DAVID MARCUS: Experts crowned El-Sayed — then Michigan voters turned his win into a near-disaster

A guy walks into an election with a double-digit polling lead and walks out having barely survived. That's not a rounding error, that's a signal flare. Abdul El-Sayed was supposed to coast. Instead Haley Stevens nearly ran the table on him, and the same class of people who told us the race was decided weeks ago are now trying to explain why it wasn't.

This is the part of the story that deserves more attention than it's getting. Polling in Democratic primaries has been wrong before, sure, but a near-total collapse of a double-digit lead isn't just noise. It means the people paying attention to actual voters, actual door-knocking, actual retail politics, understood something the modeling didn't. Stevens apparently ran like someone who knew she was behind. El-Sayed, and the experts backing him, ran like someone who'd already won.

There's a broader lesson here that Michigan Democrats should sit with, and it has nothing to do with us. When your own primary electorate nearly rejects the candidate anointed by the credentialed class, that's a trust problem, not a turnout problem. Voters increasingly don't want to be told who's inevitable. They want to be asked what they actually think, and then believed when they answer.

We'd add one more thing. This kind of gap between confident forecasting and messy reality keeps showing up, cycle after cycle, race after race. At some point it stops being an anomaly and starts being the pattern. Maybe it's time the people making these predictions admit that.