Pre-election polls
8TM
North America
People are correctly worried about privacy. Some stranger is reaching out to you and wants to know all your political opinions, how do you know they’re legit? With modern phishing techniques it’s easy for a stranger to impersonate a legit company. If they are legit, how do you know your responses are really anonymous and won’t get hacked or leaked? Oh yeah, and the call might be recorded.
I think the whole industry is obsolete, it’s simply not possible to just invent a better model. The problem is there’s so much $$$ in campaigns and those campaigns will keep spending on polls because that’s what they’ve always done.
If I make one prediction, I think by 2024 that mainstream news outlets will start reporting on betting market odds as the main predictor for the big races.
Got something to say?
Start your own discussion
5 comments
Latest
Both of the simulators had Biden winning easily with 538 giving Trump a 10% chance of winning and The Economist only 3%. Most of the polling averages of battleground states on 538 were off quite a bit. Still, both predicted the qualitative features of the election correctly with Biden taking Arizona and the battleground states in the Midwest. All in all, I think both sites are still worth following.
Predictit varied all over the place and the polling/simulations were more reliable than the betting market.
Don't understand the privacy issue? If someone wants to know my voting preference, I don't see the big deal.
Overall the polls were ok, a bit too down on Trump and Republicans in general. But 538 gave Trump a much higher chance than most other sites.
Good polling is going to get harder and harder especially at the state level. Nationally the numbers were decent overall, biden won the popular vote by a lot.
https://news.gallup.com/poll/15370/party…