All errors should be reported to

Tuesday, October 18, 2016

Most accurate 2012 poll shows Trump up 1.6 points

The Los Angeles Times poll is an outlier in the club of  pollsters. The other pollsters question its methodology of asking the same people the same questions each day. The other polls randomly call different voters each day because that is what the textbook says you should do.

But the reality is that the most accurate national poll in 2012 was the RAND Corporation poll. It had Obama winning by 3.8 points. He won by 3.9.

The Los Angeles Times poll this year is the RAND poll

From Sean Trende at Real Clear Politics:
In the end, though, the RAND poll basically got it right. The national polls (though not so much the state polls) were off in 2012. During the closing month of the campaign, they showed, on average, a 0.3 point Romney lead. The RAND poll, by contrast, showed a 3.8 point Obama lead – which was almost exactly correct.
Does that mean the Times poll will be correct this year? Absolutely not. We should treat it as one poll among many, and should note its outlier-ish tendencies. It may be worth watching for trend lines. We might also note that this cycle, it runs contrary to both the national and the state polls, and tends to be off the RCP Average by an even larger margin.
At the same time, though, we should recall that almost all of the objections lodged against the poll could have been lodged against it in 2012. Many were. The poll may well be flat-out wrong in 2016, but its history cautions heavily against dismissing it outright.
History also cautions against believing any of the rest of the polls.

They were wrong in 2012.

So wrong that Gallup dropped out of the election polling business, concentrating on the more lucrative economic and trends polling, which unlike election polling can never be proven wrong.

The RAND poll is expensive. You have to keep track of people for six months. Randomly calling people each day is cheap. And inaccurate. One poll keeps calling me asking about the Monongahela County races. That's a couple of hundred miles away. It takes a certain genius to not understand that a "755" number is not in Monongahela County.

Early in his piece, Trende wrote: "First, truth is not decided by committee. That is to say, the fact that the L.A. Times pollsters weight their poll in a different manner than other pollsters do doesn’t make them wrong."

The pollsters got 2012 wrong. RAND did it a different way and got it right.

Trende is most correct in stating RAND's success does not guarantee success this year.

But he left out the real lesson of 2012. By getting it so wrong, the pollsters showed there is nothing scientific about their polls. Might as well throw spaghetti at the wall.

Today, the Los Angeles Times poll has it Trump 44.9, Clinton 43.3.


Please read "Trump the Press," a fun romp that skewers the media experts. It's available as a paperback, and on Kindle.


  1. How can this poll be accurate? Trump said they are all rigged. Especially one done by a media outlet, and as we know, all MSM outlets are part of the Clinton conspiracy to make Donald lose. Just ask Trump or Rudy Giuliani.


    “How the USC/L.A. Times Daybreak poll works…Results are weighted to match demographic characteristics, such as race and gender, from the U.S. Census Current Population Survey, and are aligned to the 2012 presidential election outcome using self-reported votes in that election…”

    The USC/LAT poll weights its outcomes by 2012 party affiliation. This weighting process is where the dirty work is done in the propaganda-media polling. But even with 2012 party affiliations based on Obama-Romney, Trump is still ahead by almost 2%.

    But the relative level of enthusiasm between the Harpy and Trump is vastly different than that between the Obama and Romney. Seriously, who could have told the difference between Romney and Obama? One wanted open borders and the other wanted comprehensive immigration reform. One designed Obamacare and the other implemented it. Obama and the media called Romney a fascist. Romney cowered and apologized. With Trump, Americans have a real choice.

    Imagine what the poll numbers would look like with realistic party affiliation weighting. Hopefully the landslide will be big enough to overcome Democrat vote fraud.

    Steve in Greensboro

    1. I agree with your analysis Steve. Let's hope we aren't deluding ourselves.

      And this early voting for an entire month thing is ripe for rampant votefraud. That has to go -- there is no good reason to allow voting for so long, that I can think of.

    2. Can anyone provide a link that documents voter fraud is "rampant" (when, where, how) and how such fraud - if it really exists - favors solely the Democrats. I think Trump is Jim Jones incarnate the way he makes some of you believe the most outlandish lies. To be clear, Hillary is one step above scum. But Trump, what an absolute tool who would have thrown this country in turmoil in no time flat. Hillary is capable of doing the same but at least there is hope she will be gone in four years before doing what Trump would have done in months.

    3. Nice troll, tool. Your check from Soros will be arriving shortly. Did ya see the 100 page leak today? You tool.

    4. Im familiar with the illegally obtained emails being published and yet they do not answer my question and neither do you Mr. Tool. Nice try though with your simplistic Trumpian boomerang reply.

    5. I take you are absolutely and positively sure that all the stories about the dead voting in Chicago and other big cities are voting Republican?

    6. Start here but I've seen several stories like this lately, always in battleground states and what a coincidence THAT is, eh?

      It doesn't need to be rampant. They only manufacture votes where they need to, like putting your thumb on the scale.

      Its not brain surgery. Elections have always been rigged, to one degree or another. This is not news.




    10. Lots of articles about allegations. Nothing more. So who is doing the rigging in key battleground states. Smoke this you dopes.
      In Arizona, Colorado, Georgia, Indiana, Iowa, Michigan, Nebraska, Nevada, New Mexico, Ohio and Utah, the chief elections officer of the state is a Republican, elected by voters of the state. Most are secretary of state; Utah’s lieutenant governor oversees elections there.
      In Florida, the secretary of state is appointed by the state’s Republican governor, Rick Scott — a Trump supporter.
      In North Carolina, the state board of elections has five members, appointed by the governor — currently a Republican. Its current chairman and three out of five members are Republicans.
      Yep. Lots of bona fide fraud. Why aren't any of current top Republican office holders joining Trump in saying the election will be rigged?

  3. It simply boils down to how big a hidden 'Brexit' vote there is for Trump. If it would normally move polls by 3.5% or less, Trump loses. If it's 4%, as was the difference with the Brexit vote, then he'll squeak out a win/

  4. One should use the LA Times poll as a measure of the fluidity of support.

    Personally, I think at least 95% of the people who turn in a ballot have already made up their minds by Labor Day in any presidential election. If a poll is showing a swing in a candidate's support by more than 5% from one polling to the next, I simply dismiss it altogether since I assume it is a result of sampling error.

    I want to believe the LA Times poll has an accurate sampling, but it is possible the initial sample was skewed towards Trump right from the start- we simply won't know until the votes are actually counted. However, since they are polling the same people over and over and allowing those people to indicate their own conviction about what they will do in November (or early voting), you can at least determine those events that have measurably changed the dynamic without having to worry about a new sampling error.

  5. So, here is how I used the LA Times poll recently after the Trump tape hit- it measurably caused him a decline in support in the LA Times poll of about 3-4%, but still left him with a lead. I take that as evidence that he lost 3-4% to Clinton in whatever the vote in November is- it could mean he loses in November by 1%, or it could mean he wins by 1%- we simply don't have a way to determine the absolute, only the relative change to whatever that absolute was prior to the tape revelation.

  6. If, in the coming two weeks, he regains a 3-5% lead in the LA Times poll, I will take that to indicate that, in the end, the tape didn't hurt him with the electorate. It still doesn't mean he wins in November, only that the tape didn't significantly alter the outcome.