Things are not as rosy as CH wants them to be. Another group of polls showing some interesting data, kinda like the 0% black support for trump yesterday that CH with his infinite wisdom discounted the polls as bs. Oh well, can't have your cake and eat it at the same time. LOLOL This is getting to be a lot of fun for someone like myself who enjoys watch CH spin like a top when it comes to serving cakes.
Hillary Clinton is ahead of Donald Trump in several key battleground states, a series of NBC/Wall Street Journal/Marist polls of registered voters released on Friday found.
Clinton held an 8-point edge in Colorado, according to one poll — a similar margin to other surveys released in the last several weeks.
The former secretary of state garnered 44% support to Trump's 37% in Florida, the battleground state with the largest number of electoral college votes.
Clinton held a 6-point lead over the real-estate magnate in North Carolina, and a 9-point advantage in Virginia.
"With 66 electoral votes at stake in these four states, Donald Trump is playing catch-up against Hillary Clinton," Marist College polling director Lee Miringoff said in a release.
The polls come days after a series of surveys showed Trump gaining on Clinton in states like Pennsylvania and Ohio, which some observers attributed to FBI Director James Comey's critical assessment of Clinton's handling of classified information in emails while she was secretary of state. Friday's NBC/WSJ/Marist polls, however, were largely conducted after Comey's announcement.
Did you see the LA Times came out with their first tracking poll today? As much as they tried and tried in their article to excuse the results, it shows Trump up 3 points.
That's three national polls in three days without a lead for Hillary.
btw... the NBC/Marist polls have been on the heavy side for Clinton. The previous NBC/Marist poll in Pennsylvania for instance, showed Hillary leading by 15. It came out at the same time three other polls came out. Those three showed the race tied, a one point lead, and a three point lead. The Marist poll was nearly 14 points higher than the average of the other three.
Hmmmm. It seems there were quite a few back in the day. Coolidge was the last before Agnew. Teddy Roosevelt was another. Tyler, Van Buren, and Jefferson were all governors who became vice presidents (although Jefferson was "elevated" as the Electoral College runner-up).
As much as they tried and tried in their article to excuse the results, it shows Trump up 3 points
And your opinion is duly noted. BTW, who cares that you think they tried to white wash a number. The bottom line is you liked the results they reported, which you still managed to criticized. Wow, imagine if they didn't. LOL!! Dayum CH, your bias is in full bloom. Plus I'd bet those numbers will change many times in the next few weeks, especially after the RNC convention that will be a disaster. Oh, well.
Rep. Darrell Issa (R-CA) told Brietbart News Daily that now might be a good time for the Republican leadership to shut down the federal government, in protest of what he called “an imperial president” who will not “enforce criminal charges against a criminal.”
Issa was referring to the lack of charges filed against Hillary Clinton for her private email server.
Said Issa: “We should be willing to shut down the government if the president won’t limit his power.” ___________ He should go against the recommendation of the Republican head of the FBI?
“I don’t know how women can vote for him… It’s incomprehensible to me.” — Barbara Bush, in an interview with CBS News, on Donald Trump as the Republican nominee.
I looked at Silver's opinion on the fivethirtyeight blog, and apparently I was right that he's making some assumptions about swing states and battleground states.
I quote: "It’s not quite correct to characterize the race as a tossup, however. A relatively emphatic majority of recent swing state and national polls still have Clinton ahead, but often by narrower margins than before." but the degree of uncertainty is "High, or perhaps very high, for a variety of reasons." he says.
So his analysis, based on slim and dwindling leads in swing states, is highly or very highly uncertain - which necessarily means that he is relying on assumptions about them. It's rather foolish, IMO, to try to distill such small differences having high uncertainty into such precise number as his "37% likelihood" of a Trump win, or even your 2-1 odds favoring Clinton.
Nate Silver betrays a certain amount of personal bias by dismissing certain poll trends as "fairly normal fluctuations" yet relying heavily on those polls for his pseudo-precise handicapping of the race. Indeed, Silver characterizes some of the literally most important factors (everything "apart from head-to-head polls") as something he worthy of no more than "a love-hate relationship". I know what it means to be a primarily numbers guy, since I'm one of those myself, but Silver's whole approach on things leaves me skeptical, with "High, or perhaps very high" uncertainty regarding his methodology.
WP, Silver has always relied on assumptions in his predictions. He runs hundreds or thousands of scenarios, and when he says "Trump has a 33% chance of winning" he means Trump won 33% of the scenarios he ran. I do like that he's acknowledging the uncertainly in his underlying assumptions this race creates. In previous elections there really hasn't been a heck of a lot of difference among each party's nominees so there was a lot of historical data to work with in those scenarios. But both Trump and Clinton fall outside the norm for the previous four elections, bringing new factors into voters' decisions.
Indy, I don't believe that there's any meaningful way of "running hundreds of thousands scenarios" without intelligently varying input parameters according to some justifiable reasoning and with some assessment of probability.
Even some of the state of the art classification algorithms which process hundreds of thousands of scenarios require someone with a feel for the scheme, choosing input variables and algorithm parameters (such as number and depth of trees and forests in the random forest algorithm for example). You vary all of these parameters, and check the results against training data sets to determine the accuracy of the classification (which would be "winner" in this case) But all of these depend on huge training sets, which we don't have for national elections. So if Silver is doing this, his result is very weak and more dependent on his assumptions than on the technical classification algorithm.
But I don't believe that he's manually altering dependent variables on hundreds of thousands of trial runs, not with any probability other than "random". I don't think that he has the training sets for a big data network-style classification, So I believe that all that's left are his own assumptions, and as noted earlier I have reason skeptical of his bias in that regard.
There are a lot of what if questions that can be asked (or modeled) in his scenarios. What level of support does each candidate get from whites, blacks, men, college graduates, etc.? How much variability is there in various types of support? How likely are different groups to actually vote? And so forth.
What I suspect Silver does is set up models for the various groups and even subgroups and then feeds those results into his master model. And does this hundreds or thousands of times.
But Bush, McCain and Romney were much closer to each other in how the public viewed them than any is to Trump. Similarly, Gore, Kerry, and Obama were close to each other in how the public viewed them than any is to Clinton. Thus, Silver could use historical data fairly well in predicting how uncertain or volatile his averages were, but with these two candidates those old measures of volatility may be useless.
26 comments:
That tweet fell far short of the showy rollout in a midtown Manhattan hotel that was originally planned.
Seems like the Trump campaign finally decided they could not afford to p. o. the establishment portion of the GOP that had its heart set on Pence.
Anne Coulter and many others of her ilk will not be pleased.
Hillary Clinton has four foundational issues that advertising spending cannot overcome: Authenticity, Honesty, Likeability and Competence.
https://theconservativetreehouse.com/2016/07/14/rasmussen-national-poll-donald-trump-44-hillary-clinton-37/
Strange that Hillary can be so tremendously disliked and yet---
"It’s Still Clinton’s Race to Lose"
at political-wire.com
and she has a two to one likelihood of defeating Trump
according to Nate Silver's FiveThirtyEight and
Sam Wang's Princeton Election Consortium.
LOL, She is losing ground and in many battle ground states falling behind.
James is the perfect Socialist , he cries that Citizen United is unfair, yet, having Hillary dumping a water fall of cash into Advertizing is ok.
Jane does it bother you, that the money she spends is from Huge Wall Street Hedge Funds and Banks?
Attorney General Loretta Lynch that the “most effective” weapon at America’s disposal against Islamic terrorism is “love.”
She is perfect, saved Hillary from Charges that any other American would have been charged.
Her idiot Boss Obimbo feckless and a disaster agrees with the AG.
It is a god damn shame .
Things are not as rosy as CH wants them to be. Another group of polls showing some interesting data, kinda like the 0% black support for trump yesterday that CH with his infinite wisdom discounted the polls as bs. Oh well, can't have your cake and eat it at the same time. LOLOL This is getting to be a lot of fun for someone like myself who enjoys watch CH spin like a top when it comes to serving cakes.
Hillary Clinton is ahead of Donald Trump in several key battleground states, a series of NBC/Wall Street Journal/Marist polls of registered voters released on Friday found.
Clinton held an 8-point edge in Colorado, according to one poll — a similar margin to other surveys released in the last several weeks.
The former secretary of state garnered 44% support to Trump's 37% in Florida, the battleground state with the largest number of electoral college votes.
Clinton held a 6-point lead over the real-estate magnate in North Carolina, and a 9-point advantage in Virginia.
"With 66 electoral votes at stake in these four states, Donald Trump is playing catch-up against Hillary Clinton," Marist College polling director Lee Miringoff said in a release.
The polls come days after a series of surveys showed Trump gaining on Clinton in states like Pennsylvania and Ohio, which some observers attributed to FBI Director James Comey's critical assessment of Clinton's handling of classified information in emails while she was secretary of state. Friday's NBC/WSJ/Marist polls, however, were largely conducted after Comey's announcement.
Of more interest is that there's still 0% black membership in Bill and Hillary's old country club in Little Rock Arkansa.
Actually Opie...
Did you see the LA Times came out with their first tracking poll today? As much as they tried and tried in their article to excuse the results, it shows Trump up 3 points.
That's three national polls in three days without a lead for Hillary.
btw... the NBC/Marist polls have been on the heavy side for Clinton. The previous NBC/Marist poll in Pennsylvania for instance, showed Hillary leading by 15. It came out at the same time three other polls came out. Those three showed the race tied, a one point lead, and a three point lead. The Marist poll was nearly 14 points higher than the average of the other three.
First governor picked as a running mate since Agnew.
Not sure when (or even if) it ever happened before Agnew.
Hmmmm. It seems there were quite a few back in the day. Coolidge was the last before Agnew. Teddy Roosevelt was another. Tyler, Van Buren, and Jefferson were all governors who became vice presidents (although Jefferson was "elevated" as the Electoral College runner-up).
http://governors.rutgers.edu/on-governors/us-governors/governors-and-the-white-house/governors-who-became-vice-president/
Doh! I totally forgot about Palin.
She sure helped McCain. You betcha.
Eh. I see Earl Warren was a governor when he was Dewey's running mate in 1948.
I guess while having a governor as a running mate is uncommon, it's not all that rare.
As much as they tried and tried in their article to excuse the results, it shows Trump up 3 points
And your opinion is duly noted. BTW, who cares that you think they tried to white wash a number. The bottom line is you liked the results they reported, which you still managed to criticized. Wow, imagine if they didn't. LOL!! Dayum CH, your bias is in full bloom. Plus I'd bet those numbers will change many times in the next few weeks, especially after the RNC convention that will be a disaster. Oh, well.
he NBC/Marist polls have been on the heavy side for Clinton. "
BTW, any poll that shows Hillary up is suspected by you to have bias and is NFG. That sure is amusing to someone like me. LOL!
Issa Wants to Shut Down Government Over Clinton
Rep. Darrell Issa (R-CA) told Brietbart News Daily that now might be a good time for the Republican leadership to shut down the federal government, in protest of what he called “an imperial president” who will not “enforce criminal charges against a criminal.”
Issa was referring to the lack of charges filed against Hillary Clinton for her private email server.
Said Issa: “We should be willing to shut down the government if the president won’t limit his power.”
___________
He should go against the recommendation of the Republican head of the FBI?
The new Trump-Pence logo is getting peals of laughter.
It looks like Trump has or is screwing Pence.
You can see it at electoral-vote.com.
“I don’t know how women can vote for him… It’s incomprehensible to me.”
— Barbara Bush, in an interview with CBS News, on Donald Trump as the Republican nominee.
I looked at Silver's opinion on the fivethirtyeight blog, and apparently I was right that he's making some assumptions about swing states and battleground states.
I quote: "It’s not quite correct to characterize the race as a tossup, however. A relatively emphatic majority of recent swing state and national polls still have Clinton ahead, but often by narrower margins than before." but the degree of uncertainty is "High, or perhaps very high, for a variety of reasons." he says.
So his analysis, based on slim and dwindling leads in swing states, is highly or very highly uncertain - which necessarily means that he is relying on assumptions about them. It's rather foolish, IMO, to try to distill such small differences having high uncertainty into such precise number as his "37% likelihood" of a Trump win, or even your 2-1 odds favoring Clinton.
Nate Silver betrays a certain amount of personal bias by dismissing certain poll trends as "fairly normal fluctuations" yet relying heavily on those polls for his pseudo-precise handicapping of the race. Indeed, Silver characterizes some of the literally most important factors (everything "apart from head-to-head polls") as something he worthy of no more than "a love-hate relationship". I know what it means to be a primarily numbers guy, since I'm one of those myself, but Silver's whole approach on things leaves me skeptical, with "High, or perhaps very high" uncertainty regarding his methodology.
The new Trump-Pence logo is getting peals of laughter.
Just goes to show you there are far too many lonely, obsess people with far too much time on their hands.
Including an unemployed pastor.
WP, Silver has always relied on assumptions in his predictions. He runs hundreds or thousands of scenarios, and when he says "Trump has a 33% chance of winning" he means Trump won 33% of the scenarios he ran. I do like that he's acknowledging the uncertainly in his underlying assumptions this race creates. In previous elections there really hasn't been a heck of a lot of difference among each party's nominees so there was a lot of historical data to work with in those scenarios. But both Trump and Clinton fall outside the norm for the previous four elections, bringing new factors into voters' decisions.
I'm seeing articles mocking the TP initials of this team.
Memo to HRC: do not choose Russ Feingold as your running mate.
Go to the next thread up and look at the "underlying differences" between Trump and Pence in my last post at 5:48PM.
Indy, I don't believe that there's any meaningful way of "running hundreds of thousands scenarios" without intelligently varying input parameters according to some justifiable reasoning and with some assessment of probability.
Even some of the state of the art classification algorithms which process hundreds of thousands of scenarios require someone with a feel for the scheme, choosing input variables and algorithm parameters (such as number and depth of trees and forests in the random forest algorithm for example). You vary all of these parameters, and check the results against training data sets to determine the accuracy of the classification (which would be "winner" in this case) But all of these depend on huge training sets, which we don't have for national elections. So if Silver is doing this, his result is very weak and more dependent on his assumptions than on the technical classification algorithm.
But I don't believe that he's manually altering dependent variables on hundreds of thousands of trial runs, not with any probability other than "random". I don't think that he has the training sets for a big data network-style classification, So I believe that all that's left are his own assumptions, and as noted earlier I have reason skeptical of his bias in that regard.
I said hundreds OR thousands, WP.
There are a lot of what if questions that can be asked (or modeled) in his scenarios. What level of support does each candidate get from whites, blacks, men, college graduates, etc.? How much variability is there in various types of support? How likely are different groups to actually vote? And so forth.
What I suspect Silver does is set up models for the various groups and even subgroups and then feeds those results into his master model. And does this hundreds or thousands of times.
But Bush, McCain and Romney were much closer to each other in how the public viewed them than any is to Trump. Similarly, Gore, Kerry, and Obama were close to each other in how the public viewed them than any is to Clinton. Thus, Silver could use historical data fairly well in predicting how uncertain or volatile his averages were, but with these two candidates those old measures of volatility may be useless.
Post a Comment