Why We Choose The Polls We Do


Why we use the IBD/TIPP Poll



Aggregating bad polls with good polls gives you bad results - the RCP average in 2012 missed the outcome by 3.2 percentage points.

In summary, IBD/TIPP and The LA Times were each closer to the final result of the 2012 presidential election compared to the aggregate RCP average.  The LA Times and IBD/TIPP were the two most accurate tracking polls in 2012.  The PPD poll was the single most accurate poll in the latest election (the 2014 midterms).  Pew has exceptional predictions with their final polls in national elections, but their release is sporadic and pre-election week results appear to diverge greatly from more accurate tracking polls.

In the 2016 election it is more important than ever to aggregate accurate polls since polls with the most successful track records appear to diverge greatly with RCP and 538 averages.  As of October 20th, the RCP average in a 4-way race stood at Clinton +6, while our aggregate of historically accurate polls had Clinton +0.2 - a very statistically significant difference.

Post A Comment
  •  Blogger  Comment using Blogger
  •  Facebook  Comment using Facebook
  •  Disqus  Comment using Disqus

No comments :


Election Coverage

[Election Coverage][grids]