Will pollsters adjust their likely voter models based on yesterday's election?

[ColinC]

Clearly turnout cannot be expected to be the same levels as a normal midterm. Data for progress' last LV poll still seems like it might have a poor LV model. Democrats seem to be pretty solid on all of the RV polls taken, but LVs is where they struggle the most. Will do you think those models will change before the November election?

I think they were also substantially off in 2020 when they seemed to assume larger turnout for Democrats than actually as the case.

[mopinko]

1. they have to. these are the swing voters who decide everything.

if they dont, they should go out of business.

[ColinC]

3. Yeah the last Data for progress poll was taken from the 1st to 3rd, I think.

So we might see some dramatically more favorable polling for Democrats in the near future. I'm almost worried about how that could suppress or help turnout in November.

[PortTack]

2. Pollsters...I think I'll become one. No better than they did in this primary and the last 3 election

Cycles. Most, not everyone but most including 538, RCP are worthless.

[ColinC]

4. Lol

Yeah. I mean probability/ data science, is a science and all, but their predictive qualities are super overrated.

[PortTack]

5. ☝️☝️☝️☝️☝️