I debias betting market data for favorite/longshot bias per the method described in this paper by economist David Rothschild.
I don’t want to derail the thread, but I can go into more detail on this if necessary.
Is an In-Bev product and not Molsen-Coors so that alone is enough to disqualify him even ignoring the ice. I think the real question is who gets the higher percent of votes: O’ Dea for Senator who is moderate and publicly says Trump should not run in 2024 or Heidi for Governor who is still very much a Trumpanista. That may foretell the future of Coloradan Republicans statewide.
I just cast my ballot. Ryan for Senate, Whaley for governor. Dems down ballot…until there weren’t any more, notably for the judiciary. No choice for state rep. Judging from the lawn signs in my part of Ohio, I was right and truly alone.
Excellent. So as far as we know Ryan has a 1-0 lead over Vance in Ohio. Now all he has to do is play solid defense.
RCP is now predicting GOP +4 in the Senate.
RealClearPolitics - 2022 Election Maps - Battle for the Senate 2022
Before we get too excited, note that RealClearPolitics’ reputation has been getting foggy over the past few years.
The RCP average is just a straightforward average of the polls which should give a pretty good first approximation.
Here’s the weird thing, in two of their GOP pick ups (AZ, NH) and one GOP hold (PA) all have the Dem leading in RCP average. But then RCP applies something they call “Adjusted Polling Average November 2” which I don’t recall seeing before and it’s not clear where that comes from.
If I had to guess, based on some other columns in the table that are how far the polls were off and in which direction in the previous three elections, it might be a weighted average of of those. I think this is a bad idea if that’s what they’re doing because I’m pretty sure polling error in general is normally distributed.
But these guys have been doing this, well not this exactly, for a long time, so there might be more to it than that, presuming there’s at least one person over there who knows what they’re doing and the folks are letting that dude do it.
For comparison 538 Deluxe Senate has it Hassan +3.7, the RCP average has it Hassan +1.0, and RCP Nov 2 has it Bolduc +2.1.
I think this is right. They’re basically saying that if these polls were off by 2% in favor of Democrats in the last election, then their current results should be adjusted. This assumes that the error was due to some methodological problem and not just variance in the way polling is done, which I don’t think is correct, or at least not obvious.
Another problem with RCP is recency bias. Their current numbers weight the last X number of polls. But since crappy polls are more frequent than good polls, their average probably overweights lousy polls. On the flip side, if the electorate is changing fast, relying on older but more accurate polls (when they were taken) can introduce its own errors.
But if you just mentally correct for those things when you see them, RCP is a good aggregator of polls.
They certainly used to be, but I see no real reason to prefer them over the 538 “straight polls” average. That seems to adjust for all of the things you mention (recency and poll quality in particular).
538’s modal outcome is R+1 (for a 51-seat majority) which certainly feels like the most likely outcome to me as well as of today. With an 80% confidence range between 52 and 48 seats for the GOP. 54 seats (the RCP “call” I guess) has a 7.2% of happening, so I suppose if that is the outcome we could say that RCP was “right” and 538 “wrong”.
Clearly it would mean that polls like Trafalgar and InsiderAdvantage were more accurate than others (once again).
I’m a chief election inspector (poll worker), and yesterday we pre-numbered the mail-in ballots (we can’t actually process them until Tuesday). I’m in Wisconsin, obviously with a critical Senate race.
We had about half as many mail-ins as in 2020. Normally, I’d expect about that ratio (presidential vs. midterm), but this year’s midterm is so visible (especially in a few states like ours), I would expect more like 75% of 2020’s numbers.
If the turnout IS more like 75%, that means a lot more will vote in person on Tuesday, compared to the proportion in 2020. If that’s the case, does this portend a Republican wave? (Obviously I’m extrapolating from one ward to the whole country, but we ARE very “purple” here). Recall that Republicans disproportionately voted in person (on Election Day) in 2020.
Or does it just mean more people are returning to voting in person, period, and so maybe my little data point doesn’t suggest a Republican blowout?
We’ll see soon enough (God, I hope we don’t have to wait for a Georgia runoff to know which will be majority in the Senate).
Jon Ralston, Nevada elections expert from the Nevada Independent, has put out his 2022 election predictions, and he has Cortez Masto winning narrowly.
He looks at data of course, but there’s also a fair amount of his gut in there.
The article covers a lot of offices up for election in addition to the senate race, so the whole thing might be worth a read if you’re interested.
That’s great news. I mean, at best it’s Ralston’s educated guess, but no one knows Nevada polling and voting like him.
It seems to me that a couple of days ago the Democrats were slightly ahead in all the toss up states but today they are all slightly behind. I do not understand how there are so many clearly insane people in this country so willing to vote against their own interests. I just have to keep calm and remember that nothing too bad can happen for the next two years. The Republicans will not get 67 in the Senate. My new mantra is VETO.
Republicans are very good indeed at inspiring moral panic through a relentless campaign of propaganda and lies. It’s not that people are “insane”; they’re just afraid and angry and thus not thinking or amenable to doing so. And once they’ve bought into the lies, it’s very hard to dislodge them.
Here’s a fitting poem, which I first heard on NPR many years ago:
THE POOR VOTER ON ELECTION DAY (1852)
by John Greenleaf Whittier
The proudest now is but my peer,
The highest not more high;
To-day, of all the weary year,
A king of men am I.
To-day alike are great and small,
The nameless and the known
My palace is the people’s hall,
The ballot-box my throne!
Who serves to-day upon the list
Beside the served shall stand;
Alike the brown and wrinkled fist,
The gloved and dainty hand!
The rich is level with the poor,
The weak is strong to-day;
And sleekest broadcloth counts no more
Than homespun frock of gray.
To-day let pomp and vain pretence
My stubborn right abide;
I set a plain man’s common sense
Against the pedant’s pride.
To-day shall simple manhood try
The strength of gold and land
The wide world has not wealth to buy
The power in my right hand!
While there’s a grief to seek redress,
Or balance to adjust,
Where weighs our living manhood less
Than Mammon’s vilest dust, -
While there’s a right to need my vote,
A wrong to sweep away,
Up! clouted knee and ragged coat!
A man’s a man to-day!
(Not to mention the vital other half of the electorate now… but Whittier was writing long before the 19th Amendment).
If so, do Murkowski and Collins become the new Manchin and Sinema? That may not be a bad thing. And of note I don’t trust Collins anymore so maybe the Senate breakdown would more properly read as
50 D + VP / Murkowski + 50 R. I think that might be the most palatable Senate the Dems and Moderate Pubs could hope for.
I guess, but it’s not like the GOP house will be approving any palatable bills anyway. Other than mandatory budget bills I don’t expect anything other than hearings and impeachments from their next Congress, barring a miracle today.
I guess Collins and Murkowski might matter if a Scotus seat opens up. But I doubt it.
Nowhere idea else to post this… Lisa Desjardins on PBS Newshour last night laid out a clear picture on the Secretary of State races with an excellent map showing where the GOP candidates of each state stand on the 2020 election results: which ones deny, which ones fuel doubts and which ones defend the results.
Start at 9:25.
FYI Murkowski may lose to a Trump republican. She is in a 4-way ranked choice general election with 2 lther Republicans and 1 Democrat with her and the top challenger to Murkowski (Tshibaka) is a Trumper who is polling close to her.
I thought the election had loosened up in favor of Murkowski since September.