News for hackers. Stuff that matters.

A comment on Code for The Economist's election model for 2020. Parent comment

Re:Code for The Economist's election model for 2020 by seanmcdirmid on Saturday August 22, 2020 @11:11AM (#24244638)
> Some of the other models for the last election were really bad. Giving something like a 95-98% chance to Hillary was arguably a fundamental failure.

The page is still up, you don’t have to pull that number from memory. At the end, it was also nowhere near 95-98%. https://projects.fivethirtyeight.com/2016-election-forecast/

[ Parent | on HN ]
Re:Code for The Economist's election model for 2020 by CrazyStat on Saturday August 22, 2020 @11:23AM (#24244715)
By "other models" I think he means models other than 538's, several of which did put the probability of Clinton winning at over 95%.
[ Parent | on HN ]
Re:Code for The Economist's election model for 2020 by xapata on Saturday August 22, 2020 @12:56PM (#24245436)
Why would estimating a 95% chance for Hillary be a fundamental failure? 5% likelihood things happen often. You've never rolled a 20 on a 20-sided die?
[ Parent | on HN ]
Re:Code for The Economist's election model for 2020 by patmorgan23 on Saturday August 22, 2020 @01:55PM (#24245850)
Yep the odds of anyone person getting hit by lightning in a year is 1 in 700,000 yet every year several people are struck.
[ Parent | on HN ]