If you give a forecaster full freedom to tweak a model, then no matter how sophisticated the model, its output will simply reflect the incentives felt by the forecaster...
-
-
This Tweet is unavailable.
-
converting the world into nothing but impressions
End of conversation
-
-
-
I think that’s good in this case. Their incentive is to accurately convey margin of error and how this is related to the time until the election Nate has already mentioned that the model would indicate a 90-10 chance if the current polls hold until Election Day
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
usually they do one final move of it to get more attention if the big event of their year
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Your thinking is so appreciated. For real.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
The race has tightened. It will be a close election.
-
Can you back up this claim?
- Show replies
New conversation -
-
-
Polls are a point in time prediction, 90-10 may be based on polls but polls don't capture the potential changes or events. Given the very closely divided nature of the US population and 80K out of 120MM swinging an election; I wouldnt go outside 30-70.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
I think this is a bit unfair to 538. They’ve said that if polls stay around where they are now, then their model will show Biden at ~90% on election day. This is roughly what happened in 2012, though with a dip toward Romney in between. But they had Obama at ~90% at the end.
-
And they already said that if elections happened ~today, Biden would appear with a ~90% chance.
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.