Nate Silver’s analysis of poll-herding leads me to trust polls less than ever, yet, I have come to believe that it’s a fool’s errand to bet against the average of polls. This year, the correct bet was to assume that virtually all the polls were way too favorable for the Democrats, but there really wasn’t any way to guess that they would be slanted that way rather than in the other direction. In fact, my bet was wrong. Based on the available information, I thought it was very unlikely that the polls were biased against the Democrats.
My predictions were based on the assumption that the average of polls was pretty accurate but was modestly undercounting Democrats. This led me to predict that Democrats who were narrowly leading in the polls would, in fact, go on to win. And, in at least one case, it led me to predict that a Democrat who was losing would nonetheless win.
I had no way of knowing that the pollsters were deliberately making the races look better for the Democrats in order not to have their polls look like outliers.
Polls take the temperature of an electorate. But that electorate is not a predictable entity until the vote happens.
Doesn’t turnout play a huge part in whether the polls will have been accurate or not?
You could have a candidate who is polling well out in front, but if turnout goes against him/her then the poll will have been wrong. But then again it won’t have been wrong at all; it’s just that people did not vote in predictable numbers to reflect the averages.
Exactly.
What was the weather like on election day around the country? Given the low turnout, it wouldn’t surprise me if rain and cold here and there were enough to dissuade a substantive amount of voters from coming out to vote.
It is why all elections should take place during election week. And why you should be taxed for staying at home. Go vote for Garfield or Captain America if you really don’t want to be a citizen. It’s better than being a corporate victim.
It’s a funny thing about the seers who read the entrails of goats. They are remarkably accurate, except for those times when they are not.
Polls are fun, as were the drug addled ravings of the Delphic Oracle. But, even with all their scientific screenings, polls are only marginally better at predicting the future than she was.
As Bob Dylan put it, “you don’t have to be a weatherman to know which way the wind blows.” The palpable disinterest of Democratic voter coupled with the negative October Surprises (ISIS and Ebola) meant that Democratic losses were predictable without annoying people with dinnertime phone polls.
One of the reasons that I kept talking about the width of the variance is that that band is the prediction, not just the mean (expected) value. Errors outside the top and bottom limits of that band demand explanations other than random variation.
Herding is one explanation, which just so happens to make Selzer’s methods look very good. It also happens to legitimize the election as being devoid of voter suppression so we can move on.
This makes me think that it will be very difficult to determine the extent to which voter suppression tactics by the GOP actually worked and the extent to which failures of message, candidate, and activity by the Democrats caused defeat.
Thinking that the GOP got away with another stolen election delegitimizes the electoral process and reduces turnout in elections. Aside from anectdotal stories, sworn complaints, it is very hard to assess how widespread electoral fraud and voting suppression were this year. The report of Interstate Crosscheck having some 6 million names potentially dropped at the last minute from the voting roles is troubling. How would you know how many people in this group were not able to vote and just fell silent, except for randomly showing up in a polling sample?
Under conditions of political intimidation, voter suppression and electoral fraud, the opinion polls have become worthless for campaigning or prediction. Even as they miraculously predict the correct results, except for all the other polls.
More fundamentally for the future, discussion of the polls too much not distracts from discussion of the issues and sets expectations that themselves suppress voter turnout. Likewise the early voter readouts, which actually tell you zero information even when there are party labels attached.
In the last few cycles – maybe back to 2002 IIRC – there’s been a bias against the party that “won” the election. It could be that discouraged voters are over-reporting their actual desire to vote – or perhaps encouraged voters are under-reporting.
Non-presidential elections are notoriously hard to poll. And when turn out drops as far as it did in 2014 plus given the structure barriers, e.g. mobile phones, to getting an accurate sample polling becomes little more than a WAG.
CBS Sunday Morning this a.m. ran a brief graph, front-ended and back-ended by 5 ads, about the extraordinarily low voter turn out, providing the facts behind the assertion: number of registered voters versus the 1/3 that actually voted and another about this being the lowest turnout since maybe 1972?
So, I thought, here’s CBS doing a little bit of real journalism albeit that they provided no explanation for their visual graph and texts.
O/T but has anyone noticed that all the political operatives, elected and otherwise, who’ve been on TV since Nov 4 are largely conservative GOP ones? I flip by when they bring the next Obama nay-sayer on cause I just can’t take it anymore. Even MSNBC’s Chris Matthews is giving equal time to the GOP post election. As a frequent BJ commenter said recently, Matthews is a schmuck.
By golly, we Dems neeed to do something about this media disparity. it’s going to be difficult enough to climb back from this disaster of an election and the solution is NOT for the various Democratic fund raising arms to continually bombard my inbox with requests for funding and/or petition signings.
Whatever happened to reality? I don’t care if the polls don’t reflect my (admittedly skewed) world view. I want to know the truth as far as these guys can discern it.
Anyone can be wrong, but don’t be wrong deliberately.
I said as much before the election.
But a big problem is that except for outfits like the Register who have an interest in being right, most pollsters have an interest in getting repeat business.
Or is their interest in selling more advert time?
If “getting it right” were the primary concern, they aren’t very good at their job.
The Register? Their polling brand is about being better at polling Iowa than anyone else. Being right accomplishes allowing them to charge higher ad rates becausece it makes them more famous a gets more eyeballs.
Maybe instead of watching the polls, we should watch the adverts. Two most effective, IMHO, were Joni Ernst’s and Raul Ruiz’s.
There was little Ruiz’s opponent could do to reduce the appeal of Ruiz in that ad. This is not a “blue” House seat, but has long been red: Sonny Bono, Mary Bono.
Ernst’s ad is exceptionally good. But an equally good answer ad was possible.
If you watched the Thom Tillis ads, you saw that he was pitching the audience of Hulu craft channel and who knows where else that he was the NoBama. And the Tillis ad saturated based on a North Carolina IP address.
But the ads hide the person. The Joni Ernst ad was a key example.
Watched one of the Tillis ads — it was okay, but nowhere near as good as Ernst’s.
Not so much hide the person but hide exactly how what they stand for means in the real world. That’s why an effective answer to their ads wasn’t the difficult. More so if the answer ad is aired by other than the campaign. And once the campaign staff sees the “answer” ad, they follow up on it with a candidate ad that cleverly reinforces the answer.
Cory Gardner’s ad was better than most, mostly fluff but did have a bite that likely resonated. The fluff was more on the order of John Edwards’ “I’m a son of a mill worker.” The bite was that Udall “inherited” a seat in the Senate. Wish Democrats would pay attention to this — legacy candidates weren’t favored in 2014 and they won’t be in 2016. For that reason, the GOP will not run Jeb! They have a deep enough bench that they don’t have to.
The Democrats are extremely incompetent at fostering new political blood. Newcomers succeed because of their own luck and skills. It’s like Joe Mauer, he’s so talented even the Twins coaches couldn’t ruin him, but they sure could ruin other players like David Ortiz.
Neither party is particularly good at that. It’s likely more that the GOP recently got lucky. In part because the top of the heap has been somewhat empty for some time now. Take the legacy guy GWB out of the equation and they had nobody from 1992-2012. Still aren’t a large number that can step up to higher levels. What will prop them up are the okay looking/sounding loons that a lot of money sloshing around in the GOP managed to turn into higher elected officials. Majority parties at the State and Federal level do offer an advantage for their POTUS nominee. Unless they are reviled by the next election — and considering that Brownback, Scott, Snyder, and Walker were re-elected, doubt that will happen within two years.
To me that’s basically the definition of fostering new blood. Cruz is going to be a shadow majority leader, Walker has crushed every challenge against him. The Democracats? Its so much legacy and seniority.
Harry Reid has been a terrible leader and has been humiliated by McConnell at keast twice as often as the reverse, but even after thus debacle who will oppose him? Where are challengers to Schultz and Israel’s positions in charge of recruiting and electing House Dems?
Right, like famous scion of a noted American family Barack Hussein Obama.
Did you completely miss the phrase own luck and skills? Joe Mauer is (was?) a supremely talented athlete, intelligent, morally upstanding at least so far. I think we the comparrison to Barack Hussein Obama almost writes itself.
Don’t you appreciate that one advantage Obama had in 2008 was that he WASN’T a legacy candidate? No daddy or other relative to pave his way through life and into high public office. It was still and advantage in 2012 against the legacy candidate Romney that in several tries only managed to get elected to a single four year term in public office.
Pollsters, like humans, are just natural born liars.