Outside NC redistricting, do any seats flip D-->R? (user search)
       |           

Welcome, Guest. Please login or register.
May 19, 2024, 01:52:15 AM
News: Election Simulator 2.0 Released. Senate/Gubernatorial maps, proportional electoral votes, and more - Read more

  Talk Elections
  Other Elections - Analysis and Discussion
  Congressional Elections (Moderators: Brittain33, GeorgiaModerate, Gass3268, Virginiá, Gracile)
  Outside NC redistricting, do any seats flip D-->R? (search mode)
Pages: [1]
Poll
Question: Outside NC redistricting, do any seats flip D-->R?
#1
Yes
 
#2
No
 
Show Pie Chart
Partisan results

Total Voters: 38

Author Topic: Outside NC redistricting, do any seats flip D-->R?  (Read 1070 times)
UncleSam
Sr. Member
****
Posts: 2,520


« on: April 30, 2024, 01:35:01 PM »

At face value this may seem statistically unlikely, but the reverse almost happened in 2020 with only one seat flipping R-->D for non-redistricting reasons; GA-07, a suburban Atlanta seat that is zooming left. The other 2 D flips were ironically from NC redistricting. Republicans were able to hold Clinton won seats many expected to flip like CA-25 and TX-23.
2018 was a blue wave and D was over stretched. 2022 was not.

There are 10 seats where R has at least 20% to flip. The probability of none of them flipping is (0.Cool^10*(0.9)^10=0.037.

The issue is that those are not completely independent events. If Dems hold PA-07 for instance, their chances in nearby PA-08 are also probably pretty good.
I already considered this by using probabilities like 20% and 10%. Otherwise seats like CA47 and MI07 have about 40% chance to flip.

That's still not how probability works; even if you raise the base percentage, you're still treating them as independent events.

One way to think about it is if you give Biden a healthy 60% chance of winning each of the main swing states individually (AZ, MI, NC, NV, PA, WI), by your logic the chance he wins all 6 again is only 5% which is way too low. Likewise for Trump.
I know they are not completely independent. This is just a very rough estimation, since there is no way to know the conditional probabilities like P(AZ|PA) . Using the Biden example, his chance of winning each of them is about 50, since they are not independent, let's raise them to 80%, and the result is 26.2%. Not very high but still notable.
I think at that point you might as well just estimate the aggregate probability, rather than using an inaccurate formula with known inaccurate assumptions in attempt to get them to cancel out.

I will say that while not independent, house races tend to be much more localized than president. It is very unlikely that one party will sweep all of the competitive races, and if one does then by November the races we will be viewing as competitive will have shifted drastically.
Logged
Pages: [1]  
Jump to:  


Login with username, password and session length

Terms of Service - DMCA Agent and Policy - Privacy Policy and Cookies

Powered by SMF 1.1.21 | SMF © 2015, Simple Machines

Page created in 0.023 seconds with 12 queries.