Also correlated risk. Also modeling from tiny n in rapidly changing times. 2016 was a fail in data journalism and risk communication.
-
-
-
A year when attempts to model risk and polling margins backfiredhttps://twitter.com/atoker/status/796176641600974851 …
End of conversation
New conversation -
-
-
A nice, explicit way to make people understand probabilities is to make it life & death: "if you pull the trigger, you are a 34% of dying."
-
People will quit thinking 34% means 0% if their life's on the line.
- Show replies
New conversation -
-
-
This Tweet is unavailable.
-
Yes. They are trained in polls; also the odd don't reflect the volatility that came from correlated errors, for example. Very misleading.
End of conversation
-
-
-
Mentioning it in text is a rounding error to the graphic everyone refreshed. And he was honestly the best in a dismal field.
-
538's graph's volatility was useful for the corner of the Clinton campaign I was in—I don't think we'd have gotten such high turnout w/o it
- Show replies
New conversation -
-
-
34% explained: "Here is a revolver with two loaded chambers and four empty chambers. {Spin} Now pull the trigger."

Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
With politics, people have been trained in polls and at most margin of error. They are not used encountering probability models there.
-
Yes, precisely. They read/see "he has a 30% chance of winning" and think "oh, so only 30% of people are voting for him".
- Show replies
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.