As we go through the next few years towards the 2016 presidential election, I intend to often remind you of this. It's the final Gallup Poll on the 2012 contest and it shows Romney beating Obama by a point. In reality, of course, Obama got 51.1% of the popular vote and Romney got 47.2% so instead of Romney over Obama by 1, it was Obama over Romney by 4.
Keep this in mind: This was what Gallup was predicting the day before Election Day. It stands to reason that the farther out you are from the actual voting, the less reliable the polling will be.
I am not suggesting that all polls are inaccurate, particularly in the aggregate. Nate Silver famously nailed the vote by analyzing all the polls with a skeptical eye. I'm just saying that when you see Gallup say that Rand Paul would beat Hillary Clinton by X%, remember: These guys couldn't get it right the day before. A week before stands an even worse chance of being right and a month stands less…and as for three years…
And by the way: When the Gallup folks issued the above Final Prediction, they said it had a margin of error of plus or minus two points. In other words, they said Romney was going to win but would have claimed the poll was accurate if Obama had won by a point.