Polling in general has been pretty abysmal during the course of this election season, and national polling has been particularly unsteady. Two new national polls have been released yesterday, one by CNN/ORC and one by CBS/NYT. They show widely divergent pictures of the race, but moreover, they have absolutely no usefulness in predicting the rest of the race going forward, for reasons I'll explain here in a moment.
First, both CNN and CBS show Trump with about 46% of support, nationally. This is in line with the facially insane previous CNN/ORC poll which showed Trump with about 49% nationally. Where it gets dicey is how far behind Ted Cruz is for second place. CNN/ORC shows him at 33%, whereas CBS shows him at 26%.
Unfortunately for the credibility of these polls, Donald Trump has yet to register more than 47% of the vote in any single state. You cannot be at 47% support nationally if you aren't above 47% in any given state. That's just not how math works, and the pollsters are beclowning themselves (or selling an agenda) when they try to pass this stuff off as legitimate.
Moreover, the very idea of trusting a national poll when more than half of the electorate has already voted is, quite frankly, absurd. This is especially so when two of the candidates have a demonstrated base of support that has a strong regional component. It is painfully clear that on both the Republican and Democrat side, the front runners have been racking up their highest level of support in the deep south. Those states, however, have already voted, and the map going forward looks completely different.
I would be at least interested to see what the national polling would say for the region(s) of the country that have not voted yet, but until then, these worthless data points are pretty much what we have.