The widgets and charts are below. First, a brief tangent:
* * * * * * *
Rasmussen, and the watering down of the polling field …
About a hundred state polls have been released since the end of the primary season a couple weeks ago, and we have seen the number of undecided voters from these polls range anywhere between 5% and 18%. On Wednesday alone there were nine polls released, and only two of those polls had greater than 10% undecideds,
both by Rasmussen. Then on Thursday, Rasmussen released three more polls, all of them with undecideds greater than 10%. So I thought I would dig deeper.
Of all the state polls released in the last two weeks, these are the total average undecideds by polling agency (first tier agencies only):
7.5% Average Undecided Voters – Survey USA
8.5% Average Undecided Voters – Research 2000
8.5% Average Undecided Voters – American Research Group
9.5% Average Undecided Voters – Quinnipiac University
9.5% Average Undecided Voters – Public Policy Polling
12.25% Average Undecided Voters – Rasmussen ReportsNow the percentage of the total state polls released in the last two weeks (first tier agencies only):
46% - Rasmussen Reports26% - Survey USA
11% - Quinnipiac University
6% - Public Policy Polling
6% - Research 2000
5% - American Research Group
The average number of undecideds from these polls is 10.14%. Of all the polling agencies, only one agency has an average of undecideds above 10.14% …
Rasmussen. Since Rasmussen is releasing about half of all the state polls, and since Rasmussen’s margin of error is always 4.5% (higher than all the other agencies), it’s easy to see how Rasmussen weakens the field with less reliable polls.
This is why I don’t average the polls together for each state for projections, unless they are taken during the same time period and meet certain criteria. Averaging polls together adds the weakest of polls into the mix and waters down the results.
I use a
strength of projection as a methodology for projecting electoral votes. Under this methodology, all the polls are still used but the weakest of the polls (being older than 35 days or having undecideds greater than 10%) are given less weight. Doing so, along with other methodologies, makes my projections accurate to +/-4%, better than Rasmussen’s +/-4.5%.
* * * * * * *
I return you now to your regularly scheduled widget. :7 (By the way, Obama is within one point in Georgia, and Barr is taking 6% there! Har!)
Yesterday’s WidgetFull Weekly TrackingPast editions of THE MATH* * * * * * *
TRACKING^ Things are leveling off at a nice place to come to a rest! Demeter says there's something wrong with this chart: I need to make room for 400, lol! That's a good problem to have!
^ Speaking of watering down the polling field, RCP's poll average was affected by the national Faux News poll yesterday. The Faux News poll shows
19% Undecided. Talk about a useless result!
^ Another day, another all-time high. Ohio take a big leap north, moving much further away from the zone. Traders are beginning to believe that Ohio won't be as much in play this year as other states for Obama. And Iowa jumps back into the 80's today.
Both Intrade and Rasmussen Markets move down today from their all-time highs yesterday for Obama.
* * * * * * *
^
PLEASE NOTE: This map is not a prediction.
It is a map of the states which are currently
either polling or trading blue. :D
Sources:
Pollster.comFiveThirtyEight.comElectoral-Vote.comIntradeRasmussen MarketsRCP AverageRCP Latest PollsDonate to the Obama Campaign today through DU’s donation link at Barack Obama’s website:
http://my.barackobama.com/page/outreach/view/main/duWhat is this? THE MATH’s Daily Widget was created to track multiple daily sources which measure the potential success of Barack Obama’s presidential campaign. Projected Electoral Votes are on a scale of 0 to 538, with 270 being a majority, or projected win. Trading and Averages are on a scale of 0 to 100, with 50.1 being a majority, or projected win.The Wigand Electoral Average uses public polling results to determine base states for both parties and a short list of swing states, and combines the actual electoral votes from base states with a weighted Intrade calculation for electoral votes from swing states to derive an average on a scale of 0 to 100. On this scale, greater than 50.00 is considered a win for Barack Obama. An average of 50.10 would yield 270 electoral votes, and an average of 100.00 would yield 538 electoral votes. More Information .