Post

Obverse is true for Republicans. In any event, you know they would rather be in our shoes.

Not Elliott, but in our sims a single Senate race's margin swings about 7 points from night to night, and the part all races share is about 4. So close races tend to break the same way, just not in lockstep.

I've noticed with the lack of any polling, the senate odds tick towards Republicans in the toss-ups generally. Is that just a function of the existing polling just being farther away from present day and fundamentals making a bigger difference?

Ours barely noticed over the same two weeks: 50.9 to 51.0 expected Democratic-aligned seats, and 54 to 56 in 100 for control. Is your half point coming from a couple of races or spread across the map?

Is the 0.5 drop because of the roughly half-billion dollar ad blitz?

Sample size is the smaller problem. A clean random sample of 600 lands within about 4 points of the truth 95% of the time. The hard part is that the people who pick up aren't random, and no sample size fixes that. That's the error a forecast has to leave room for.

Those get weighted for partisan lean and past misses

Below error in senate you mean?

And the money effects. Why I’m not buying the wave yet.