Welcome to DU! The truly grassroots left-of-center political community where regular people, not algorithms, drive the discussions and set the standards. Join the community: Create a free account Support DU (and get rid of ads!): Become a Star Member Latest Breaking News General Discussion The DU Lounge All Forums Issue Forums Culture Forums Alliance Forums Region Forums Support Forums Help & Search

ecdab

(930 posts)
Sun Nov 8, 2020, 11:12 AM Nov 2020

Polling versus Election Results

The below presentation of data compares the final polling averages from 538.com to election results reported as of 11/8/2020 at 6AM eastern time.

T=Trump, B=Biden, U=Uncommitted (either undecided or favoring 3rd party in polling, voted for 3rd party in results)

Please note that some states have still reported less than 90% of their vote total, and those states could see major changes to the percentages between now and when their counting is complete (see Alaska, California, DC, Illinois, Maryland, New Jersey, New York, Utah).

Initial thoughts:

There were a number of States that were not polled very much, and the primary pollster for many of those States was Survey Monkey. Those States were all non-battle ground States and tended to have low electoral vote counts. The results in those States do skew in Biden’s favor, but that is not a reflection on polling in general, that is a reflection primarily on Survey Monkey.

Most States were polled fairly extensively. In those States, Biden’s polling numbers were very accurate. The pattern that emerges over and over again is that people that were uncommitted to Trump or Biden in the polling ended up voting for Trump.
Most people, including myself, expected those uncommitted people to break for Biden – because the pandemic, the floundering economy, Trump’s behavior, and the fact the undecided voters almost always break against the incumbent. That would have created the Blue Wave that was predicted by many.

That leaves us with the question; why would these people break for Trump? Did all the gas lighting that the GOP did actually work? It would appear so. How do we combat that going forward? That’s beyond me.

In terms of the polling, I think this data shows that the polling, for the most part, was quite good. I think the explanation from most media sources regarding the meaning, or even the existence, of respondents to polls that were not committed to Trump or Biden and what that meant to possible outcomes was quite bad. The numbers below show that 538 deserves a lot of credit.

STATE/ % REPORTED/ POLLING DATA %/ RESULTS TO DATE %/DIFFERENCE %

ALABAMA/ 99/ T57.4 B37.8 U4.8/ T62.1 B36.5 U1.4/ T+4.7 B-1.3 U-3.4
ALASKA/ 47/ T51.2 B43.6 U5.2/ T62.9 B33.0 U4.1/ T+11.7 B-10.6 U-1.1
ARIZONA/ 97/ B48.7 T46.1 U5.2/ B49.5 T48.9 U1.6/ B+0.8 T+2.8 U-3.6
ARKANSAS/ 99/ T58.9 B36.2 U4.9/ T62.6 B34.6 U2.8/ T+3.7 B-1.6 U-2.1
CALIFORNIA/ 77/ B61.6 T32.4 U6.0/ B65.1 T33.0 U1.9/ B+3.5 T+0.6 U-4.1
COLORADO/ 95/ B53.6 T41.1 U5.3/ B55.3 T42.1 U2.6/ B+1.7 T+1.0 U-2.7
CONNECTICUT/ 97/ B58.6 T34.6 U6.8/ B59.3 T39.1 U1.6/ B+0.7 T+4.5 U-5.2
DELAWARE/ 99/ B58.9 T 34.6 U6.5/ B58.8 T39.8 U1.4/ B-0.1 T+5.2 U-5.1
DC/ 80/ B90.8 T5.8 U3.4/ B92.6 T5.2 U2.2/ B+1.8 T-0.6 U-1.2
FLORIDA/ 96/ T46.6 B49.1 U4.3/ T51.2 B47.8 U1.0/ T+4.6 B-1.3 U-3.3
GEORGIA/ 99/ B48.5 T47.4 U4.1/ B49.5 T49.3 U1.2/ B+1.0 T+1.9 U-2.9
HAWAII/ 99/ B64.3 T30.0 U5.7/ B63.7 T34.2 U2.1/ B-0.6 T+4.2 U-3.6
IDAHO/ 99/ T56.6 B38.5 U4.9/ T63.8 B33.1 U3.1/ T+7.2 B-5.4 U-1.8
ILLINOIS/ 89/ B55.0 T39.0 U6.0/ B55.3 T42.7 U2.0/ B+0.3 T+3.7 U-4.0
INDIANA/ 99/ T52.9 B42.0 U5.1/ T57.0 B40.9 U2.1/ T+4.1 B-1.1 U-3.0
IOWA/ 92/ T47.6 B46.3 U6.1/ T53.1 B44.9 U2.0/ T+5.5 B-1.4 U-4.1
KANSAS/ 99/ T53.9 B41.0 U5.1/ T56.7 B41.1 U2.2/ T+2.8 B+0.1 U-2.9
KENTUCKY/ 98/ T55.6 B39.9 U4.5/ T62.2 B36.0 U1.8/ T+6.6 B-3.9 U-2.7
LOUISIANA/ 99/ T57.6 B37.1 U5.3/ T58.5 B39.8 U1.7/ T+0.9 B+2.7 U-3.6
MAINE/ 91/ B53.3 T40.3 U6.4/ B53.5 T43.4 U3.1/ B+0.2 T+3.1 U-3.3
MARYLAND/ 70/ B63.1 T28.9 U8.0/ B63.1 T35.1 U1.8/ B+/-0.0 T+6.2 U-6.2
MASS./ 92/ B64.6 T28.9 U6.5/ B65.3 T32.4 U2.3/ B+0.7 T+3.5 U-4.2
MICHIGAN / 99/ B51.2 T43.2 U5.6/ B50.5 T47.9 U1.6/ B-0.7 T+4.7 U-4.0
MINNESOTA/ 96/ B51.8 T42.7 U5.5/ B52.5 T45.4 U2.1/ B+0.7 T+2.7 U-3.4
MISSISSIPPI/ 86/ T55.5 B39.6 U4.9/ T59.5 B38.9 U1.6/ T+4.0 B-0.7 U-3.3
MISSOURI/ 97/ T51.6 B43.6 U4.8/ T56.9 B41.3 U1.8/ T+5.3 B-2.3 U-3.0
MONTANA/ 99/ T49.8 B45.4 U4.8/ T56.7 B40.4 U2.9/ T+6.9 B-5.0 U-1.9
NEBRASKA/ 99/ T52.1 B42.5 U5.4/ T58.5 B39.1 U2.4/ T+6.4 B-3.4 U-3.0
NEVADA/ 95/ B49.7 T44.4 U5.9/ B50.0 T47.8 U2.2/ B+0.3 T+3.4 U-3.7
NEW HAMP./ 99/ B53.9 T42.8 U3.3/ B52.6 T45.5 U1.9/ B-1.3 T+2.7 U-1.4
NEW JERSEY/ 80/ B58.4 T37.9 U3.7/ B58.4 T40.1 U1.5/ B+/-0.0 T+2.2 U-2.2
NEW MEXICO/ 99/ B53.9 T42.1 U4.0/ B54.2 T43.6 U2.2/ B+0.3 T+1.5 U-1.8
NEW YORK/ 81/ B62.3 T32.9 U4.8/ B58.3 T40.4 U1.3/ B-4.0 T+7.5 U-3.5
N. CAROLINA/ 98/ T47.1 B48.9 U4.0/ T50.0 B48.6 U1.4/ T+2.9 B-0.3 U-2.6
N. DAKOTA/ 91/ T56.0 B38.7 U5.3/ T65.0 B31.7 U3.3/ T+9.0 B-7.0 U-2.0
OHIO/ 92/ T47.5 B46.8 U5.7/ T53.3 B45.2 U1.5/ T+5.8 B-1.6 U-4.2
OKLAHOMA/ 96/ T59.2 B36.2 U4.6/ T65.4 B32.3 U2.3/ T+6.2 B-3.9 U-2.3
OREGON/ 97/ B58.7 T37.4 U3.9/ B56.5 T40.4 U3.1/ B-2.2 T+3.0 U-0.8
PENNSYLVANIA/ 98/ B50.2 T45.6 U4.2/ B49.6 T49.1 U1.3/ B-0.6 T+3.5 U-2.9
RHODE ISLAND/97/ B62.9 T32.4 U4.7/ B59.4 T39.1 U1.5/ B-3.5 T+6.7 U-3.2
S. CAROLINA/ 99/ T51.6 B44.5 U3.9/ T55.1 B43.4 U1.5/ T+3.5 B-1.1 U-2.4
S. DAKOTA/ 98/ T54.5 B39.0 U6.5/ T61.8 B35.6 U2.6/ T+7.3 B-3.4 U-3.9
TENNESSEE/ 99/ T55.1 B41.4 U3.5/ T60.6 B37.4 U2.0/ T+5.5 B-4.0 U-1.5
TEXAS/ 97/ T48.6 B47.4 U4.0/ T52.2 B46.3 U1.5/ T+3.6 B-1.1 U-2.5
UTAH/ 88/ T51.9 B42.1 U6.0/ T58.7 B37.2 U4.1/ T+6.8 B-4.9 U-1.9
VERMONT/ 95/ B66.5 T27.8 U5.7/ B64.9 T31.7 U3.4/ B-1.6 T+3.9 U-2.3
VIRGINIA/ 99/ B53.7 T41.9 U4.4/ B54.0 T44.4 U1.6/ B+0.3 T+2.5 U-2.8
WASHINGTON/ 96/ B59.4 T36.4 U4.2/ B58.8 T38.7 U2.5/ B-0.6 T+2.3 U-1.7
W. VIRGINIA/ 99/ T62.1 B33.5 U4.4/ T68.2 B30.0 U1.8/ T+6.1 B-3.5 U-2.6
WISCONSIN/ 99/ B52.1 T43.7 U4.2/ B49.4 T48.8 U1.8/ B-2.7 T+5.1 U-2.4
WYOMING/ 99/ T62.3 B30.8 U6.9/ T69.9 B26.6 U3.5/ T+7.6 B-4.2 U-3.4

4 replies = new reply since forum marked as read
Highlight: NoneDon't highlight anything 5 newestHighlight 5 most recent replies

Cicada

(4,533 posts)
1. I got attacked here for arguing that betting markets have predictive validity
Sun Nov 8, 2020, 11:21 AM
Nov 2020

Betting markets historically have been pretty good predicting elections. Maybe better than polls. In 2016 and 2020 they gave Trump about a one third chance of winning. Maybe that was better info than we got from polls.

Another interesting fact is that USC Dornife has been including a weird question asking who your social contacts will vote for. This year that question accurately predicted Biden’s margin in the popular vote. They claim that question has been more accurate in past elections too.

Tribetime

(4,692 posts)
2. So Colorado California D.C. and Georgia
Sun Nov 8, 2020, 11:34 AM
Nov 2020

Were the only states where the polling matched. With Colorado and D.C. the only only one's that were a tad off in Biden's favor. This gives me hope for the runoffs...but we need to get 100% paper ballots

ecdab

(930 posts)
3. Georgia polling turned out to be quite good. let's hope the run off election polling is good.
Sun Nov 8, 2020, 01:04 PM
Nov 2020

Figuring out why folks appear to have broken late for the GOP is critical to understanding what to do in the future - including the immediate future in Georgia.

Tribetime

(4,692 posts)
4. Its odd that all the polling that was off 46 of 50
Sun Nov 8, 2020, 02:11 PM
Nov 2020

All favored Trump......what are the odds 10 billion to 1 ?

Latest Discussions»General Discussion»Polling versus Election R...