Home | Entries [Previous |Next]
3 Comments
Hrrm...I'm not sure if that works or not. "Voter share" is only tenuously related to turnout. In other words, I'm not sure what I called "voter share" (county vote total / state vote total) is the same thing as anticipated turnout - which is what you have it represented as in your second equation.
You'll notice how most of the share numbers listed are >1%. If we're trying to project likely turnout, I think it's safe to assume turnout in most of these counties will be far greater than 1% of registered voters.
So that number (something closer to 30%, I would think) times your number of registered voters would be LV. I typically get projected turnout number by taking an average of similar elections and doing some sort of growth exercise as you note above.
Posted by: Adam | August 1, 2006 10:20 AM
What is a GOP primary election I live in Great Britain and enjoy whating the WEST WING I am not sure on certain terms
Posted by: geoffrey anderson | August 23, 2006 05:21 AM
Analysis for July 31, 2006
Pick Your Own GOP Primary Winner
Here's something for everyone interested in the GOP Senate primary to play with.
We all know that a higher margin of victory in Shelby County is more powerful than an equal margin in say- Polk County, so now you can put your own margins in and see who comes out a winner, with properly weighted data. I have some default numbers in here to get you started, with documentation in the extended entry. All you have to do is enter in how you think the vote will split along the three grand divisions and see how your candidate fares state-wide. Post different scenerios you come up with in the comments!
Update -- Now allows totals under 100%.
Documentation -- A simple "voter share" calculation is used to determine the relative percentage of that votes cast statewide that each Tennessee county contributes (county vote total / statewide vote total). These counties are then categorized either East, Middle, West. The total voter share percentages are aggregated based on the three grand divisions.
| VS | EMW | |
| Anderson | 1.65% | E |
| Bedford | 0.53% | M |
| Benton | 0.43% | W |
| Bledsoe | 0.35% | E |
| Blount | 1.29% | E |
| Bradley | 1.30% | E |
| Campbell | 0.88% | E |
| Cannon | 0.41% | M |
| Carroll | 0.65% | W |
| Carter | 1.26% | E |
| Cheatham | 0.78% | M |
| Chester | 0.34% | W |
| Claiborne | 0.88% | E |
| Clay | 0.30% | M |
| Cocke | 0.56% | E |
| Coffee | 0.90% | M |
| Crockett | 0.24% | W |
| Cumberlan | 1.22% | E |
| Davidson | 6.21% | M |
| Decatur | 0.36% | W |
| DeKalb | 0.40% | M |
| Dickson | 0.92% | M |
| Dyer | 0.80% | W |
| Fayette | 0.72% | W |
| Fentress | 0.60% | E |
| Franklin | 0.78% | M |
| Gibson | 0.76% | W |
| Giles | 0.78% | M |
| Grainger | 0.28% | E |
| Greene | 1.32% | E |
| Grundy | 0.24% | M |
| Hamblen | 1.09% | E |
| Hamilton | 4.01% | E |
| Hancock | 0.27% | E |
| Hardeman | 0.59% | W |
| Hardin | 0.67% | W |
| Hawkins | 0.79% | E |
| Haywood | 0.51% | W |
| Henderson | 0.29% | W |
| Henry | 0.48% | W |
| Hickman | 0.57% | M |
| Houston | 0.26% | M |
| Humphrey | 0.43% | M |
| Jackson | 0.40% | M |
| Jefferson | 0.68% | E |
| Johnson | 0.49% | E |
| Knox | 5.64% | E |
| Lake | 0.23% | W |
| Lauderdale | 0.53% | W |
| Lawrence | 1.01% | M |
| Lewis | 0.39% | M |
| Lincoln | 0.73% | M |
| Loudon | 0.73% | E |
| Macon | 0.36% | M |
| Madison | 1.81% | W |
| Marion | 0.22% | M |
| Marshall | 0.61% | M |
| Maury | 1.12% | M |
| McMinn | 0.93% | E |
| McNairy | 0.63% | W |
| Meigs | 0.31% | E |
| Monroe | 1.07% | E |
| Montgome | 1.64% | M |
| Moore | 0.19% | M |
| Morgan | 0.54% | E |
| Obion | 0.52% | W |
| Overton | 0.50% | M |
| Perry | 0.13% | M |
| Pickett | 0.25% | E |
| Polk | 0.38% | E |
| Putnam | 1.34% | M |
| Rhea | 0.75% | E |
| Roane | 1.35% | E |
| Robertson | 1.21% | M |
| Rutherford | 2.43% | M |
| Scott | 0.66% | E |
| Sequatchi | 0.37% | E |
| Sevier | 0.79% | E |
| Shelby | 16.16% | W |
| Smith | 0.52% | M |
| Stewart | 0.30% | M |
| Sullivan | 1.98% | E |
| Sumner | 1.98% | M |
| Tipton | 0.90% | W |
| Trousdale | 0.26% | M |
| Unicoi | 0.25% | E |
| Union | 0.54% | E |
| Van Buren | 0.22% | M |
| Warren | 0.92% | M |
| Washingto | 1.47% | E |
| Wayne | 0.45% | M |
| Weakley | 0.41% | W |
| White | 0.60% | M |
| Williamson | 2.22% | M |
| Wilson | 1.76% | M |
The following are the sums of the grand divisions, which are used as unrounded numbers for the purpose of this calculator: East = .369233392; Middle = .350179311; West = .280587298; All of the preceeding numbers were determined based on the previous "marker election" (2002 Republican Primary). Voter share can flux with turnout, which may be an issue with the 2006 GOP primary, but this margin typically is not significant.
"Voter share" should be termed "POE" for percentage of effort.
POE = GeneralElection02 County total / GE02 State Total
POE x Registered Voters as of now = LV (Likely Vote)*
*Let me note here that the LV is without "growth". We would usually build some type of growth multiplier that would help us estimate the actual expected vote. Often times the growth multiplier would be combined with the POE to create a EVE (expected vote effort). We would use the EVE in place of the POE for the functions below.
We used to build "quota" tables, which would indicate exactly how many votes in each county we needed to win.
((County Registered voters x POE)/2)) + 1 = Votes Needed to Win (VNTW)
We would then build a "forcast" which was how many votes, based on our latest polling, we thought we were going to get in the county.
VNTW x (Current polling percentage in county / 100) = Likely Earned Vote (LEV)
On election day we would match the returns to our LEV and VNTW. This is how we tracked statewide returns.
In statewide races we would break the vote down by DMA (media markets) and not major geographic divisions (ie: Middle, East, West), because all of our targeted media was executed by DMA and not major geo divisions. So, if we were polling soft in an area, we could adjust our buy/placement/mail, etc.
Anyway, I like your calculator. It's a great way to play with the numbers. But, I think we all know who is going to win.
Posted by: Rex | August 1, 2006 07:43 AM