Good polls/Bad polls

[PresidentSamTilden]

It looks like 2016 will go down in the popular memory as a great polling miss. In reality, though, the national polls were off less on average than 2012. Many polls herded to +4 at the end, and it looks like the final result will be ~ +2. Hillary was overestimated by a hair, but nothing crazy. The winner, and his strength in the Midwest, were the true surprises.

Anyhow, we all spent some time following these polls. I thought it might be a good time to "shout out" some of the polls that nailed it, and shame some that were way off. Here's a few I've looked back on:

Good Polls:

Rasmussen Daily: Not taken seriously by many after 2012, their final result of +2 Clinton will probably be the most accurate of any poll. Surprise!

Mcclatchy/Marist final: This poll went against the herd and called for a Hillary +1 result. Not too shabby.

Trafalgar: Went against heavy consensus and called MI and PA for Trump.

Selzer/Bloomberg, FL, T+2: Was dismissed by many as an outlier, ended up being the truth that we didn't want to hear.


Bad Polls:

W&M/Target Smart Florida, Clinton +8: Technically, this nailed Hillary's vote share, so that's nice lol. Trump's, not so much.

Columbus Dispatch, Ohio, Clinton +1: I was curious to see how this one would do. Well, it missed.

Marquette, WI, Clinton +6: Muh gold standard! Thanks for the complacency, guys. I suppose Hillary's vote share was accurate, though.

Alaska Polls: Most had a close race or even Clinton leading. Nah.

UNH, NH Clinton +11: Wow.

LA Times Tracker: Final poll had Trump winning popular vote by 3 points. Pretty much everyone thought it was off the whole way...and it was.

Special Mention:

Nate Silver. Blasted by many for sewing doubt and for not giving Hillary a 100% chance to win, he comes out on top of the aggregation game once again.

Feel free to add any polls that you think deserve a second look, or criticize my placement of these, lol.

[boske94]

Bad polls:
National- one day before election
LA Times/USC Tracking Trump +3 on Nov 8th
Monmouth Clinton +6

State polls:
CBS/You Gov       Ohio               Trump +1
Mitchelll/FOX2     Michigan         Clinton +5
Loras                  Iowa, Wisconsin  Clinton+1, +6
Morning Call        Pennsylvania   Clinton +4
CNN/ORC            Nevada           Trump+6

[Rules for me, but not for thee]

What's funny is the much maligned LA Times poll was probably the most accurate in the end, whether by sheer luck or better methodology I'm not sure.

[Dr. Arch]

What's funny is the much maligned LA Times poll was probably the most accurate in the end, whether by sheer luck or better methodology I'm not sure.

Except it wasn't? They missed the popular vote by 5 points.

[PresidentSamTilden]

Except it wasn't? They missed the popular vote by 5 points.

Yep, this was actually probably the least accurate, lol. Except for the St. Leo poll, but I don't think anyone took it seriously...

[muon2]

I used a filter on HuffPo based on nonpartisan polls of likely voters only. A 2.5% margin for HRC was what it was predicting. If you account for the undecided and excess Johnson/Stein vote, the predicted margin is a reasonable match to the result.

If you use HuffPo 3-way and limit the polls to nonpartisan pollsters and likely voters the current numbers from their trend lines are Clinton 45.1%, Trump 40.7%, Johnson 5.1%. For the 2-way it gives 45.8% to 42.4% (moderate smoothing)

Why would you do any of this?

This in response to posts on both sides complaining about inclusion of partisan pollsters. HuffPo lets me filter them out so I did this as a point of comparison.

By further comparison that same exercise today has Clinton 45.2%, Trump 42.5%, Johnson 4.9% in the 3-way and Clinton 45.4% to 42.8%  in the 2-way. Clinton's share in both is essentially unchanged from my post 4 days ago.

The Trump line in my exercise is still rising but more slowly than it did last week.

3-way: Clinton 44.5%, Trump 43.6%, Johnson 4.4%
2-way: Clinton 45.3%, Trump 43.6%

The Clinton data is still very steady as it has been over the last month. Note that when Johnson is not included Hillary gets a quarter of his vote, but the rest goes to undecided. To eliminate recent noise and movement between undecided/third party I can put on more smoothing on the two-way results and she is at 46.0%. However, over time Trump continues to gain at the expense of Johnson, but the trend lines don't show him catching her by Tuesday.

I think 538's model is reacting like the trendlines I'm seeing with HuffPo's tool. The Trump poll numbers have stopped moving up so his trendline comes down and hers moves up. Here's what it gives me this afternoon.

3-way: Clinton 45.5%, Trump 42.9%, Johnson 4.5%
2-way: Clinton 46.3%, Trump 43.3%

[PresidentSamTilden]

Yeah, that's pretty good. What polls did it filter? LA Times and SurveyMonkey?

[peterthlee]

In retrospect, Columbus Dispatch nailed OH Senate race exactly for Portman 58-37. Sad!