You’ve heard the term, “click bait.” It’s those wonderful, tempting links all over websites like “The top-10 sexiest actors,” or “You’ll never believe what she looks like now.”

For political junkies, our click bait is the latest poll numbers. We mostly use them to reinforce our already held beliefs, because nothing is as addictive as a hit of confirmation bias.

The latest polling by the Colorado Polling Institute shows Gov. Jared Polis’ favorability ratings are sinking underwater (45% favorable). But not as badly as U.S. Sens. John Hickenlooper (43%) and Michael Bennett (41%), or Attorney General Phil Weiser (24%).

So, who’s going to win the elections next year? Thank God we have polling to let us know — like any of us have faith in polls.

Every election season, America pretends to be shocked — shocked! — when the polls turn out to be about as reliable as a fortune cookie written by a stoned political science major. Pollsters blame turnout models, politicians blame the media, and the public blames politicians, pollsters and the media.

Polling is notoriously imprecise — an art, not a science. Each pollster has his own secret sauce, a special way to calculate his findings, or a better way to find a more representative sample of voters. But voters fib. They fib to themselves and to pollsters.

Some things you just don’t wanna tell a stranger who calls during dinner. Donald Trump famously under-polled because folks didn’t want to admit to some unknown caller they voted for him. It’s way too embarrassing. It’s like the women who have gone out on dates with me. They tell no one for fear of ridicule and shame.

Polls, for all their usefulness, are simply not built to predict the future. They measure what people say, not what people actually do.

A much more accurate, and occasionally profitable, tool has been gaining steam, ignored by pundits and despised by consultants and pollsters: prediction markets. That’s a fancy way of saying “betting on election outcomes.”

Betting platforms like Kalshi (Kalshi.com), PredictIt (PredictIt.org) and Polymarket (PolyMarket.com) are basically sports-betting apps that include elections and future events like “Will the Federal Reserve cut interest rates?” The odds change depending on how people place their bets.

As I write this, Kalshi gamblers give 95% odds a Democrat will be Colorado’s next governor, 86% a Dem will be the next state AG, and 97% our next U.S. senator is going to be… wait for it… a Democrat (seems about right to me).

Oddly Kalshi only gives Hickenlooper an 80% chance of winning his Dem primary. Polymarket has him at only 66%. Are the markets saying, “tired and long in the tooth”?

To be clear, the apps don’t make these predictions. The individuals betting their hard-earned money do. And like any gambler, they are very self-interested.

These markets don’t get invited onto cable news panels or hold press conferences. Unlike pollsters, they have no ego. But they do make people put money where their mouth is.

A strange thing happens when people risk their own money: they suddenly get very real.

Because their money’s on the line, participants search for information, corroborate sources and absorb every rumor and scandal floating around a campaign. You’ve seen people do something similar on sports betting apps, trying to gather objectively true data.

If new information emerges, markets adjust within minutes — not days. Unlike polls, which update in slow, lumbering costly steps, prediction markets behave like the stock market. New information gets “baked in” almost immediately. Stock prices aren’t about what a stock is worth, they’re about what investors think the stock will be worth. Prediction markets do the same.

Over the last two decades, prediction markets have routinely outperformed polling averages, especially in the final stretch before Election Day. It’s not magic; it’s economics.

Since newspapers and TV news stations no longer pay for pollsters, if you’re trying to get a sense of what’s likely to happen in next year’s elections, turn away from pundits and tune into betting sites for truly objective predictions.

A site called Manifold.Markets currently has state Sen. Barb Kirkmeyer in the lead for the Republican nomination for governor and shows Michael Bennet likely to win the Democrat nomination at a whopping 71%.

If democracy is going to be a gamble, we might as well listen to the people who actually place bets.

Jon Caldara is president of the Independence Institute in Denver and hosts “The Devil’s Advocate with Jon Caldara” on Colorado Public Television Channel 12. His column appears Sundays in Colorado Politics.