American Research Group, Inc.

 

2017 Alabama US Senate Special Election Pollster Ratings

 

The following table lists the accuracy of pollsters in the December 12, 2017 Alabama US Senate special election using the measure of polling accuracy proposed by Martin, Traugott, and Kennedy.

An accuracy measure of 0.00 indicates that the odds ratio of the survey results matches the odds ratio of the actual results (the closer to 0.00, the better).

 

2017 Alabama US Senate Special Election
Jones Moore Odds
Ratio
Accuracy
Washington Post
50% 47% 1.03 0.03 
Monmouth 46% 46% 0.97 -0.03 
Gravis 45% 49% 0.89 -0.12 
Trafalgar 46% 51% 0.87 -0.13*
JMC Analytics 44% 49% 0.87 -0.14 
CBS/YouGov 43% 49% 0.85 -0.16*
Strategy Research 43% 50% 0.83 -0.18*
Fox News 50% 40% 1.21 0.19*
Emerson 44% 53% 0.80 -0.22*
Actual 49.9% 48.4%

 

*The accuracy measures for five of the nine polls fall outside their respective 95% confidence intervals.

The average absolute accuracy for all polls is 0.134.

The average absolute accuracy for the four polls within their respective 95% confidence intervals is 0.079.

The average absolute accuracy for all polls in the 2017 Virginia gubernatorial election was 0.133.

The average absolute accuracy for all polls in the 2016 US presidential election was 0.070.

Special note: SurveyMonkey provided results from 10 alternative likely voter methods, with no single prediction for the race. The average absolute accuracy for all ten alternatives is 0.101. Four of the alternatives produced accuracy measures that were outside of their respective 95% confidence intervals. The most accurate alternative was standard and 2016 vote weights for all registered voters.


ARG Home