
I got a lot right about the 2026 midterms, and I feel like many Slow Boring positions have been vindicated — most notably the argument that winning a Senate majority is possible if and only if Democrats make some moves to the center on cultural issues.
We also called some small, specific things correctly, such as that nominating Matt Dunlap would cost Democrats ME-2 and that Josh Turek was the right candidate for the Iowa Senate race. There’s been an incredible amount of beefing on the internet about the fact that Abdul El-Sayed won comfortably in Michigan while putting up a mid-sized underperformance, but this is exactly what I said all along would happen.
That said, it’s my job to comment on issues in the news despite uncertainty, and when you do that, you get things wrong.
My view is that the only way to improve your judgment is to try to be honest with yourself about errors. There are prominent members of the media who I see out there every day full of bluster claiming to have been right about everything. To me, people like that are basically Elizabeth Holmes telling Nathan Fielder she’s never lied — it’s a huge red flag that means your heart and mind are completely closed and you’re not learning anything.
Obviously, in a sense, basically everyone was “wrong,” and the final national popular vote was about D+13 rather than D+9 as the polling averages said.
What I would say is that even though this was a decent-sized polling error, it was still a victory for the general concept of trusting the polls. Not because polling is precisely accurate but because errors happen in both directions — they don’t always underestimate Republicans — and it’s not possible to predict in advance which way they will be off. Nobody has come up with a reliable way of doing better than looking at polling averages. What I want to focus on today, though, are more interesting mistakes: things that destabilize my worldview a little and will cause me to adjust my takes going forward.
This focuses mostly on the role of turnout, the success of a particular way of ducking inconvenient issues, and the ways that these two things maybe interacted in putting a few races on the table that didn’t seem winnable to me.
Turnout seems to have mattered a lot
Everyone is aware that turnout matters in politics. But my sense is that after almost every election, many people massively overstate its significance in explaining the outcome. The facts are not currently 100 percent known yet, and we’ll get a better picture in the weeks and months to come, but that seems to have not been the case this year.

