I’m putting together a much longer, much more comprehensive takedown of the 538 and RCP polling aggregates, and when I post it in a day or two I hope you’ll rec the fuck out of it, because I’m putting a lot of effort into it; but with Markos making a supposedly encouraging but actually depressing post, using 538’s numbers, I’m going to summarize here.
There are dozens of red wave pollsters, whose numbers are always way off, certainly deliberately so, and they’ve been flooding the aggregates, just as they did two years ago. Simon Rosenberg says, as of yesterday, the red wave pollsters have produced 58 polls just this week! The purpose is obvious, as they craft a false narrative that Trump is winning. He’s not.
538 knows better. They “weigh” pollsters based on past record, but no matter how much they weigh them, the sheer quantity of shit polls overwhelms any plausible quality in the weighing. Shit pollsters shouldn’t be weighed, they should be excluded. If one wants to have credibility.
So, here, in brief, are the numbers I will show in a longer post, with screen grabs and quotes, hopefully tomorrow. And, btw, RCP is even worse. And yes, 538 had some misses on the GOP side, too, such as having Desantis up only 12.1, when he won by 19.4, launching him into the stratosphere as the leader of the Republican Party. Until people around the country got a look at him.
But these are the 538 aggregates from two years ago compared to the actual results, in what were considered to be the tightest races. These are a lot of misses, in four cases having the wrong winner, and in the rest significantly underestimating the Democrat’s margin of victory.
Two years ago, 538 had Lake winning AZ-Gov by 2.4. Hobbs won.
They had Kelly up a mere 1.5 for AZ-Sen. He won by 4.9.
In NV, they had Laxalt up 1.4. CCM won, although it was close.
In PA, they had Oz beating Fetterman by 0.5, Fetterman won by 4.9.
In WI, they had Michels up by a slight 0.3, but Evers won by 3.4.
In GA, they had Walker up a point. Warnock won by 0.77, then went on to win the runoff.
In MI, they had Whitmer up only 4.8. She won by 10.5. (The absurd Trafalgar poll actually had her down 1)
In NH, they had Hassan up 2.2. She won by 9.1.
In WA, they had Murray up 4.7. She won by 15.5.
In CO, they had Bennet up a robust 8.4, and he won by 14.6
How 538 retains any credibility baffles me. Don’t read them, don’t give them clicks, and question anyone who ever cites 538.
