28% is indeed quite a different number than 44% but it still reveals a horrid methodology as do the other cross tab numbers Nate cited.

Tbh though I’d still be willing to wager on the under for 28% on black support in Michigan for Trump.

I'd also take < 28% of black support in MI for Trump.

I'd also take JJ vs AQ

The crosstabs Nate Silver is linking to seem to have been taken down. I did find this previous Trafalgar Michigan poll

https://drive.google.com/file/d/1bfV...9ZIThV_7x/view

Response Rate: 1.37% - in other words "We called 75000 people and here's what the 1025 people who picked up and completed the survey had to say".

This might explain some of the wilder results posted earlier.

Also I checked out their numbers for Pennsylvania 2016:
https://projects.fivethirtyeight.com.../pennsylvania/
Their survey was done from 3-5 Nov. Similar dates were

3-5 Nov You Gov Clinton +2
3-6 Nov Gravis Clinton +6
1-7 Nov SurveyMonkey Clinton +3
3-5 Nov Trafalgar Trump +2
2-6 Nov AA/ESA/POR Trump +2
1-7 Nov Google Tie

Actual Result: Trump +0.7%

So in fact google were closer to the result.

For Michigan they polled Nov 6 and said Trump +2 when it was Trump+ 0.2% (i.e. they were as far off as the national "polling miss" that is often criticised) but the only other poll taken that late was the D-rated Mitchell Research one which gave Clinton +6 so they win the direct comparison. (although Google did 1-7 Nov and nailed it as a tie).

Yes, and isn't it the case that for these polling companies, politics polling is mostly about marketing themselves for their actual business of selling polling services to companies for non-political polling. It would seem to make sense that they would want to stand out a bit in some way.

In non-political polling, of course if I want to know if the public prefers celebrity A or celebrity B so i can decide who to get to endorse my product, it's important I find out whether the public prefers someone 75-25, or 60-40. If it's anywhere near 50-50 it means that it really doesn't matter which one I pick. That's very different to politics polling where the difference between 51-49 and 49-51 is massive.

I find the Trafalgar poll very interesting. Now that I see the demographic, I believe the poll accuracy. Only thing is, it's heavily skewed towards older voters. That's where I think it falls short of giving a completely accurate picture.

I’d probably be down for some kind of <28% black support bet. Maybe for an autographed book or something if I win.

Art of the Deal?

Or like $100k

The cross tabs were taken down because they were being ridiculed mercilessly. Ironically by taken them down they just further cemented the reality that they are a partisan grifting operation who is literally just making stuff up.

Regardless how they may claim they ask questions we don’t even know if they talk to anyone so perhaps they can be shelved from this discussion entirely.

At this point we can’t assume they are even slightly legitimate. With so many other polling outfits out there, perhaps the discussion should go that direction.

One fraudulent pollster should not derail a thread where mason is sharing his knowledge on polling in a more general sense.

Along this line it appears that Trafagar says their response rate is a little over 1 percent. I remember at Census we were concerned when the response rate went under 95 percent, and I did not make a mistake with these two numbers. So, here is an area where I would have no understanding of their methodology (unless this number was a misprint). Does anyone know the response rate that some of the other polling organizations have?

Best wishes,
Mason

Betting markets and 2+2 thread starters it seems.

If Trump wins again this huge polling houses should retire! How accurate are these polls, considering some people are shy to say they would vote for Trump when they are asked?

The Shy Trump Voter theory is going to become self-fulfilling - so many Republicans watching Hannity explain this is how Trump voters act that they are going to start acting this way.

Is it more embarrassing to say you support Trump, or to say you support an arbitrary GOP Senator? Because people seem to be way more willing to tell pollsters that they support Trump than that they support many GOP Senators, but no one is concocting wild theories about how socially unacceptable it is to support McSally.

lol WTF.

HE PUBLISHES A MODEL THAT PEOPLE CAN READ AND BET ON

We'll see if this is just the genius of Trafalgar and their extremely well-designed questions!

Get your fainting coaches ready. Mason's favorite polling outfit has been less than forthright about whether the polls they are releasing have been paid for by partisan clients, calling into question whether we are only seeing the ones they got permission to release.

I never said that Trafalgar was my favorite polling outfit. What I said was that it appeared they were using a different methodology than most other pollsters and that might explain their differences. The election itself should determine whether their methods are superior or not.

As for Silver and his comment, he seems to be implying that the purchaser of the poll is influencing its outcome. I don't think he has anyway of knowing this.

You could also argue that since Silver thinks poorly of Trafalgar, he also feels it's necessary to point out possible problems with it while he doesn't feel the same thing is necessary for those polls that he likes better. I don't know the answer here.

A simple analogy is that it's well known that purchasers of a product tend to think higher of the specific product than those who don't purchase it. But the quality of the product, whether good or bad, doesn't change.

Mason

No it won't. For reasons I explained in my previous post.

You just completely made this up. He never implies that. A ton of data from previous elections shows that internal or partisan polls are not random and have particular and predictable bias. The 538 model corrects for this.

You also completely made this up. He is not saying Trafalgar is inaccurate. He is simply saying his model needs to treat them like other internal or partisan pollsters.

That's a terrible analogy. A good one would be movie reviews that are shown during a commercial for the movie. They are always completely legit reviews, but they will obviously be biased towards critics who like the movie.

This tweet kinda buries the lede in that Mason's favorite new pollster casually says, at the start of this clip, that he thinks Trump needs to win Pennsylvania "by 4 or 5, to overcome the voter fraud that's gonna happen there"!!!

Mason, has your opinion changed at all upon hearing this guy speak and hopefully realizing that he's a conspiracy-addled moron?

If I intentionally design my polls to show my client in the most favorable light possible, and my client wins, does that mean my polling method was superior?

Sometimes donkeys get there; that doesn't justify getting it in bad.

Is the poker player who won the hand the one who played the hand the best?

LOL I came to post just that image of the bow tie dude from fox news

Well before he learned this about Trafalgar he build a election forecast model, and found that partisan polls do in fact have a partisan lean.

Not saying they're intentionally putting a thumb on the scale, but when the people who cut your checks want to hear X it can be easier to back into a methodology that leads to X.

Depends on whether or not I'm the one stacking up the chips

Is this the real Mason Malmuth posting or did we get duped by some troll account? This is hard to swallow, reasoning this bad, ugh.