because of you being IAADS that I posed the questions. Models (in whatever form they take) are simplifications of the real world; in this case polling as predictive models of the election.
Who is sponsoring that site? It's a lot like WND or Breitbart. RWNJ propaganda and nothing else.PatGund wrote:Well, I think we have the "Unskewed Polls" site for 2016
https://www.longroom.com/polls/
Reality Check wrote:They claimed that using their methodology they would have correctly called the 2012 and 2008 Obama wins. I call BS. They show almost every poll is off by a large margin.
Responding to your post because it is so succinct, but this is to all those who enjoy the discussion. I enjoy it too--just not here. It is, IMO, worthy of its own thread. But I for one come here to see the polls, not to read polling theory. It's a totally different head set, y'know?Flatpointhigh wrote:As someone who suffers from dyscalculia, I really like the discussion on the maths of polling. I may not be able to wrap my head around all of it, but I like it.
Respectfully, I disagree. I think that polling and aggregation analysis* is as appropriate (and necessary) here as legal analysis is on any of the myriad threads we have on legal issues. Sorry.p0rtia wrote:Responding to your post because it is so succinct, but this is to all those who enjoy the discussion. I enjoy it too--just not here. It is, IMO, worthy of its own thread. But I for one come here to see the polls, not to read polling theory. It's a totally different head set, y'know?Flatpointhigh wrote:As someone who suffers from dyscalculia, I really like the discussion on the maths of polling. I may not be able to wrap my head around all of it, but I like it.
edited for grammary.
Still respectfully disagree. The methodological discussion is inseparable from the current results and forecasts if there is any attempt to understand what the polls (or the the aggregators) mean. Again I make the analogy to how discussion regarding legal matters works on the Fogbow. Remember, there might be far more lawyers on the Fogbow than mathematicians, but we are way prettier than you*.listeme wrote:The topic is 2016 polls, though.
Sure, some analysis. But I'm going to disagree with you on this, you guys -- this topic is a more specific topic than polling. This is recent polling, polling in the news, et cetera, or at least that is how a lot of us are interpreting the title.
Analysis of methodology is a much more general topic.
It's a granularity problem.
538: Who’s Behind A Mysterious Website Saying Polls Are Skewed Against Trump?:esseff44 wrote:Who is sponsoring that site? It's a lot like WND or Breitbart. RWNJ propaganda and nothing else.PatGund wrote:Well, I think we have the "Unskewed Polls" site for 2016
https://www.longroom.com/polls/
But LongRoom and whoever runs it has gone out of its way to obfuscate its identity. The site has an “about us” page which lists four people associated with the site, but they each seem to be without any semblance of an online paper trail, an odd thing in the age of the internet.
“Michael Ellis,” the man listed as LongRoom’s managing editor, is described in only the vaguest of terms as “an Internet Executive with over 23 years of experience, including general management of mid to large sized publications. He has been involved with internet community management his entire career.” The three other staff members have similarly indistinct bios, and rather than photographed headshots, the staff is depicted in sketches. None of the staff appears to have Twitter accounts, let alone follow the @LongRoomNews account. Searches for the staff on other social networking sites did not lead anywhere and there is no listed point of contact for any of the LongRoom staff members anywhere on the site. FiveThirtyEight reached out to the site’s only point of contact for comment — a support email address — and did not hear back. A public records search for LongRoom yielded no results for the business. (In addition to its “unbiased” polling operation, the site aggregates news stories.)
But an analysis of the site’s IP address showed that in April 2015, LongRoom switched its registration to a domain that for a fee, allows registrants to keep their names private — Domains By Proxy, LLC. The last name associated with the website, as recently as January 2015, is Fred Waid, who listed the site’s associated organization as “American Separatist” based out of New Mexico. FiveThirtyEight reached out to Waid but had not heard back as of publication.
x6 
x2 
x4 x2 
Many people have so speculated.tek wrote:Sounds like the KGB to me.
It would be irresponsible not to speculate.
Good numbers.Reality Check wrote:Edit: Adding link: http://www.wsj.com/articles/hillary-cli ... 1471015822New state polls:
New NBC/WSJ/Marist polls
CO: HRC 46, Trump 32 (+14)
FL: HRC 44, Trump 39 (+5)
NC: HRC 48, Trump 39 (+9)
VA: HRC 46, Trump 33 (+13)
Aug 4-10
