Nate Silver: “So if we had some sort of probabilistic forecast of the race, it would undoubtedly think that Zohran’s chances are high. The de facto margin of error is almost certainly large here: the further you go downballot, the higher polling error tends to be (races to the U.S. House are harder to nail than presidential elections, for instance). Furthermore, New York has a lot of groups that are hard to poll (including plenty of voters who are not native English speakers). It’s a multi-candidate race, which considerably increases the variance. And the polls were way off in the primary — although they lowballed Mamdani rather than his opponents.”
“Still, let’s say that Mamdani’s lead is really ‘just’ 19 points, closer to the average of recent polls than his 22-point lead in the Times/Siena poll. And let’s say the practical margin of error in the race is 20 points — in other words, very large — partly because there are still almost two months of campaigning left. A formula based on the normal distribution would nevertheless imply that Zohran has around a 95 percent chance of winning.”
