Tarnished silver
Nov 15, 2018
If we'd given Trump a 30% chance and everyone else had given him a 45% chance, we wouldn't consider our 2016 forecast to have been too great. But the consensus was about a 15% chance. Some people gave him a 1% chance. So, yeah, we think 2016 counts as a good forecast.
— Nate Silver (@NateSilver538) 14 November 2018
Nate Silver, you funny guy! N. N. Taleb knows how to handle this idiot:
In a nutshell, what Klueless Nate @NateSilver538 doesn't get is the PRINCIPLE that it is not ONE forecast that matters for track record, but ALL the forecasts made per election.
— Nassim Nicholas Taleb (@nntaleb) 15 November 2018
That principle we've known since De Finetti; I adapted it to continuous time & derived properties.
Here's a small sample of all those "good" forecasts leading up to the 2016 election:
And his headlines .... pic.twitter.com/gMqonUdqYu
— Jᴀqᴇn H'ghᴀr (@RealFacelessMan) 15 November 2018
Here's another beautiful collection of those lovely forecasts. The 2016 presidential election was a colossal failure for Nate Silver. He got basically everything wrong. Even during the election night, his ridiculous model increased Hillary's chances of winning as it became clear that the Witch was going to crash and burn. Instead of owning up to his bias and incompetence, he lies and sugarcoats his shitty forecasts.
Comments