re: Nate Silver issues Blue Tsunami Warning!

quote:

---

I think Nate’s problem is his formula is based entirely on media polls.

---

His models include internal polls, consulting polls, campaign donations, different turnout models, historical trends, and does a great job of grading/weighing particular polls based upon previous performance (this is all available on his website).

As far as 2016, he had Hillary at something like 70-75% on election night which, considering Trump won by 78,000 votes spread across WI, PA, and MI, I'd say he basically nailed it. 75% isn't 100%. He was also one of the few statisticians to annotate a very high chance of a Dem popular vote victory and Trump electoral victory. He was pretty much dead-on with Hillary winning popular vote by 3%.

quote:

---

They're really just hurting themselves with these type proclamations.

---

yeah, Nate Silver and other pre-election "polls" are a big reason DJT is president.

I wonder if anyone really trusts them anymore though. It's increasingly hard to get a representative sample even via mixed methodology polling.

quote:

---

and does a great job of grading/weighing particular polls based upon previous performance

---

Disagree. His track record is not that great, especially THIS far out.

Go look at his 2016 predictions on 10/23/16 and tell me "he does a good job"...he doesn't.

The only thing Nate Silver does a good job at is hedging his bets with other peoples information the closer it gets to the election.

quote:

---

I don’t think the media polls have figured out a way to poll trump supporters. Partly because most don’t answer polls

---

How does one even become a part of these polls? I've never once been asked to be a part of a political poll or survey

quote:

---

Partly because most don’t answer polls

---

Or we tell them we are voting for the commie.

That way it makes the commies over confident and pisses off the republicans.

win\win

You can check all Nate's 2016 forecasts out here.

LINK

For the particular states, just hover your mouse over them to see the chances of a win for each candidate.

I get why Nate isn't a beloved figure in republican circles....he's goofy, he's gay, he's unabashedly liberal. That said, he's a stats guy through and through and his track record has been much better than most (Larry Sabato is the only name that comes to mind who may have a better track record than Nate). His methodologies are an open book and free for anyone to see and evaluate.

Most importantly, as a statistician he doesn't deal in absolutes. Yes, the dems have a great chance of taking the House this cycle. Yes, Nate says it's about an 85% likelihood it happens. But it also means there's a 15% chance it DOESN'T happen. If the election is held 7 times, his extensive modeling shows that the Dems would likely take the House in 6 of those elections. In addition, his numbers are snapshots of current trends. He's not saying that the dems WILL have an 85% chance on November 6th. He's saying if the election were held TODAY, the dems would have that 85% chance.

He's really no different than a Vegas handicapper. You take all the data available and make your best guess of what's going to happen. Any sports bettor knows Vegas handicappers know what the frick they're doing...Nate is no different.

quote:

---

he's goofy, he's gay, he's unabashedly liberal.

---

This has absolutely NOTHING to do with what anyone in this thread is talking about, and why would republicans care someone is gay?

quote:

---

his track record has been much better than most

---

Again, no it's not.

quote:

---

he doesn't deal in absolutes.

---

Then how can he have such a good track record?

quote:

---

But it also means there's a 15% chance it DOESN'T happen. If the election is held 7 times, his extensive modeling shows that the Dems would likely take the House in 6 of those elections.

---

So he can't technically be wrong...EVER. How do you know he's anymore right versus someone that says 50/50?

His odds for 2016 control of the house leaned the wrong way, his odds for 2016 control of the senate leaned the wrong way, his odds for 2016 presidential election leaned the wrong way...yet he has a great track record because he didn't chose the numbers 0 or 100...

quote:

---

In addition, his numbers are snapshots of current trends. He's not saying that the dems WILL have an 85% chance on November 6th. He's saying if the election were held TODAY, the dems would have that 85% chance.

---

Here's where the real issue with Nate comes in, the tightening of his polls is unjustified given the information he releases about his "models" and from a academic/statistical standpoint it is bordering disingenuous hedging.

quote:

---

He's really no different than a Vegas handicapper.

---

dude you have no idea what you're talking about and you 100% have no idea how Vegas handicappers work

Nate said in his final prediction in 16 that Hillary had an 81% chance of winning. Nate is not good at his job.

Lol