Some truths are eternal, no matter who said them. Whether Niels Bohr, the Nobel Prize-winning physicist and atomic pioneer, or Yogi Berra of the New York Yankees first uttered these words, they are true: “Predictions are very difficult, especially when it comes to the future.”

This week’s US presidential election seems so important that it’s only natural for people to try to predict the outcome or find an indicator to latch on to. For centuries, political forecasting markets have emerged to meet this need, and it has happened again. In recent weeks, several new projects have been launched that enable political futures trading. The movements in the various markets were strange at best and led to accusations of manipulation. And just as it is always difficult to predict the future, it also seems to be a fact that markets are overshooting. That’s exactly what just happened.

Starting in early October, all major markets saw a steady increase in Donald Trump’s chances of winning. This was led by the offshore company Polymarket, with the help of a series of large purchases by a Trump “whale” (and a brief blip caused by what appeared to be a Kamala Harris whale). Then came Trump’s rally at Madison Square Garden a week ago. That changed the narrative. The PredictIt market, where political junkies hold limited positions, began to falter. Polymarket, increasingly optimistic about Trump, followed suit.

Then on Saturday came the shocking Des Moines Register poll from one of the country’s most reliable pollsters that suggested Kamala Harris was leading by three percentage points in the heartland state of Iowa, which has increasingly voted Republican in recent elections. If true, it means Harris is winning other states where expectations appeared to be more stringent; Proponents dared to speak of an impending landslide. As with any outlier, the survey should be taken with a grain of salt, but the predicted market reaction was dramatic: