Political observers were shocked over the weekend after a new poll revealed what one reporter called "the largest lead for either party on the congressional ballot in the NBC News poll since the 2018 midterms."Jake Sherman, the founder of Punchbowl News, flagged the NBC poll on Sunday morning."NBC P...
Is there any reason to trust these polls? I feel like they’ve been wrong since 2016
Maybe?
I don’t have the time or energy to dive into the methodology to figure out if any poll is conducted scientifically. What I do know is that it’s way easier to build a bad poll, and easier still to hit the bricks and generate heaps of bad data.
What I really wish these would ask are things like: “if you had the only vote and could write-in anyone that is eligable, even if they’re not on the ballot, who would you choose?”… or: “Who is your favorite person in politics right now?” Disregarding of all the nonsense answers, you’d probably find out a lot more about hearts and minds versus whatever controlled choices major parties are propping up.
The problem with 2016 was that the polls accurately predicted who would win the popular vote, because that’s what they measure. The 270toWin poll based outcome showed Trump’s chance of victory LINK HERE, and so did 538 which is no longer available as it was purchased by ABC News and taken offline.
Trump’s loss in 2020 and victory again in 2024 were also predicted by both aggregate polls and online betting platforms.
The problem with poll inaccuracy is solved by two methods: increased sample size and giving priority based on perceived bias (gaining back points for having clear and accurate data available to the public).
I think 2016 proved that we should actively ignore them and not get complacent based on poll numbers.
Since only 2016?