Two weeks till the real test.

On March 1, there will be 12 Democratic contests and 13 Republican contests. That is just two weeks away. It is easier to focus on the two closer contests (Nevada and South Carolina) than it is to focus on so many contests in so many places. Reporters, operatives and regular voters tend to do that for cognitive reasons. They also justify this focus based on the idea that events, momentum and trajectory matter and reshape subsequent races.

Despite the belief that events trigger later election outcomes, in fact coalitions are likely to have already been built and that means events do not register quite as much as expected. Just this year, for example, the results in Iowa seem to have had little impact on the results in New Hampshire, and that was true on both sides of the aisle. Likewise, in 2008 the results from the first four contests would have been good enough to predict the outcome of the Democratic race in all 50 states. Who was strong in what coalitions, how important caucuses were going to be for Obama, all of that was in place early on.

This is why South Carolina and Nevada may seem important but the new National NBC Survey Monkey poll could be more revealing.

On the Democratic side what is clear is that while Sanders now leads with white voters 47-44, Sanders’ effort with non-white voters is going exceptionally slowly. This is a problem because he only has two weeks to close a big gap. Although Sanders has closed a bit nationally overall in this poll, his New Hampshire win did not even come close to ending the wide split.

The more voluminous polling is South Carolina is more instructive because serious efforts are being made to close the gap among African American voters there. Sanders seem to have made some gains but nowhere near at the pace needed to fix the problem by March 1st. The pattern of whites supporting the non-establishment candidates and minorities not has appeared in many Democratic primaries in the past, particularly in Andrew Cuomo’s 2014 primary challenge. The insurgent closed hard with whites but never got close with non-whites. Sanders in short has two weeks to solve this very large problem across the eight non-caucus, non-Vermont states: Alabama, Arkansas, Georgia, Massachusetts, Oklahoma, Tennessee, Texas, and Virginia. Two weeks is not a lot of time and the door is closing.

There is much more flux on the Republican side because five candidates are still somewhat viable. But if Trump’s lead holds up in South Carolina, his national lead means the other scramblers will have only 10 days to change everything. Since the Republican rules allow for winner- take-all contests, things can change quickly. However, it should be disconcerting for those watching on the Republican side to see Trump very close to 40% nationally with 13 states coming in two weeks.

While the caucus states are a wild card, the rest of the states look like tough nuts for any one else to crack, with the exception of Ted Cruz’s home state of Texas. Trump seems likely to win at least eight primaries, with Cruz and Rubio having a decent shot at two or three states. (What winning means in Colorado is difficult to define because the state will not report results but will send delegates based on who won the delegates.) It is hard to believe that Trump can win 10 or 11 out of 17 contests and will then see a reversal four days later when four other states vote, or seven days after that when another four states vote. Maybe the Trump train can be derailed on March 15th when the big winner-take-all states of Ohio and Florida vote, along with Missouri, Illinois and North Carolina. Maybe, but it is certainly not a sure thing.

We will key in on the next two states soon but the decisive races may be two weeks ago.