Diceheist posted...
According to the article landlines made up less than 6% of the poll. Cell phone is the most accessible means of polling that can't easily be cheesed (anyone can bot an internet poll). Just talking to people in person would suffer an even bigger sampling bias.
NYT's methodology adjusts for the fact that certain demographics are significantly underrepresented by cellular polling by weighting their responses more greatly. This has produced accurate results so far.
Accessibility doesn't mean reliability.
Modern polling has a lot of issues.
Instead of using opinion polls, why not use actual poll data? Ya'know all the elections that have happened from 2020 til now, where the Democrats have been obliterating in almost every single race, but yeah keep clinging to polls.