������Ƶ

Conventional wisdom says that Gov. David Ige will lose his re-election bid.

Not so fast. You need to rethink, or, better yet, to think differently.

As far as election predictions are concerned, we all need to remember the key mistakes analysts made in predicting the outcome of the 2016 presidential election. Those mistakes offer two lessons:

First, there is an enormous difference between a high probability and a sure thing.

Second, the more you think unequivocally, the less prepared you are to accommodate exceptions and anomalies.

The analysts’ 2016 mistakes (this includes me) were more psychological than mathematical.

In fact they made the same hardwired brain-based errors that ordinary folks regularly make in everyday life.

They displayed confirmation bias, succumbed to the non-wisdom of the crowd, and bought into a conventional wisdom that missed cues regarding Trump’s strength that were (barely) hidden in plain sight.

How and why did that happen, and how does this apply to the governor’s race?

When my wife was pregnant with our first child, she bought a book on birth defects. The book described a malady, then stated the probabilities of a child being born with it.

These odds reassured me, so I interpreted them optimistically — as few as a one in 10,000 chance of your newborn having a particular defect.

The numbers did not reassure my wife. They scared her, so she interpreted the probability as “as many ����”.

Hers was not a mistake or a mathematical error. It was an interpretation of a statistic based on the narratives she carried in her head, her own anxieties, visions of a terribly challenged infant that were so strong they dominated any discussion about odds.

She exaggerated one interpretation and pretty much ignored the other, as I did in the opposite direction.

That is what most of the best political analysts did regarding Trump.

It wasn’t the data. As a rule, probability models are excellent sources of prediction.

So if a reputable 2016 probability model said that Hillary Clinton had a 70 perent chance of winning, those were well-informed odds.

A couple of boo-boos messed things up. People had to understand how to interpret probabilities. A 70 percent probability does not mean that a candidate will get 70 percent of the vote, which was a very common misperception.

For the analysts, though, the problem was not lack of grammar school math skills. The culprits were those visions embedded in their hearts and minds.

That gets us back to the birth defect story. If you are so wrapped up with one story and one interpretation, you will ignore others.

The media generally ignored this simple truth: A 70 percent probability means that for every 10 elections, Trump would win three of them.

Analysts rolled right over the fact that, as Nate Silver put it in his of the coverage, a 30 percent or so probability is within the “cone of uncertainty.”

The story in the election analysts’ head was that Trump’s campaign was unconventional, which it was for sure. But they went on to assume incorrectly that unconventionality could not be effective.

To these experts, Trump’s innovations indicated amateurism and incompetence, sure signs that Trump could not win.

Consequently that 30 percent chance that he would win got left in the punditry dust.

So, in 2018 don’t do that.

What should you do?

To those of you who are not political junkies but still reading this, stick with me. What I am going to offer is not just a political lesson.

It’s a life lesson. Really.

First, think in terms of probabilities, not pound-on-the-table absolutes.

That’s how the people who are on all kinds of issues — they are almost never experts, by the way–act.

You don’t need statistical data to do this. In fact you couldn’t find such information even if you wanted to. Hawaii is a data desert.

Even without these aids, you should think in terms of probabilities —“I give Colleen Hanabusa a 55 percent chance of winning” — and you should regularly re-examine these odds when you find new information.

There is a bunch of stuff to consider. The one most frequently ignored in the governor’s race is incumbency. Nationally about 80 percent of governors running for re-election win. Historically an incumbent governor almost never loses in a primary.

Yes, Ige was an exception to that rule in his 2014 victory over incumbent Neil Abercrombie, and maybe Hanabusa is a strong enough candidate to pull it off as well.

But incumbency is generally such a fundamental factor that it should not be unequivocally dismissed.

Here are some other things:

• Approval ratings: Ige’s have been fairly low but quite stable. Hanabusa’s are higher but not by much.

• Crucial unanticipated events: For Ige the fake nuclear disaster was a personal disaster. The Big Island volcanic eruption could turn into an opportunity.

• General uncertainty: As Civil Beat’s pollster Matt Fitch recently said, right now there is no good explanation why Ige did considerably better in a second gubernatorial poll than in the first.

Use any information you think is relevant, including the wisdom of others. But be careful.

Pundits frequently rely on the same old sources and carry in their heads the same old perceptions about elections and politics. That has its advantages.

But it also has, as we saw in 2016, its limitations, like the inability to understand anomalies or take them seriously.

Keep in mind that your passions and narratives, which are powerful drivers of your political values, are unruly and undependable when it comes to understanding the full picture.

So when you say something could not possibly happen, like women supporting Trump after the release of his “pussy” conversation or people voting for Ige after the “Gong Show”-like North Korean missile debacle, remember two things.

One, you are thinking unequivocally in a world filled with contingencies.

Two, as the birth defect book story showed, your predictions are affected by your predilections.

And that’s not just true about politics. It’s life.

Thoughts on this or any other story? Write a Letter to the Editor. Send to news@civilbeat.org and put Letter in the subject line. 200 words max. You need to use your name and city and include a contact phone for verification purposes. And you can still comment on stories on .