London Mayoral Forecast
As a tradition, pollsters lie about their data in the UK. If not for misinforming the public and disregarding the rules of basic science, it would be funny.
Pollsters in the UK report all of their numbers as if there were no undecideds. When there are only like, 2% undecided, it’s still terrible practice.
One problem.
Uh oh.
Now, three polls is far from an authoritative poll average (sorry, RealClearPolitics) but in this case, there are only three who, despite lying about their data by reporting “zero undecided,” are kind enough to let you download reports you can dig through to find how many undecideds there are.
I’m sure the average poll-follower does this, and doesn’t simply take pollsters at their word.
I'm joking of course, even alleged experts take the pollcast reported by pollsters as if it were the poll.
There are a lot of “Undecideds” and they ignore them. It's a disgrace.
Here's what YouGov reported
Khan by 22. Good luck.
Undecideds have a known and measurable impact on the range of possible outcomes: it makes it much wider. You can assume you know all you want, you don’t.
And if you can’t separate your assumptions from the data given by the poll, you’re terrible at your job and should quit forever.
That is to say, all of these “pollsters” have reported numbers for Khan in the mid-high 40s, some over 50, and I’d take lots of money on that “under.”
My median forecast has them in the LOW 40s.
There’s good reason to believe undecideds will lean Conservative, that Reform will underperform and some will break Conservative, and while some Lib Dems will do the same for Labour, it’s likely to be lower in number.
Oh, and thanks to the misinformation of the “Labour got this in the bag up 24 points” reporting, a lot of would-be Labour voters won’t be bothered.
This doesn’t mean Labour won’t win - they’re huge favorites, and rightfully so. But there’s a better chance Labour wins by single-digits than by 20+. This is not the conclusion you would draw from the (lying) poll averages. And in the event Labour loses or the election ends up close - which is not impossible - get ready for more whining about the polls being wrong from people who don’t understand them.
Here’s my forecast for the election
Labour is “red” and Conservative “blue” by convention, so that's what I've done here.
Here's how it breaks down:
That's a median forecast of 12.77. Far from the 20-point margin in many pollcasts.
It's more likely to be a single-digit margin than over 18!
If you believe my forecast, anyways.
Pollcasters disagree.
We'll see tomorrow.
And to be clear, the range of possible outcomes (few polls + high undecided) leads me to believe 1.5% is low for Conservatives.
Given the media and pollsters irresponsibility, reporting Khan “up by 20+” I'd be shocked if many Labour voters didn't stay home.
If I were Labour, I'd vote like the race were within 2.