Thread by @AbdulElSayed

Awful take, habibi.

Tell me @NateSilver538, what’s “actual science”?

Political polling is in shambles, why don’t you stick to that and let “actual” scientists handle this, mmkay?

https://twitter.com/NateSilver538/status/1340473435156307968

habibi, what’s actually happening is that you’re just arrogant enough to misinterpret models you don’t understand & write whole damn tweet thread about it. And you forget this isn’t sports or politics. Trust matters here, & you’re undermining it out of ignorance/arrogance.

https://twitter.com/NateSilver538/status/1340363906590650370

So while we’re at it, take some notes: predictive modeling, which you do, is not the same as causal modeling, which we do.

We don’t just throw a bunch of variables into a model to maximize fit. We ask how variables may explain impacts of other variables.

https://twitter.com/natesilver538/status/1340361755403366400

This model was adjusted for a number of key chronic disease risk factors, which ultimately blunt the impact of the chronic diseases they cause in the model.

How many people have you met with renal disease who don’t also have hypertension?

Age clearly matters, but much of the impact of “age” in these models isn’t simply age in years, it’s also residency in long term care facilities (40% of deaths), which obviously tracks with age—and those folks were among the first to get vaccinated.

So, you’re interpreting models you don’t actually understand and haven’t thought through to make policy claims that are already being addressed.

Now that I come to think of it, it’s not surprising that political pollsters get things so wrong if this is the slapdash work you do.