patrick.net

From today:

he he he

I think there is a non-zero chance that Trump drops out in the next month. And Paul Ryan is the next President.

Tenpoundbass says

There's just no way Trump would be polling at 15% cut it out.

Do you not understand what that graphic is saying? It's the probability that either candidate wins the White House based on state polling data. It's not saying that Trump is polling at 15%, it's saying that Trump is behind (3-7%) in most swing state polls and therefore has a small chance of winning enough states to become President.

This one kind of lags.

Remember last week when Trump was winning ? Sigh...those were the days...

dont forget assange says he still has emails that will put Clinton behind bars.

You'll crawl back in your swerer when you are proved wrong again.

Ironman says

Ha Ha ha, clueless, as usual.

The current polls are OVER SAMPLING Dems by 10% - 20% to get Hillary a few point advantage, yet, in reality, there are approximately the same percentage of Dems/Repubs in the country.

And we've come full circle. Despite claiming that polls were "skewed" in 2012 before the election, and then seeing that the polls were actually biased FOR Republicans, CIC is going to start all over with the polls are biased again. How idiotic are you? Do you ever learn anything?

Today's update:

Arizona and Georgia ready to flip next.

http://dailycaller.com/2016/08/04/shock-poll-trump-at-over-30-percent-with-black-vote-in-north-carolina/

Trump At Over 30 Percent With Black Vote In North Carolina
A poll released Wednesday of likely North Carolina voters shows Donald Trump leading Hillary Clinton while at the same time gaining 32 percent of the black vote.

Ironman says

It's just factual numbers, tatty boy, from the major polling companies, but I wouldn't expect you to understand them based on your 2nd grade math skills.

Yep--and those numbers have proven to be VERY accurate predicting actual voting totals. Any bias is FOR Republicans.

It's hard to believe, but you may wish to at least consider that companies that do this for a living and depend on accuracy of results may have better and more sophisticated algorithms than assuming the country is 50/50.

zzyzzx says

Then how come nobody is showing up for Hillary rallys to the point where she has to cancel them?

Same reason you can have a huge KKK rally next to a nearly empty public library. Would you conclude that therefore hate is better than knowledge ?

marcus says

Same reason you can have a huge KKK rally next to a nearly empty public library. Would you conclude that therefore hate is better than knowledge ?

That's not an answer? You don't have an answer so I will have to give it to you. It's because nobody really likes Hillary.

Yep, 538 now-cast is saying about 10 to 1 odds against Trump winning if the election were held today.

zzyzzx says

marcus says

Same reason you can have a huge KKK rally next to a nearly empty public library. Would you conclude that therefore hate is better than knowledge ?

That's not an answer?

I thought it was a perfect answer. I like Hillary infinitely more than I like Trump, but I'm not someone that would go out of my way to go to any political rally, especially if I could see it on you tube.

What,... you think that because Trump has rock star or messianic following in some circles, more so than Hillary, that makes him a better candidate ? That's off the charts stupid. Even if the average IQ at a Trump rally was over 100 this argument wouldn't hold water. But we both know that isn't the case.

http://www.nytimes.com/video/us/politics/100000004533191/unfiltered-voices-from-donald-trumps-crowds.html

Surely it can't go much further than 1/11 that is 10:1 odds in favor of Hillary, before the end of the week. Right ?

marcus says

Surely it can't go much further than 1/11 that is 10:1 odds in favor of Hillary

In all seriousness, there are signs Trump has bottomed out, and is rebounding.

He has doubled his support among blacks: Marist has him at 1%, but only 3 hours later, NBC shows him at 2%.

TRUMP WHINING=TRUMP WINNING

Ironman says

Really?? They can actually control who picks up the phone and answers their polls? Wow, talk about power!

Instead of continuing to spew your lies and dis-information, why don't you dig into the recent polls and look at the data of the sample.

Nah, you won't do that, it will interfere with your delusional narrative.

I can't even believe you are this big of an idiot. Do you think they just call 1000 people at random and report the results without adjusting the data?

Ironman says

tatupu70 says

Any bias is FOR Republicans.

Do you think reporting the same lie over and over without proof makes it true?

No, I think showing data makes it true. Polls are overall very accurate, but any bias was favoring Republicans.

Sure enough--Arizona and Georgia flipped. Next up: South Carolina, Utah, and Texas.

I love the fact that Hillary is counting Jill's unhatched eggs in her basket. There's going to be some Nader tears November 9th...

I would like that.

Ironman says

So Einstein, please tell us why there was an oversampling of Dems in the CBS poll, which results in a higher percentage for Clinton.

Not sure how many more times I can explain this to you--I'm afraid it's over your head. You see the column where they talk about "weighting"? That's where the pollsters (guys who do this for a living) apply proprietary algorithms to adjust the results so they represent the overall population. You're basically saying that these guys don't know how to count to 100. Despite the fact that looking at polling results vs. actual results shows that their results are actually quite good. Meaning their weighting is correct. And you continue to be an idiot.

Tenpoundbass says

I love the fact that Hillary is counting Jill's unhatched eggs in her basket. There's going to be some Nader tears November 9th...

I would like that.

She doesn't and doesn't need to. This graphic shows the average from a lot of national polls. It lags and hasn't settled yet after the craziness of this past 1.5 weeks. But I'm posting it for you to observe that there is a reason why the percentages add up to 88.1 . In spite of the fact that a lot of people simply dislike Hillary way less than they dislike (or fear) Trump, it's still always true on election day, that most people choose to vote for someone that can possibly win.

Better picture of Amish billboard:

Ironman says

tatupu70 says

You see the column where they talk about "weighting"? That's where the pollsters (guys who do this for a living) apply proprietary algorithms to adjust the results so they represent the overall population. narrative pushed by the MSM.

There, I corrected it for you.

The problem with that narrative is that the polling results in the last election (when folks like you were saying the same thing and were "unskewing" the polls) were spot on. Accuracy was good. Predictive value was good.

Face it--your guy is losing now. It may change. And if it does, the polls will change too.

Time to get the supremes on the Monaco diet...

tatupu70 says

Face it--your guy is losing now.

Ironman says

Please get out of your time warp. This isn't the "last election" from years ago. All your Libbie MSM sources, Nate Silver (and you) have been completely wrong this cycle. It's nice you want to reflect back 4 or 8 years ago to try and save face, but your smoke screen isn't working.

Nate Silver has been 100% correct with his predictions based on polling. Because the polling has actually been quite good through the primaries. As it was in 2012.

I know you need to grasp at straws to find some hope. So, I won't continue to burst your bubble. Go ahead and pretend that the polls are biased. Enjoy your playtime.

Wow--talk about dishonest. How about you post my whole quote?

tatupu70 says

Nate Silver has been 100% correct with his predictions based on polling

When you have to resort to misquotes you are really losing it.

Well, Nate doesn't really "predict" the winner. The site gives odds on who will win based on several models. Nate saying Clinton has a 93% chance of winning isn't him predicting she'll win. It's saying that based on the model, she'll win 93 times out of 100. So, if there were 100 contests in the primary (with those odds), you'd expect seven where the 93% favorite lost.

You'd scream that he was wrong on all seven of those results. When, in reality, those 7 results prove he was RIGHT

I'm laughing. You continue to post links that agree with me. Please continue.

Backpedaling? Are you kidding? Like I said--you are proving my argument and making yourself look silly.

You have to resort to purposely misquoting me to have any chance.

I love it. You double down on your despicable behavior of purposely misquoting. You really are a sad person.

Ironman says

Misquoting, except WaPo and CNN disagree with you.

You really are a joke!

Nope--they agree with my actual quote:

tatupu70 says

Nate Silver has been 100% correct with his predictions based on polling

Which is why you have to misquote me. Otherwise you'd look like the usual idiot that you are.

Ironman says

Despite the pretense of scientific detachment, Silver’s models are hardly unbiased. The moment you decide to weight some data sets over others, you’ve introduced bias. Silver’s failed Polls-Plus model incorporated indicators that had virtually no predictive value this year, like endorsements and fundraising totals.

Lies, Damn Lies, and Statistics.

Like I said, we're entering a new political system, the 6th or the 7th depending on how you count it. Models for the previous system will yield incorrect results. Or...

"Past results do not equal future returns."

Modern Liberalism - From Teddy to Carter - is going to return. Neoliberalism is on the way out.

tatupu70 says

Well, Nate doesn't really "predict" the winner. The site gives odds on who will win based on several models. Nate saying Clinton has a 93% chance of winning isn't him predicting she'll win. It's saying that based on the model, she'll win 93 times out of 100. So, if there were 100 contests in the primary (with those odds), you'd expect seven where the 93% favorite lost.

Yes, exactly.

It's saying that based on the model, she'll win 93 times out of 100. You might add "she would be expected to win 93."

And to take it one step further, if there were 100s of times that there were 100 contests, sometimes she would win all 100, other times she would win less than 90, but the average number won would be 93. That is, if he is absolutely correct, which he isn't guaranteeing, it's simply his best estimate.

Btw, this is the aspect of statistics that's hardest for students to grasp, and hardest to teach (given time constraints). For further understanding, read about "sampling distributions."

thunderlips11 says

"Past results do not equal future returns."

This is correct. I imagine that Mr. Silver is busy analyzing his model and deciding how to improve it.

thunderlips11 says

Modern Liberalism - From Teddy to Carter - is going to return. Neoliberalism is on the way out.

That would be wonderful

Ironman says

FiveThirtyEight’s homebrew “Polls-Plus” model, which weights several factors based on a secret formula, has been worse at predicting outcomes than a weighted average of the most recent polls.

That's why Silver has a polls only model and a polls plus model. What's your point?