Why the Polls Got It So Wrong (and How a Frenchman Named Théo Saw It Coming)

When I woke up on Wednesday, one question was bouncing around my head: How did the polls get it so wrong?

I mean, we had more data than ever before:

• Voting data from two elections with Trump and one with Harris.

• Social media, with 24/7 sentiment analysis from TikTok and X.

• Actual face-to-face conversations this time around (unlike the lockdown days of 2020) to help us gauge what people were thinking.

And yet, the polls missed the mark. Badly.

It wasn’t just the polls either—I had plenty of conversations with friends from both sides of the political spectrum, and no one had a clue which way the election would go. Heck, five days ago, I thought Trump was toast. By 10 PM Tuesday, he was practically planning his victory lap.

So, how did everyone get it so wrong?

Here’s a theory: we had too much information. Instead of making us smarter, it made us more confused. If you spent 30 minutes on Twitter, you’d think Trump was about to win in a landslide. But hop over to Threads, and you’d be convinced Harris had it in the bag.

Chat with a student at Berkeley, and they'd struggle to name a single Trump supporter on campus. Meanwhile, talk to a coal miner in eastern Kentucky, and you'd hear a completely different story.

And when you ask someone who they’re voting for, there’s always a chance they’ll tell you what they think they should say, not what they actually believe. You could Google all the stats, maps, and migration patterns you want, but with so many ways to slice the data, it’s no wonder everyone was confused.

Well, almost everyone…

Enter "Théo," the Frenchman Who Saw it Coming

One of the more interesting stories to come out of this election was about a mysterious French trader who goes by “Théo.” This guy, who bet around $30 million on Trump’s victory in prediction markets, had no political agenda—he just believed the polls were underestimating Trump, like they did in 2016 and 2020.

Théo wasn’t just throwing his life savings into the wind, though. He took a fresh angle on the polling problem. His theory? Polls were missing the “shy Trump voter.” These voters either didn’t want to admit they were supporting Trump or just refused to participate in polls altogether.

To solve this, Théo suggested an ingenious solution: instead of asking people who they were voting for, ask them who they think their neighbours are voting for. People might not reveal their own preferences, but they’ll often spill the beans when talking about others.

Théo even commissioned his own surveys to test this theory, and the results were apparently mind-blowing (though he didn’t share them—something about a deal with the pollster). He argued that U.S. pollsters should use this “neighbour” method to avoid another embarrassing miss in future elections.

The Real Lesson: Ask the Right Questions

Théo’s big win wasn’t just luck—it was because he asked the right question. In a world drowning in information, it’s easy to find a stat or story to back up whatever you already believe. But if you want solid answers, you’ve got to ask the right questions.

Pollsters made the mistake of asking people who they were voting for directly, and, surprise, they didn’t always get honest answers. Théo’s approach—asking about the neighbors—cut through the noise and got closer to the truth.

This doesn’t just apply to elections. Whether you’re trying to figure out if a stock is a good buy (just Google “Is Tesla undervalued?” and then “Is Tesla overvalued?” for a fun exercise), deciding on a career move, or diagnosing a headache (three WebMD searches and you’re convinced it’s an aneurysm), our world is full of conflicting information.

And it's only going to get worse. With AI tools cranking out content at lightning speed, the amount of noise is about to hit new levels.

The key to surviving this data deluge? Distillation—figuring out which information actually matters and tossing the rest. But before you can do that, you need to start with the right questions. Ask better questions, and the good info will come to you.

Remember: it’s not about having all the data—it’s about using it wisely.