There's an axiom in polling that goes back to 1936. If you expect to win the presidency, you'd better be polling well as September begins.

Nobody really 'likes' to admit that polling is as good of a science as it is; it tends to undermine our sense of control and autonomy -- that less than 1,000 people polled can represent the population of 146 million or so voters in the United States.

But that's the nature of statistics -- as well as the numbers being criticized for any number of reasons.

But the tea leaves of political fortune telling are starting to settle, with bad news for third-party candidates, in particular.

In possibly one of the most inclusive polls to date, Gallup offers a lot of insight into the demographics of the third-party candidates --  most damaging in that they just can't seem to get traction outside of the under-30 voters, possibly the 'least critical' of all the demographics in American politics.

Love it or hate it, the Clinton campaign has done a good job neutralizing the effects of 'The Bern' on swaying support to the Johnson or Stein campaigns.

With Johnson drawing from the Republican base at a two-fold rate, compared to the Democratic base (22% to 11%, respectively), Johnson's gamble of attracting Sanders voters with the libertarian social ideology has seemed to fail.

---

But August is also the time when the news outlets start making their solid, but still conservative projections on the electoral map.

In 2012, CNN projected Obama as clearly ahead (but not outright winning) in August, but he then won the toss-ups with a clean sweep to overwhelmingly win the Electoral College.

Even the more conservative Fox News begrudgingly agreed, though still hopeful of a strong Romney surge.

The point being that it's not just the liberal media trying to sway voters with 'cooked up' polling numbers -- this has been a pretty accurate science.

But this August, it seems that all of the networks and polling companies are agreeing that Clinton has an overwhelming lead in the Electoral College, and a strong 10+ point lead in the national polling.

The interesting part, however, is that this new data completely takes the so-called 'Holy Trinity' of battleground states (Ohio, Florida, and Pennsylvania) out of the equation -- with Clinton easily winning only capturing Pennsylvania.

But this new data is going to redouble the Republican efforts to proclaim Johnson the 'spoiler,' with traditionally deep-red Georgia being counted as a true toss-up due to the plurality in polling.

Clinton, trying to exploit the plurality weakness, has opened an office in Utah, where no Democrat has won in the past 50 years -- crazy things happen in a plurality.

The next 11 weeks will not be kind to the third-party candidates, with candidates employing all of the traditional tactics to neutralize them.

---

So are the fortune-telling political tea leaves set for good -- or is there something that can be done about it?

We're sadly a nation that has turned non-voting into a national institution, with almost half of the eligible population choosing not to vote on election day.

Pollsters know this, and structure their polls around 'likely-voters' instead of just eligible ones.

Independents are too often stuck in the mode of trying to draw away the votes from the major parties, when we already have a large pool of untapped potential to work with.

If independents ever want to have a real say in the election process, we need to do a much better job of 'getting out the vote' for our cause among non-voters than the major parties have ever done.

Because that's part of the winning combination for independents -- fielding candidates at all levels of government, ideas for positive change, and enfranchising voters who have long given up on the process.

When all three of those have been done, the current polling models will be completely shattered -- and political fortune telling will just have to wait for election night.