Within the next three weeks, by March 15, both the Republicans and the Democrats will have assigned more than half of the total delegates it will take to win each party's nomination.

You want to see a graph to that effect, you say? Sure thing.

There are 26 individual votes on the Democratic side (including territories and the conclusion of Democrats Abroad) and 31 on the Republican.

Or, to cut to the chase: This is a critical period for figuring out who will go on to represent both parties in November. And the data we have about how this will all turn out is very sketchy.

In most states, on both sides, there hasn't been any new polling conducted since the Iowa caucuses. In some cases, the most recent polls in a state show people like Chris Christie winning, because they were conducted in 2013. In others, conducted last fall, Ben Carson is doing disproportionately well. If we look only at new polls conducted this month -- allowing at least for the re-shaped race after voting began -- here's what we know about the upcoming states. (Public Policy Polling surveyed a number of states on the Democratic side. Most of the other polls are from in-state pollsters, compiled at Huffington Post Pollster.)

Since it's usually Donald Trump and Hillary Clinton in the lead, when anyone is, we've scaled the graphs against those leads.

There are a number of other factors at play here, including that delegate allocation varies from state-to-state, but the point is: Lots of delegates are about to be allocated and we don't know who's going to get them.

We can assume that Clinton and Trump will win a number of them because they're winning nationally. Part of the challenge of Super-Tuesday-like days is that it's a lot harder to campaign in all of the states over two weeks than it is to campaign in Iowa and New Hampshire over ten months.

The assumption that the national leader will win those states, though, has a big flaw in that national polling averages aren't predictive. The closest any candidate's actual results have been to their polling average on the day of an election (or to the most recent polling average) was John Kasich in Iowa who was polling very poorly nationally and did very poorly in the state. Otherwise, the results have been all over the place.

On average, Trump's and Ted Cruz's results have been at least six points away from their national polling. Marco Rubio's have been nine points. Bernie Sanders's? Thanks to New Hampshire, he's averaging 14 points.

Of course, three weeks ago, we also didn't know who'd win the very first vote, in Iowa. We were as blind about where we'd be now as we are now blind about where we'll be on March 15.

The point being? Take predictions of where the race is headed with a nice grain of salt.