As early as last November, we felt compelled to rebut an argument that was then (and is now) in vogue: Hillary Clinton led in 2008, too. The rebuttal is simple. Hillary's lead in 2016 is much larger than the lead she enjoyed in 2008.

There is an equally important point that's worth making, now that the conversation has shifted, at times dramatically, in the other direction. Our own Jennifer Rubin asks, "Is GOP 2016 becoming a two-man race?" referring to the men Jeb Bush and Scott Walker, both of whom have seen bursts in recent polling. To answer Rubin's question: Maybe! But the lesson of past polling is that picking winners from crowded fields 18 months out is a recipe for failure.

We'll start with an exception: The GOP field in 2012. There was Mitt Romney, floating along on the surface in early 2011. His is the dark red line on the chart; on all of the charts that follow, it's the dark line that denotes the winner.

A helpful visual cue when we get to the 2008 race with which we began. There's Hillary, floating along on the surface, with a shark named Barack swimming just underneath her.

Notice, first, how little space there was between Clinton and Obama. The scale on those first two graphs isn't the same, and Clinton led by more in June 2007 than Romney did in June 2011. But there's a gap of only about 11 points by June, far from the 50-point average she sees now.

But back to the point of this article. Here was the Republican field in the first half of 2007.

And here was the Democratic field for 2004 in the first half of 2003. Notice that the horizontal axis isn't to scale, but that doesn't muddy the point very much.

It's only when we get back to races with a prohibitive favorite, that led widely, that the polling during the first half of the year prior to the election clearly matches the end result. As in 2000 and 1996.

(We had to include that surge by Newt Gingrich for the sake of what-might-have-been.)

When we get back to 1992, polling is less frequent. But a poll from March 1991 shows which people were leading the Democratic field at that point. Not mentioned: Some dude named Clinton.

Okay, you say. I get it. But look at that first Romney chart! A big lead now can mean that you win! Yes, that's true. The recent past doesn't mean that front-runners won't win — just that it's not a given. These graphs, though, only look at the first half of the year prior. There were a lot of changes in the second halves of those years, too, like the late 2003 surge of former Vermont governor Howard Dean.

Or the trouble that Mitt Romney faced in 2012, which we revisited last week.

If prior-year polling were a reliable guide to who would win, the 2004 Lieberman-Bush race would have made way for the Clinton-Giuliani contest of 2008. To the best of our knowledge, that didn't happen.