![salt shaker - no salt EDIT](https://floridapolitics.com/wp-content/uploads/2015/07/salt-shaker-no-salt-EDIT-1024x1024.jpg)
The poll aggregating website, Real Clear Politics, shows a half-dozen polls released in the past three days alone. Additionally, Associated Industries of Florida has released a series of polls and a detailed analysis of their findings.
And if you are paying even the slightest bit of attention, you are probably scratching your head over the differences – the WIDE differences – in Trump’s lead. Some polls have his lead over Rubio at more than 20 points, while others have the race down to single digits.
So why the differences?
For the most part, polls with a looser sampling model tend to give Trump a larger lead than those polls using techniques which increase the likelihood that the respondent is an actual voter and has a verified history of voting in primaries. To put it in very simple terms: tighter screen = tighter race, looser screen = larger lead for Trump.
So who is right?
If you are a super-geek and read Ryan Tyson’s analysis closely you probably know the answer. Trump seems to be doing well (and I know this is a shock) among non-traditional voters or even non-voters. So polls that pull in self-identifying frequent GOP voters (“you betcha I’m gonna vote” and “yeah, sure, I’m a Republican”), even if they are not registered Republicans, tend to give Trump a lift.
This isn’t an editorial comment, it’s an observation of the data and Tyson explained it perfectly – in fact, his analysis is the first we have ever seen where he gives his own methodology a grain of salt. How cool was that?
But here’s where that gets interesting. As of Thursday, about 17% of those who have cast ballots have not voted in a recent primary. While there are always first time voters in each cycle, the fact that approximately one in six voters are the above-referenced non-traditional voters might suggest that the looser polls may be closer to the mark than we would expect.
Not so fast.
Trump appears to be doing well among that group (at around 40%) but that is just not enough to pull his numbers by that large of a margin – keep in mind that 60% of those new voters are picking someone other than Trump. It likely only impacts his actual lead by about a point or two.
So what’s the bottom line?
With about half of the election already in the can and with Trump solidly and consistently in the lead, irrespective of the methodology of the poll, the safe odds are that come Wednesday morning most of the conversation will center on Marco Rubio post-mortems.
Oh, and about the salt shakers.
Take all of these polls – ALL OF THEM – with a grain of salt. Nobody knows what the final turnout will look like as this election is unlike any we have ever witnessed. In the final analysis, some will be correct and some will be wrong. But if you like to bet, the safe money is on another bad night for the junior Senator from Florida.
At least that’s what the polls say.