Jim Savage points us to this explainer, How do pollsters predict UK general election results?, by John Burn-Murdoch of the Financial Times.
It’s bittersweet seeing my method described by some person I’ve never met. Little baby MRP is all grown up!
Being explained by the Financial Times—that’s about as good as being in the Guardian and the Times. Not quite as good as being mentioned in Private Eye—that’s still the peak of the news media, as far as I’m concerned. But still pretty good.
It’s kind of amazing seeing the phrase “MRP election poll” in a headline, with the implication that’s just a standard phrase now, not even needing a reference. A lot can happen in 22 years.
Maybe I’m totally off but I feel like it misses a key point? Pooling. It really didn’t look at the way one area might be estimated based on another’s samples.
Anon:
Yes, I agree, the video would’ve been improved by an explanation of partial pooling.
What prominent US polls or election forecasters make use of MRP?