Thread by @studentactivism

I'd say tonight's results indicate pretty powerfully that "No On One, Leave Two Blank" was good, simple, effective messaging strategy.

https://twitter.com/NateSilver538/status/1429865259804790784

Like Silver, I was made nervous by "No One One, Leave Two Blank." Unlike Silver, (1) I understood the argument in its favor, and (2) nobody's paying me gazillions of dollars a year to be smart about politics and elections.

https://twitter.com/studentactivism/status/1429862711731662854

This is the crux of where Silver (I'd argue) whiffed it on the "Leave 2 Blank" issue—an overconfidence in his own gut instincts, dressed up as spurious statistical precision. (I know, you're shocked.)

https://twitter.com/NateSilver538/status/1429947869474312194

In Silver's calculations above, "Leave 2 Blank" increased the chances of the recall resulting in a GOP governor by 50%. But in order to reach that result, he had to claim the strategy hiked the odds of a bad result by 1000%, and the odds of a good one by only 7%.

And he also has to ignore the fact that losing the recall would have ITSELF been bad news for California and national Dems, even if Newsom's replacement had also been a Democrat.

This isn't even math. This is pure politics. In Silver's "calculations" there were only two possible outcomes:

1. Dem Gov (good)
2. GOP Gov (bad)

But in reality there were three:

1. Newsom stays (very good)
2. Newsom replaced by Dem (bad)
3. Newsom replaced by GOP (very bad)

He makes it explicit here: Newsom staying is good for Newsom, but Newsom being recalled and replaced by another Dem is good for the Dems.

https://twitter.com/NateSilver538/status/1429946341992566793

But of course a successful recall would have been lousy for the Dems, both in California and nationally. We'd all be waking up to a needless, chaotic gubernatorial transition and "Dems In Disarray" headlines that would persist for months.

And of course the claim Silver calls merely "vaguely plausible"—that refusing to put a Dem face on a GOP recall effort, demanding instead that voters choose between the Dems and the GOP as parties, would boost Dem outcomes in a Dem state—turns out to have been exactly right.

And right not just in the sense of "allowing the Newsom to eke out a victory" but also in the sense of "turning a close race into a slam-dunk blowout and shining a sustained national spotlight on one of the Republican Party's most odious clowns."

If you tried to put all that into an equation, though, you'd have four or five different outcomes to account for, each of ambiguous likelihood, contentious value, and ambiguous relationship to the "Leave 2 Blank" strategy. Such an equation would, in short, be obvious bullshit.

So Silver tossed a giant pile of relevant political considerations overboard, pulled some percentages out of his butt, and voila: A calculus. Numbers! Yay!

And I know dunking on Nate Silver is easy and petty and unedifying. If this was just "look at the wrong guy being wrong" I like to think I'd restrain myself.

What's interesting is when he's wrong in a way that has an obvious allure—a way that offers an opportunity to tell a story.

Because I want to be less stupid about politics than I am, and working through how someone whose brain works kind of like mine whiffs something is an opportunity to achieve that, at least a little.

(Also my eldest is in training to be a numbers-heavy poli sci person, and I have to take my opportunities to influence her path where I can find them. Hi, Case!)