Poll Average Details- LA US Senate - 2008-11-04 20:00:00

The Poll Average is a weighted and adjusted number.
[Mid Date Column] All times are calculated by the middle date of the poll if it spans multiple days.
A time window of 2 months prior to the most recent poll is used.
[Time Weight Column] Polls are weighted in the amount they contribute to the final results in a linear fashion in this time window - the newest poll has a weight of 1, a poll 30 days prior will have a weight of 0.5 and a poll 60 days out will have a weight of 0. If the same pollster has more than one poll in this time period, only the newest will be used.
[Internal Column] If a poll is marked as being an internal poll for one of the candidates, it's weight will be halved.
[Candidate Columns, Bias] OurCampaigns has calculated partisan bias figures for all Polling Firms based on data from final vote totals in races they have polled. These amounts are subtracted from the poll result - in the case the bias figure is negative it will effectively be added to the poll result and appear as two dashes (minus a negative = plus). [TOTAL Row] The adjusted weights are summed for each poll result and the adjusted poll results for each candidate are multiplied by the poll weight and then summed. Each candidate's poll sum is divided by their weight sum to give the total.

PollMid DateTime WeightInternal Mary Landrieu John N. Kennedy Richard M. Fontanesi Robert Stewart Jay Patel
YouGov11/1X 1.00000   52.0000 - 1.4134 bias 39.0000 - 0.0288 bias      
OnMessage Inc. (R)10/29X 0.95000   45.0000 - -1.9372 bias 44.0000 - 3.3406 bias      
Ed Renwick/Loyola Institute of Politics10/26X 0.90000   49.0000 - 7.5932 bias 34.0000 - -5.7575 bias      
Southeastern Louisiana University10/23X 0.85000   53.0000 - 0.1162 bias 34.0000 - 1.0603 bias      
Mellman Group (D)10/21X 0.81667   54.0000 - 3.6270 bias 34.0000 - -2.5182 bias 3.0000 - 0.6331 bias    
Rasmussen Reports10/21X 0.81667   53.0000 - 0.4180 bias 43.0000 - 1.5977 bias      
TOTAL 49.026414 38.439975 2.366900 0.000000 0.000000