Members of the cleaning crew sweep the floor at the University of Illinois at Chicago Pavilion after Donald Trump canceled his rally on Friday. (Reuters/Kamil Krzaczynski)

The remarkable rise of Donald Trump as GOP front-runner continues to confound the pundits and much of official Washington. Playing off Churchill’s famous quip about the Russians, Trumpism remains for many a “Jersey Shore” riddle wrapped in a “Duck Dynasty” mystery inside a “Kardashians” enigma.

Mostly missing from the competing explanations for Trump’s emergence have been insights from prospect theory, an influential model of the choices people make under conditions of uncertainty. Originally advanced by Daniel Kahneman and Amos Tversky in a 1979 article published in Econometrica, prospect theory has since developed into a complex scholarly paradigm that has been applied to everything from financial investments to negotiations between nations. But much of the underlying intuition can be captured with a simple word game.

Imagine that you are confronted by two sets of choices. In the first, you are asked to choose between receiving a \$5,000 check for sure, or having a 50 percent probability of either receiving \$10,000 or nothing at all. Of course, the presumed payoff from these two options is identical because the expected value of the lottery (0.5 multiplied by \$10,000 plus 0.5 multiplied by zero) is precisely the same as the value of the sure thing (\$5,000).

Now consider a second set of choices. But here, the decision is between losing \$5,000 for sure, or having a 50 percent probability of losing either \$10,000 or nothing at all. Once again the expected return from the two options is the same because the value of the lottery (0.5 multiplied by -\$10,000 plus 0.5 multiplied by zero) is identical to the value of the sure thing (-\$5,000).

The standard approach in economics and political science is called “expected utility theory.” In this model, you should prefer the first set of choices to the second, of course, because a gain is always better than a loss. But within each set of choices, this theory goes, people should be indifferent between the lottery and the sure thing because the expected value from each is the same. Or at least, they would be if human beings were calculating, hyperlogical Vulcans.

Prospect theory, by contrast, takes into account how ordinary folks actually respond to uncertainty. Its core tenet is that in the first set of choices, which offers the potential for gain, people will tend to choose the sure thing, while in the second set of choices, which offers the potential for loss, people will go for the lottery.

That’s because, in the real world, people’s attitudes about risk depend on what is at stake. When confronting possible gains they tend to play it safe, and when confronting possible losses, they prefer to take chances and hopefully minimize the damage. Or as academics put it, in a “gain frame” people are risk averse, and in a “loss frame,” they accept risk.

This behavioral pattern has been confirmed in numerous empirical studies conducted by social scientists. Many scholars believe that it may be hard-wired into us via evolution. In Darwinian theory, a “loss frame” might have involved a choice between slow and certain starvation, or the risk of sudden death (during a hunt, in a raid, or on a trek, say) versus the possibility of thriving. With conditions like that, the risk-taker would be more likely to live and have children than the one who stayed home. However, when resources are plentiful, someone who chooses a modest but stable supply of food might be more likely to send her genes into the future than a counterpart who risks it all on the same hunt, trek, or raid.

So what does this have to do with Donald Trump and the 2016 presidential campaign?

Mr. Trump is nothing if not unpredictable, a walking, talking, political lottery. One week he supports taking out the families of accused terrorists; the next he pledges to obey international law. He favors the massive tax cuts of Republican orthodoxy, but also promises to protect major entitlement programs and supports the individual mandate for health insurance. On abortion, he has flipped from pro-choice to pro-life, but still endorses the core mission of Planned Parenthood. Over the years, he has supported and funded both Republican and Democratic candidates for office.

No one, in other words, can easily predict what he would actually do as president.

Trump’s volatility may come from a lack of philosophical or ideological moorings, but it also appears to be conscious and strategic. Recently, TV host Bill O’Reilly asked Trump about his plans, and the billionaire responded, “I want to do what’s right. I want to be unpredictable. … The voters want unpredictability.”

Why might Trump supporters prefer a political lottery to more predictable Republican Party figures like Sens. Marco Rubio and Ted Cruz or former Florida governor Jeb Bush?

According to surveys and demographic trends, Trump’s voters believe they are in a socioeconomic equivalent of a loss frame. White middle- and working-class voters have suffered stagnant wages for decades, and the 2008 financial crisis erased much of the wealth they had accrued through home equity.

What’s more, by the 2030s, white people will no longer be the demographic majority in the United States. The coming diversity explosion is already visible in America’s public schools, most major cities, popular culture and the White House.

No wonder large subsets of the population – often white, less educated and economically lagging – want to roll the electoral dice, opt for risk and vote for Trump.

Much has been made of Trump’s alleged misogyny, xenophobia and racial insensitivity on the campaign trail. Trump may be insulting female journalists and Mexican immigrants, and encouraging the Klan, as ways to articulate social views and policy preferences shared by many voters. Or it may be something a little different. Richard Fenno, the preeminent scholarly observer of representation, emphasizes that candidates’ public utterances are part of their overall “presentations of self” — that is, the stylistic images that they manipulate to shape how voters see them.

From this perspective, Trump’s apparent spontaneity and periodic outrageousness reinforces his image as unpredictable, a different kind of politician – a kind of lottery. And so when establishment figures like Mitt Romney criticize Trump as unreliable or not mainstream, Trump’s ties to his core supporters are strengthened, not weakened.

For Trump, unpredictability may have electoral value for an important but still limited set of voters. One of the challenges in applying prospect theory to real world behavior comes in trying to identify the differences between gain frames and loss frames. In 2016, how many voters truly believe that the future is bleak and that our best days are behind us? Many, to be sure. Maybe even a significant voting bloc. But perhaps not enough to win in November.

Moreover, if Trump’s popularity continues into the fall, voters will learn a lot more about him. Campaigning in a general election would force him to clarify positions, to put out fires, to become more predictable. In the standard political science view, general elections, unlike primaries, push candidates toward the ideological center. For nominee Trump, the main impact actually might be to push him toward clarity. If so, he’d lose his attractiveness for pessimistic voters looking to roll the political dice.

C. Lawrence Evans is the Newton family professor of government at the College of William and Mary.