You had to know it would only be a matter of time. As soon as the Internet started offering online gambling sites, and states like Kansas began legalizing fantasy sports betting, it was only a matter of time before the odds makers in Vegas would figure out a way to tap into the political market.

The truth of the matter is that gambling on the outcome of elections is nothing new. But what is new is the ability to take data from people who are willing to lay down money on the outcome of an election and turn that into a form of crowdsourcing that yields a barometer that appears to be at least as accurate as any other prediction model floating around these days.

So, if you’re wondering who the odds-on favorite is to win the White House in November, all you have to do is Google the term “presidential prediction markets” and you get your answer: Democrats stand a 71 percent chance of winning the general election, and right now, Hillary Clinton has roughly a 90 percent chance of being the Democratic nominee.

Put another way, Donald Trump is viewed as having an 80 percent chance of being the Republican nominee, giving Republicans only a 29 percent chance of winning the White House.

That’s the current (as of this writing) assessment from the website PredictWise, founded by Microsoft Research economist David Rothschild, which aggregates data from a number of different sites.

One of the sites PredictWise uses goes by a similar name, PredictIt, which gives users the chance to buy, sell and trade shares in the outcome of an electoral event, such as the outcome of a primary, a nomination, or the general election. So, for example, ‘Candidate A wins the nomination’ would be an event, and traders will speculate on what the percentage chance is of a particular outcome of that event, either “yes” or “no.” Percentages are then translated into U.S. cents. The sum total of “yes” and “no” bids add up to $1.

As of Sunday afternoon, people willing to bet money on a “Clinton-Yes” outcome of the general election were buying at 61 cents. People betting on a “Clinton-No” outcome were selling for 39 cents.

To see how accurate that model is, we only have to look back at some recent primaries. Leading up to the March 1 Super Tuesday primaries, PredictIt was forecasting that Trump would win in 10 states and lose only in Texas to that state’s favorite son Sen. Ted Cruz. And it showed Florida Sen. Marco Rubio would finish second in the Minnesota caucuses.

PredictIt got every one of those right, except Minnesota, where Rubio eked out a win.

That kind of information can be fun and entertaining, depending on which side of the race you’re on, as long as you take it with a grain of salt. There’s still a whole lot of race left, and a scandal here, or a misstep there still could greatly affect the outcome. But these prediction markets are also grabbing serious attention from academic circles.

Kansas University political science professor Burdett Loomis called attention to them during a recent talk he gave to the Douglas County Democratic Party. When I emailed him later to get more information, he suggested the Iowa Electronic Markets, one of the oldest prediction markets around, and one originally set up by academics.

IEM has been around for a few election cycles now, and in 2008 it outperformed all the major public opinion polls for accurately predicting the outcome of the election.

IEM’s model, which looks a lot like commodity futures trading, offers two different types of “contracts,” or estimates of the outcome: “vote shares,” or the percentage of the total popular vote either party will get; and “winner-take-all,” which predicts the outcome, regardless of point spread.

At last check, contracts for a Democratic popular vote win in November were trading at 59.8 cents, compared with 40.5 cents for a Republican win.

In the winner-take-all contracts, Democrats were up 71 cents to 31 cents over Republicans.

Prediction markets are essentially a variation on a theme that has been developing in the field of public opinion polling for some time. Originally, pollsters would ask (and still do ask), “Who do you intend to vote for in the upcoming election?” That would give an accurate snapshot in time of where the race stood at that particular moment, but it often had little predictive value because people change their minds.

More recently, pollsters have started asking a different question: “Regardless of who you intend to vote for, who do you think will win the race?” That question turns out to have much more predictive value because it acknowledges the tendency of people, in the end, to gravitate toward the norm. In other words, most people want to be on the winning side.

The new, Web-based prediction markets are another step in that direction because they basically ask, “Regardless of your own opinion, if you were to put actual money on the line, where would you place your bets?” The theory is, people start getting a lot more honest when they have money at stake, even if it’s only a couple of bucks.

Bear in mind, however, the old saying in Las Vegas: “All those glitzy casinos weren’t built with money people won at the slot machines.” Gamblers often have a propensity for losing, so these current market-based predictions should be taken for what they are: predictions. It’ll take until November to find out how accurate they are.