PDA

View Full Version : Nate Silver: polls not accurate for primaries and caucuses




nikcers
02-01-2016, 06:37 PM
Here it is, the elephant in the room that has been ignored up until the votes are counted. Polling wizard says that the polls aren't accurate. In other words water is wet. i trust someone who understands polling science over any pundit.


Ben Carson in second place? Rand Paul in third? The odds are against it — but equally strange things (http://fivethirtyeight.com/features/how-much-can-really-change-before-iowa-votes/) have happened in Iowa before.

We say this for the same reason we can sometimes issue highly confident forecasts just before a general election: It’s what the data tells us. That data tells us that polling in general elections is pretty accurate, at least in the final few weeks before the election. The data also tells us that polling in primaries and caucuses is not very accurate. (http://fivethirtyeight.com/features/iowa-is-the-hardest-state-to-poll/) Historically, the average error of late polls in presidential general elections is about 3.5 percentage points (http://fivethirtyeight.com/features/is-the-polling-industry-in-stasis-or-in-crisis/).1 (http://fivethirtyeight.com/features/iowa-is-the-hardest-state-to-poll/#fn-1) By contrast, the average polling error associated with presidential primaries is more like 8 percentage points (http://fivethirtyeight.com/features/is-the-polling-industry-in-stasis-or-in-crisis/), more than twice as high.

spudea
02-01-2016, 07:03 PM
we really need that 8-point error swing in Rand's favor. That really is the best case scenario and probably lands him tied 3rd with Rubio.

dannno
02-01-2016, 07:14 PM
we really need that 8-point error swing in Rand's favor. That really is the best case scenario and probably lands him tied 3rd with Rubio.

That's the average polling error, I have a feeling Rand's polling error is a bit larger..

craezie
02-01-2016, 07:22 PM
don't we hope.