away with the primary. This is true nationally (where he’s hovered around 30%) and is especially the case in Iowa, where he’s polling lower.

Based solely on the polling, history suggests that Biden has a considerably better chance of winning the caucuses than any other candidate. Still, his chance of winning is less than 50%.

I updated an analysis from a few months ago where I looked at where all candidates were polling at this point in open Iowa caucuses (i.e. ones without incumbents running) since 1980. From that data, I produced the chance that someone polling at a certain level would win the caucuses.

A candidate polling in Biden’s position would win 30% to 40% of the time, depending on whether he was polling closer to the 24% he drew in a CBS News/YouGov poll produced in July or the 28% he had in the Monmouth poll.

On one hand, this does not seem high, given that this means someone other than Biden would be projected to win 60% to 70% of the time.

On the other hand, Biden is doing significantly better than a number of past national front-runners who fell flat on their faces in Iowa and nationally. Jeb Bush in 2016, Rudy Giuliani in 2008 and Joe Lieberman in 2004 were all losing at this point in Iowa and polling at or less than 15%. Biden has led in every CNN-approved Iowa poll this cycle.

If Biden’s polling position held through the eve of the caucuses, his chances of winning would skyrocket. At this point, however, he would need to start polling in the mid-30s to become a better than 50% proposition to win in Iowa.

If his polling falls, Biden’s chance of winning drops precipitously. Incidentally, once a candidate is in the low 20s at this point, his or her chance of winning correlates nearly perfectly with where the candidate is polling. The chance of winning dips to 20% when a candidate is at 20% in the polls; 15% chance at 15% in the polls; 10% at 10%; 5% at 5%; and 3% when they’re at 1% in the polls.

Using the Monmouth poll and applying these probabilities to the other candidates, you end up with Warren at about a 20% chance to win Iowa; Harris, Sanders and Buttigieg with about a 10% chance each; and the other 15% chance scattered among the rest of the candidates.

Another way to look at this is that Biden and Warren combined have a little more than a 50% chance of taking the caucuses. Those who are suggesting that the Democratic primary may come down to Biden and Warren seem to have a point; at least when it comes to Iowa.

When you combine the top five candidates, there’s an 85% chance you’re looking at the eventual Iowa winner (at least according to how past races have developed).

Indeed, there has been only one caucus winner out of the 14 in our dataset to be polling at less than 5% at this point (George H.W. Bush in 1980). That’s where the vast majority of Democratic candidates are right now. I will say the chance of a winner emerging from this under 5% group is probably a little greater than usual given how many candidates there are at that level.

Putting aside the model, the fact that the eventual winner in Iowa is currently polling below 30% shouldn’t be surprising. The average Iowa winner at this point in the past has been at about 26% in the Iowa polls.

The question is how far south of 30% is that eventual Iowa winner this year?