13
Comments (8)
sorted by:
-7
tuchodi -7 points ago +2 / -9

So covid19 models aren't supposed to accurately predict anything?

Read this part again:

"They do not predict the future because they inspire changes to behaviour or policy that prevents that outcome."

4
Oldbutnew 4 points ago +4 / -0

Keep in mind, they modelled full lockdowns in the past, and those numbers were bullshit.

This fascist is just gaslighting you and you’re drinking it up.

3
Scrodimite 3 points ago +3 / -0

Wasn't there a model for Saskatchewan that had something like 35,000 Covid deaths on it and in reality it was well under 100?

3
Oldbutnew 3 points ago +3 / -0

Quite possible. Anyone who knows anything about modelling can tell you that in most situations, the preparers pick the answer they want, and then choose the inputs/assumptions they need to get to their answer.

That’s why you need to back-test models. Back-testing is checking to see, using foresight, whether the models actually predicted anything.

Modelling is the new Tarot card reading or astrology being purported by snake-oil salesmen wearing lab coats.

-4
tuchodi -4 points ago +2 / -6

Keep in mind, they modelled full lockdowns in the past, and those numbers were bullshit.

What? Modeling didn't work because we till have covid?

Don't forget OP's quote: "People mistake models (for) crystal balls."

3
Ontarian [S] 3 points ago +3 / -0

Are you really this stupid?

-5
tuchodi -5 points ago +2 / -7

Are you really this stupid?

Hard to say, but it looks like maybe I can read and understand better than you.

3
Ontarian [S] 3 points ago +3 / -0

Doesn't seem like it.