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
16 replies = new reply since forum marked as read
Highlight: NoneDon't highlight anything 5 newestHighlight 5 most recent replies
I forget now-- how reliable is Quinnipiac?? (Original Post) ailsagirl Sep 2016 OP
Reliable with a quantifiable R lean Loki Liesmith Sep 2016 #1
This is true ailsagirl Sep 2016 #4
538 gives them an A- pkdu Sep 2016 #2
wow-- that high?? ailsagirl Sep 2016 #3
This message was self-deleted by its author Loki Liesmith Sep 2016 #5
The grade is heavily weighted toward it's reliability Loki Liesmith Sep 2016 #6
After the pitiful showing in last night's program, how he could be trending higher is a mystery ailsagirl Sep 2016 #8
They have an R lean, but but it is consistent. Adrahil Sep 2016 #7
Quinnipiac's Republican Bias jamese777 Sep 2016 #9
This is a bit off-topic but there's humor at the end PJMcK Sep 2016 #10
Thanks for the post! ailsagirl Sep 2016 #14
Do you forget? alcibiades_mystery Sep 2016 #11
Did I forget if Quinnipiac is reliable? ailsagirl Sep 2016 #12
Oh, good alcibiades_mystery Sep 2016 #13
Oh, I see ailsagirl Sep 2016 #15
Not as reliable as the average of all polls Cicada Sep 2016 #16

Loki Liesmith

(4,602 posts)
1. Reliable with a quantifiable R lean
Thu Sep 8, 2016, 06:57 PM
Sep 2016

See 538s pollster page for more details.

Note that after these polls came in...Hillary's numbers on 538 went up. They are GOOD numbers.

Reliability and accuracy are not the same thing.

Response to ailsagirl (Reply #3)

Loki Liesmith

(4,602 posts)
6. The grade is heavily weighted toward it's reliability
Thu Sep 8, 2016, 07:06 PM
Sep 2016

The poll gets fewer crappy samples then others. But their model is biased a bit. Nate doesn't penalize as much for bias.

ailsagirl

(22,897 posts)
8. After the pitiful showing in last night's program, how he could be trending higher is a mystery
Thu Sep 8, 2016, 07:09 PM
Sep 2016

Maybe it takes a few days for the polling to catch up with last night's debacle

 

Adrahil

(13,340 posts)
7. They have an R lean, but but it is consistent.
Thu Sep 8, 2016, 07:09 PM
Sep 2016

Not a bad methodology, but their population model favors R's. It's a bit odd they haven't corrected for it.

PJMcK

(22,037 posts)
10. This is a bit off-topic but there's humor at the end
Thu Sep 8, 2016, 07:35 PM
Sep 2016

From 1986 until 1993, I lived on City Island in New York. One of my neighbors, who became a very good friend, worked for a major television network just at the advent of the internet. He was in charge of political polling for this network and he was remarkably intense about his fealty to the integrity of their polls. He was a disciple of Stan Opotowsky who developed many of the algorithms that are fundamental to modern polling techniques. Here's a link to Stan's history:

http://www.nytimes.com/1997/10/02/business/stanford-opotowsky-73-editor-at-new-york-post-and-abc-news.html

On the evening I met Stan, we drank a lot! He expressed a lot of fantastic observations and comments with insight, humor, intelligence and a decidedly Democratic political point of view.

That evening, we went sailing on my boat and I learned the Opotowsky Traffic Theorem: When driving a car in a major metropolitan area, and while listening to the local traffic reports, always do the opposite of the broadcasted directions. This is because everyone else is listening to the same broadcast. By the time you reach that interchange, everyone else will have diverted to the alternate location.

And there's a codicil: Never comment to the other people in the car about the traffic whether it's good or bad. Acknowledging the power of the universe over traffic only increases the frustration and potential for pollution of the sample.

He was a really fun gentleman who was clearly a liberal Democrat.

 

alcibiades_mystery

(36,437 posts)
13. Oh, good
Thu Sep 8, 2016, 07:49 PM
Sep 2016

I thought is was sarcastic troll shit, or Reliving the Primaries assholistic crowing or something.

You were being authentic! My mistake.

ailsagirl

(22,897 posts)
15. Oh, I see
Thu Sep 8, 2016, 07:52 PM
Sep 2016

You know, I hate reading threads like the one I started but was hoping Quinnipiac wasn't very reliable.

"sigh"

No problem! No troll I!! (or is it me?)

Cicada

(4,533 posts)
16. Not as reliable as the average of all polls
Thu Sep 8, 2016, 09:22 PM
Sep 2016

Any one poll is unreliable. So don't pay attention to a single poll. There is really no point.

Latest Discussions»Retired Forums»2016 Postmortem»I forget now-- how reliab...