I'm sympathetic to the case, made by @zeynep and others, that forecasting has less value when the polls regularly produce big misses.
But some forecasters, like the Nates, have avoided triumphalism and been upfront about the limitations of their models.
Others, not so much... https://twitter.com/mattyglesias/status/1324507561308114947 …
-
-
"what do you think you learn from data modeling?" is actually something that could be polled and it would be good to know the results
-
What do you think people take away from seeing 96% chance of their candidate winning? https://www.dartmouth.edu/~seanjwestwood/papers/aggregator.pdf …
- Show replies
New conversation -
-
-
Also,
@gelliottmorris, you seem to have forgotten that you dunked on me for saying—in a tweet the emphasized how far Biden was ahead!—that Trump still had a very real chance of victory. You were not nearly as careful as other forecasters, and you know it.pic.twitter.com/WlemQjfayI
-
"slim chance" was the right call, Yascha
- Show replies
New conversation -
-
-
This Tweet is unavailable.
-
Yeah, I am not at all a fan of the coverage, which I had hoped would get better after models but it got worse because models felt more "scientific" rather than, wait polls are garbage, the event is rare, what are we doing here?
End of conversation
-
-
-
the predictive models don’t deserve the prominence they enjoy but empowering hyperactive horse race reporters to set the narrative is vastly more destructive
-
They actually got more empowered by the models is my observation over the last decade! At least, before, one could see they were reading tea leaves from lawn signs and treat it more like a horoscope.
- Show replies
New conversation -
-
-
Huh? The whole point of simulating outcomes is to model the uncertainty in the data. We have describing the forecast like this all along. I don't think the "almost nobody reads forecasts like this" criticism is valid. That's why we focus on the histogram of EC outcomes.
-
@zeynep point is that your audience is perceiving something different than your intent. And it’s a problem. One that could perhaps be corrected, but not unlike those of FB and endemic misinformation. The issue isn’t accuracy, but the way non-data junkies act based on the %’s
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.