I would have loved to have predicted the Brexit vote with confidence. But I didn’t. Instead I predicted a 52-48 win for Remain.
However, I did repeatedly warn (to no avail) to Matt Singh and others counting on “history”, that counting on history was a bad idea.
Comments Ahead of the Vote
In my final post before the vote I noted Last Four Brexit Polls Split: Turnout of Age 18-35 Likely the Key.
Turnout was indeed the key, but I also questioned the methodology of the latest polls.
Likelihood to Vote
In ORB’s final poll, not only did the likelihood of the two most likely age groups to vote Remain rise dramatically, the likelihood of the most likely age group to vote Leave fell.
On top of that, ORB imputed a 3-1 advantage for Remain on most undecided voters, leaving a mere 2% truly “undecided”.
What happened is all the polls that got it wrong last time, changed their methodology in belief that Matt Singh, was the latest all-knowing God, just as Nate Silver was in the US.
Neither Nate Silver, nor Matt Singh responded to numerous emails or Tweets. My Tweets to Singh were not combative (until after vote), and in many cases I simply asked polite questions because I did not understand his methods or terminology.
That ORB poll was ridiculous. I called it at the time. Undecideds suddenly became a 3-1 favorite for Remain, based on “history”.
Both Nate Silver and Matt Singh had more faith in history than changes in attitudes.
My mistake, and I made one, was believing pollsters at the end, when it was clear they changed their methodology to accommodate history, just as Matt Singh, like Nate Silver before him did.
At any rate, I take pride in pointing out the odds were much closer than the pollsters and the likes of those counting on history had you believe.
Anyone confident of victory and was correct, went out on a limb. No one knew this outcome. Anyone whom tells you they did know the outcome is a fool or a charlatan.
That said, I do get to gloat on one thing: my repeated calls that the odds were way wrong and those who had faith in the betting odds bookies offered and Matt Singh verified were way wrong.
Mike “Mish” Shedlock