Was 2026 a vindication for polling?

Many have suggested so, because the predictions for a Blue Wave were accurate.

In fact, however, WaPo points out that the polls were even less accurate than in 2022 or 2024. That fact was masked by the facts that (1) they erred in the opposite direction compared to recent history, (2) nobody (except nerds) cares about the margin as long as the polls get the winner right.

The 2026 polls consistently underestimated Democrat margins by four to nine points. That’s a dramatic level of error. The Dems overperformed the polls in every key Senate race except Alaska. The polls were especially far off in Texas and North Carolina.

* As far as I can see, only one polling or analysis organization correctly called Mississippi a toss-up. That was a newbie group called On Point.

* The prediction markets fared no better than the polls in Mississippi. Polymarket gave the Democrat a 16% chance on Monday night.

Speaking of prediction markets …

The big winner on election night, according to Kalshi? Jon Ossoff.

It’s interesting that the speculators felt: (1) that the election hurt JD Vance, but not Marco Rubio; (2) that the blue wave reduced Newson’s odds, but actually helped AOC.