Replies

Not sure yet.

My President and Senate models are a combination of the state polls and adjusted polls by statespoll.com, where available.

Because of the polling bias, it is hard to get a read on the national situation. Based on the published Presidential polls, Trump is running somewhere between 228 and 250 Electoral Votes. This range covers 30 days of polls, the lower end includes statespoll.com corrections and the higher range is ONLY statespoll.com corrections. I know these polls don't track with observed enthusiasm, so I will be watching for movement in the next two weeks.

On the Senate side, it's a different picture. I think those races come down to individual personalities.

I think Susan Collins is gone in Maine. She has less than a 20% chance of winning. There is no saving her this time.
For some reason, Colorado polling has been non-existent until this past week. Cory Gardner won his 2014 race by only 2%, and he's now running behind former governor John Hickenlooper by 9%. He has less than 1% chance of keeping his seat.
Next is Martha McSally in Arizona. The statespoll.com adjustments give her a 12% boost in her probability, but she's still running under a 25% chance of winning based on the polls.
Michigan is one of those states with big polling swings. John James is losing (38% probability) based on full polling, but is leading (69% probability) with only the statespoll.com adjusted polls.
Iowa is similar to Michigan, but more tight. Joni Ernst is tied (50% probability) with all polls, and is slightly ahead (59% probability) with only statespoll.com adjusted polls.
North Carolina hasn't seen the effects of the Democrat sex scandal in the polling. Thom Tillis is behind (36% probability) in all polling, but is ahead (68% probability) in statespoll.com adjusted polls, although Tillis doubled his position from last week in the full polling set.

-PJ

My President and Senate models are a combination of the state polls and adjusted polls by statespoll.com, where available.

That's what I figured. And there's nothing wrong with that, if the polls are accurate. I think they're pretty inaccurate, a la 2016. I have ignored Trump xx%, Biden yy% polls almost completely. I have looked at the Ras daily tracking poll, as well as polls measuring enthusiasm (out-of-WH-power Dems historically should be ahead, but they're trailing considerably). I am looking at GOP success in voter registration, as well as fundraising (2020 compared to 2016) and time the GOP has had to put ground games in place in key states. Plus how many unions have switched from Dem to GOP this election? That has to have some effect, but I'm not sure how to measure it. Certain union (mostly police, firefighters) get-out-the-vote operations are at a minimum not helping Democrats, and in some cases will be helping Republicans. Time will tell.