From strictly a horse-race perspective, here are a few of my near-final thoughts on the 2016 presidential election.

First, as with so many other presidential elections, the basic fact that produces an outcome is turn-out. My gut tells me a lower turn-out favors Trump, a higher favors Clinton. But there’s another pair of numbers that I think will matter, too.

I’m referring to the pairing of Brexit and ground-game. The Brexit refers, of course, to the British vote in the summer about whether to leave or remain in the European Union. The “leave” side of the Brexit issue bears close resemblance to many of the pro-Trump voters in the US. The Brexit result showed that roughly 4% of the final vote hadn’t been identified by pollsters prior to voting day. That number feels about right to me as I think about Trump; anything less than 4% can be overcome by the unusual coalition that Trump has constructed. If he is within 4% in any given state, the Brexit story means that Trump is within striking distance of winning that particular state.

Conversely, the Obama voter turn-out operation in 2012 succeeded in overcoming the predictions of the most well-regarded pollsters. His ability to get unenthusiastic voters to the polls in 2012 produced roughly a 4% gain in the final tally. That’s what accounted for Obama’s defeat of Romney. Similarly, I think the same thing holds true for Clinton n 2016. A vastly superior ground-game (as they call it) has the potential to collect 4% more voters than would otherwise have proven true.

So, that leaves us with a pair of competing 4%s in addition to the basic fact of turn-out. As I count it out, that means 1, 2, 3.