Professor Emeritus

Veteran
Poster of the Year
Supporter
Joined
Jan 5, 2015
Messages
51,330
Reputation
19,731
Daps
203,943
Reppin
the ether
They should then say how much of a region they are able to capture vs the last census numbers of said region. And they should also include a data unavailable number in their polls as well.
And there doesn't seem to be a mandate on accuracy. Polls are wrong all the time. No one got the presidential race right in 2016.
Most reputable poll companies include a lot of those breakdowns in the pages of supplementary information that accompany the polls. But the available solid information doesn't actually tell you what you need to know. You can know that your response rate was 76%, but you can't know how those other 24% would have chosen. So you have to do projections based on the demographics and past history. Which is what all the good poll companies do.

Saying "no one got the presidential race right in 2016" is an ill-informed way to look at how results are given. 538 predicted that Hillary Clinton has a 70% chance of winning. That means that Donald Trump had a 30% chance. They weren't wrong, because if the underdog didn't win 30% of the time with those numbers than those percentages WOULD be wrong. Their "70% chance" is only accurate if the underdog actually wins nearly 1/3 of the time. If they were giving you the absolute winner, then they would say 99% or 100%. You can't even say the polls were wrong, because those polls might have completely accurately represented the electorate, but the fact that the Anthony Weiner case dropped the weekend before the vote suppressed turnout slightly, or the weather wasn't good here and there - the margin Trump won by was smaller than the expected random variance in three states.

Polls attempt to be the most accurate representation possible. You want to treat them like they're prophesies (which is understandable, because the media and most people in the horserace game try to treat them like that too). It's nonsensical. If someone is about to roll a dice and they say there's an 83% chance it will be 5 or lower, and the dice comes up 6, were they "wrong"?
 

AnonymityX1000

Veteran
Joined
Jun 6, 2012
Messages
31,004
Reputation
3,101
Daps
70,654
Reppin
New York
Most reputable poll companies include a lot of those breakdowns in the pages of supplementary information that accompany the polls. But the available solid information doesn't actually tell you what you need to know. You can know that your response rate was 76%, but you can't know how those other 24% would have chosen. So you have to do projections based on the demographics and past history. Which is what all the good poll companies do.

Saying "no one got the presidential race right in 2016" is an ill-informed way to look at how results are given. 538 predicted that Hillary Clinton has a 70% chance of winning. That means that Donald Trump had a 30% chance. They weren't wrong, because if the underdog didn't win 30% of the time with those numbers than those percentages WOULD be wrong. Their "70% chance" is only accurate if the underdog actually wins nearly 1/3 of the time. If they were giving you the absolute winner, then they would say 99% or 100%. You can't even say the polls were wrong, because those polls might have completely accurately represented the electorate, but the fact that the Anthony Weiner case dropped the weekend before the vote suppressed turnout slightly, or the weather wasn't good here and there - the margin Trump won by was smaller than the expected random variance in three states.

Polls attempt to be the most accurate representation possible. You want to treat them like they're prophesies (which is understandable, because the media and most people in the horserace game try to treat them like that too). It's nonsensical. If someone is about to roll a dice and they say there's an 83% chance it will be 5 or lower, and the dice comes up 6, were they "wrong"?
When they gave Nate Silver credit for picking every state correctly in 2012. What did they mean?
Triumph of the Nerds: Nate Silver Wins in 50 States
He gave a higher percentage to chances to win to Obama in some states and some higher to Romney and it was all correct thus predicting them all correctly. If the media calls him prescient and called 'a perfect game' in that scenario he also gets to be called 'wrong' when it is the opposite scenario like 2016.
 

Professor Emeritus

Veteran
Poster of the Year
Supporter
Joined
Jan 5, 2015
Messages
51,330
Reputation
19,731
Daps
203,943
Reppin
the ether
When they gave Nate Silver credit for picking every state correctly in 2012. What did they mean?
Triumph of the Nerds: Nate Silver Wins in 50 States
He gave a higher percentage to chances to win to Obama in some states and some higher to Romney and it was all correct thus predicting them all correctly. If the media calls him prescient and called 'a perfect game' in that scenario he also gets to be called 'wrong' when it is the opposite scenario like 2016.

Yes, like I said right in that comment you quoted, the media is low-IQ about it too. Are you saying that an obviously incorrect interpretation of .538's numbers is right just because the media makes the same mistake?
 

AnonymityX1000

Veteran
Joined
Jun 6, 2012
Messages
31,004
Reputation
3,101
Daps
70,654
Reppin
New York
Yes, like I said right in that comment you quoted, the media is low-IQ about it too. Are you saying that an obviously incorrect interpretation of .538's numbers is right just because the media makes the same mistake?
I'm saying the way polling data is presented is not in a true statistical framework. All the things I said that need to be done differently and need to be included along with the desired information would help stop incorrect interpretation. And would prove a lot of polls are skewed toward likely voters and like the person I was quoting showed most polls skew toward older voters. They stubbornly stick to doing that based on outdated information or presenting what they want to happen. Young people vote now in 2016 & 2018 according to the data and the polls are not reflecting that in their sampling still.
When I said Nate Silver got it wrong, yes I was basing it on the praise, fame and personal success he gained for getting it right in the past. I was using the consensus framework when talking about it.
 

wtfyomom

All Star
Joined
May 9, 2012
Messages
7,704
Reputation
-757
Daps
11,438
Reppin
NULL
Your being incredibly dismissive. Read this article about Bernie's tenure as Mayor of Burlington which goes into detail how he was able to successfully carry out an agenda that helped revitalize Burlington's downtown area through his vision bringing in affordable housing, worker cooperative businesses, and environmentally friendly initiatives. The legacy he left on Burlington could still be felt through this day and he was popular pushing these agendas. I'd argue running a city is much more difficult than running some bureau because being the executive of a city no matter how small it may be still comes with it some push back from political forces.
his first book, outsider in the house, details his time as mayor and he will have to deal with a lot of the same issues as president albeit a bigger scale but hes not going into the presidency thinking theres going to be a red carpet for his agenda and it shows how he has dealt with it before
 

wtfyomom

All Star
Joined
May 9, 2012
Messages
7,704
Reputation
-757
Daps
11,438
Reppin
NULL
@wtfyomom Not trying to be one of those polls are fake people but this guy I followed during the 2018 mid terms has some analysis about the polls for Bernie. Some of them are underestimating under 50 voter turnout






yeah i agree, hes being underestimated but at the same time, even if its that hes not doing well with boomers which we know, its still a concern so still had my attention
 

StatUS

Veteran
Supporter
Joined
Apr 30, 2012
Messages
28,495
Reputation
1,884
Daps
62,603
Reppin
Everywhere
Bernie is 2nd in the latest NC poll. He's been running second ahead of Warren in the last couple of polls I've seen.
 

Berniewood Hogan

IT'S BERNIE SANDERS WITH A STEEL CHAIR!
Joined
Aug 1, 2012
Messages
17,983
Reputation
6,880
Daps
88,330
Reppin
nWg
Not important, but Private Pyle is having a meltdown because Bernie Bros replied to some if his Tweets.

 
Top