Timeline for US Election results 2016: What went wrong with prediction models?
Current License: CC BY-SA 3.0
8 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Nov 10, 2016 at 21:23 | comment | added | Has QUIT--Anony-Mousse | Why does it say as of August 30? Did you look at other Aug 30 results, too? | |
| Nov 10, 2016 at 21:07 | comment | added | Mark | The LA Times poll was correct by accident: the over-weighted 19-year-old counterbalanced the under-weighted white rural vote. | |
| Nov 10, 2016 at 20:15 | comment | added | T.E.D. | ...actually, it looks like Clinton may finish about a full point ahead of Trump in the popular vote, which means this poll was off by 4, not 3. So in theory a similar poll that had her winning by 3 points would have been twice as accurate as this one (off by only 2 points rather than 4). | |
| Nov 10, 2016 at 14:56 | comment | added | T.E.D. | Problems with this poll as an example are 1) Its polling the wrong thing. Popular vote is correlated with winning the Presidency, but that's not how its decided. 2) It got the topline wrong. Clinton won what it is measuring, not Trump. 3) It was off by the same 3ish points most of the other polls were, just in a different direction. | |
| Nov 10, 2016 at 13:49 | comment | added | AnoE | Just a slight note - would you really expect any statistic to be within less than 3.2% of an error window? | |
| Nov 10, 2016 at 4:11 | comment | added | Jon | It's not certain if the y-axis is percentage of popular vote. The questions given to the sampled population were "likelihood" of voting for a candidate as opposed to a Trump vs Hillary answer. In either case, I felt this poll was relevant to point out as it did not follow the herd and used a unique weighting methodology. You can follow the links, read through and learn more about the methodology though. | |
| Nov 10, 2016 at 1:35 | comment | added | Winston Ewert | This poll had Trump winning the popular vote by 3.2%, but Clinton seems to have won by .1%. So I don't see how you can say they had accurate numbers. | |
| Nov 9, 2016 at 19:30 | history | answered | Jon | CC BY-SA 3.0 |