Uniform swings
       |           

Welcome, Guest. Please login or register.
Did you miss your activation email?
March 19, 2024, 03:34:56 AM
News: Election Simulator 2.0 Released. Senate/Gubernatorial maps, proportional electoral votes, and more - Read more

  Talk Elections
  Presidential Elections - Analysis and Discussion
  U.S. Presidential Election Results (Moderator: Dereich)
  Uniform swings
« previous next »
Pages: [1]
Author Topic: Uniform swings  (Read 996 times)
phk
phknrocket1k
Atlas Icon
*****
Posts: 12,906


Political Matrix
E: 1.42, S: -1.22

Show only this user's posts in this thread
« on: November 02, 2011, 09:43:09 PM »

Why do people believe in this intellectual fraud?
Logged
RI
realisticidealist
Atlas Icon
*****
Posts: 14,688


Show only this user's posts in this thread
« Reply #1 on: November 05, 2011, 02:04:59 PM »

No one believes in uniform swings, but there isn't a better way to do it that isn't arbitrary. As Antonio said, there isn't nearly enough data to accurately calculate ceilings and floors and other things that happen in real life.
Logged
Person Man
Angry_Weasel
Atlas Superstar
*****
Posts: 36,681
United States


Show only this user's posts in this thread
« Reply #2 on: November 05, 2011, 02:11:17 PM »

Well, a uniform swing is like any model that doesn't make any more assumptions than the fact that the national vote has changed from last election cycle. As you put more and more variables in, the map changes to a more accurate depiction. This is political science, not Tarot.
Logged
RI
realisticidealist
Atlas Icon
*****
Posts: 14,688


Show only this user's posts in this thread
« Reply #3 on: November 05, 2011, 04:31:06 PM »

Well, a uniform swing is like any model that doesn't make any more assumptions than the fact that the national vote has changed from last election cycle. As you put more and more variables in, the map changes to a more accurate depiction. This is political science, not Tarot.

Which variables would those be? You could compenstate for home states fairly easily, but past that?
Logged
phk
phknrocket1k
Atlas Icon
*****
Posts: 12,906


Political Matrix
E: 1.42, S: -1.22

Show only this user's posts in this thread
« Reply #4 on: November 05, 2011, 04:47:50 PM »

Well, a uniform swing is like any model that doesn't make any more assumptions than the fact that the national vote has changed from last election cycle. As you put more and more variables in, the map changes to a more accurate depiction. This is political science, not Tarot.

Which variables would those be? You could compenstate for home states fairly easily, but past that?

I would gather money the candidate's campaign spent on said state, change in unemployment rate, etc.
Logged
RI
realisticidealist
Atlas Icon
*****
Posts: 14,688


Show only this user's posts in this thread
« Reply #5 on: November 05, 2011, 04:56:53 PM »

Well, a uniform swing is like any model that doesn't make any more assumptions than the fact that the national vote has changed from last election cycle. As you put more and more variables in, the map changes to a more accurate depiction. This is political science, not Tarot.

Which variables would those be? You could compenstate for home states fairly easily, but past that?

I would gather money the candidate's campaign spent on said state, change in unemployment rate, etc.

The first is irrelevant to hypotheticals, but could be used close to an election, and it definitely has an effect. The other I'm not so sure about...
Logged
phk
phknrocket1k
Atlas Icon
*****
Posts: 12,906


Political Matrix
E: 1.42, S: -1.22

Show only this user's posts in this thread
« Reply #6 on: November 05, 2011, 05:09:35 PM »
« Edited: November 05, 2011, 05:12:42 PM by phk »

Quote
You must be logged in to read this quote.

OK. Let's see.

Quote
You must be logged in to read this quote.

OK. So not a uniform swing than.

Quote
You must be logged in to read this quote.

I disagree with this. One of the implications of UNS is that every voter for the most part is a swing voter whose voting habits is perfectly correlated with the national swing. That is troubling and problematic because it is false to begin with.

Quote
You must be logged in to read this quote.

Well let's see here... I simply took each state's individual swing and divided by the national swing for 2008.

Hawaii               3.75334
Indiana   2.23124
North Dakota   1.92292
Nebraska   1.87873
Montana   1.86228
Utah               1.80062
Delaware   1.78726
Vermont   1.73381
New Mexico   1.63618
Nevada   1.54985
Illinois               1.51696
Virginia   1.49024
California   1.44707
Colorado   1.39979
Wisconsin   1.38952
South Dakota   1.34224
Michigan   1.33916
Idaho               1 31757
North Carolina1.31141
Maryland   1.28058
Oregon   1.25283
Connecticut   1.23330
Georgia   1.17163
Texas               1.14183
Kansas   1.07503
Iowa               1.04830
Washington   1.01747

Total   1.00000

New Jersey   0.90956
New York   0.88078
Maine               0.85612
New Hampshire 0.84687
South Carolina 0.83248
Florida               0.80267
Pennsylvania   0.80267
Wyoming   0.77595
Missouri   0.72559
Rhode Island   0.72456
Minnesota   0.69476
Ohio               0.68756
Mississippi   0.66906
D. C.               0.62487
Alabama   0.41521
Alaska               0.41213
Kentucky   0.37410
Arizona   0.20452
Massachusetts0.06680

Oklahoma   -0.01542
West Virginia   -0.02364
Tennessee   -0.08016
Louisiana   -0.42343
Arkansas   -1.03700

If you subtract 1 from the totals you can see how off a uni swing would be.

In fact if we are going for the range of [1.10 to .90] of the national swing you only get... 4 states.

For the sake of it.

5 states swung the other way.
18 states + DC swung in Obama's way but less than the national swing.
27 states swung greater than the national margin. 1 over 2 times greater and 1 over 3 times greater.

Quote
You must be logged in to read this quote.

No, there is no reason to assume that at all for the reasons I stated above. I even feel if in the case its right, it feels like its due to luck of the draw. Even if your not going for %'s but just the overall winner.

Quote
You must be logged in to read this quote.

Regression studies are used by experienced statisticians like Andrew Gelman and Nate Silver. Of course I take what they do relatively seriously.
Logged
Nichlemn
Jr. Member
***
Posts: 1,920


Show only this user's posts in this thread
« Reply #7 on: November 06, 2011, 08:06:57 AM »

I don't think uniform swings are particularly useful when you compare races where both of the major party candidates are different. By far the best application is with polling of a single race, e.g. if Obama's head-to-head numbers suddenly increased 5 percentage points nationally, it's likely that his numbers increased close to 5 percentage points in every state.
Logged
Pages: [1]  
« previous next »
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 0.214 seconds with 11 queries.