# Tag Archives: Trump win probability

## 2016 ELECTION MODEL (Nov.3): Trump 98% Win Probability

2016 ELECTION MODEL (Nov.3): Trump 98% Win Probability

Richard Charnin
Nov.3, 2016

The purpose of the Election Model  is to show the effects of changes in voter party affiliation (Dem, Rep, Ind). There are currently nine polls in the model. Each poll is shown using a) the actual poll shares and Party-ID weights and b) the actual poll shares using the Gallup party-affiliation survey. Gallup is the only poll dedicated to national voter party affiliation.

Undecided voters are allocated to derive the final adjusted TRUE poll share. Typically the challenger (in this case Trump) gets approximately 75% of the undecided vote.

Clinton leads Trump 44.9-43.3%  in the actual 9-poll average.

After adjusting the polls for the Gallup voter affiliation split (40I-32D-28R):
Trump leads Clinton 44.7-41.7% and by 336-202 EV before undecided voter allocation.
Trump leads Clinton 49.0-43.2% after undecided voter allocation.
There is a 98% probability that Trump will win the popular vote.

THE MODEL SHOWS THAT THE PRE-ELECTION POLLS ARE OVERSTATING HILLARY CLINTON’S VOTE BY INFLATING THE NUMBER OF DEMOCRATS COMPARED TO INDEPENDENTS AND REPUBLICANS.

As I have stated many times, each poll has a different party-ID.Theoretically, they should all have the SAME Party-ID since these are NATIONAL polls – and there is only ONE theoretical NATIONAL Party-ID split at any given point in time.

The popular Vote Win Probability and estimated Electoral Vote are calculated for each poll. The 2016 party-ID for each state is calculated by applying the  proportional  change  from the 2012 party-ID  to  the current Gallup 2016 survey Party-ID. The state votes  are calculated by applying the published national poll shares to the 2016 state party-ID. The electoral vote is then calculated.

The built-in SENSITIVITY ANALYSIS shows the effect of incremental vote shares on the total vote.

Those who have written models can appreciate the methodology. So can individuals who can apply basic logic.The model uses actual published data. If there is another quantitative modeler out there who has written a similar model to approximate the True poll shares, I would like to see it.

 9-POLL AVERAGE Gallup Pct Stein Clinton Trump Johnson Ind 40.0% 4% 28% 44% 6% Dem 32.0% 1% 91% 6% 2% Rep 28.0% 1% 5% 90% 3% Total 92.5% 2.2% 41.7% 44.7% 3.9% Votes 119,448 2,840 53,863 57,736 5,009 EVote 538 0 202 336 0
 Nov. 3 Party ID ACTUAL Ind Dem Rep HRC Trump Ipsos 11.9% 43.5% 36.6% 42% 38% IBD 27.4% 39.9% 32.7% 44% 44% Rasmussen 32% 40% 28% 42% 45% Quinnipiac 26% 40% 34% 47% 40% Fox News 19% 43% 38% 44% 41% CNN 43% 31% 26% 49% 44% ABC 29% 37% 29% 47% 45% Gravis 27% 40% 33% 46% 45% LA Times 30% 38% 32% 43% 48% Average 27.3% 39.2% 32.1% 44.9% 43.3% GALLUP ADJUSTED Elect Vote Popular Vote Undec.Alloc. 40I-32D-28R HRC Trump HRC Trump Win Prob Win Prob Ipsos 37.9% 39.4% 232 306 73.4% 99.4% IBD 40.9% 45.8% 180 358 96.8% 99.8% Rasmussen 37.2% 47.4% 46 492 100.0% 100.0% Quinnipiac 44.7% 40.8% 335 203 6.5% 35.8% Fox News 39.6% 41.6% 218 320 79.9% 97.3% CNN 48.6% 44.4% 335 203 7.0% 13.7% ABC 46.4% 49.7% 202 336 86.5% 87.4% Gravis 42.6% 45.6% 216 322 86.7% 99.1% LA Times 40.7% 49.4% 54 484 99.9% 100.0% Average 41.7% 44.7% 202 336 87.2% 98.1%
 Sensitivity Analysis 9-Poll Average Gallup 40I-32D-28R Trump % Rep Trump 86.0% 88.0% 90.0% 92.0% 94.0% % Ind Trump 48% 45.2% 45.8% 46.3% 46.9% 47.4% 44% 43.6% 44.2% 44.7% 45.3% 45.8% 40% 42.0% 42.6% 43.1% 43.7% 44.2% Clinton 48% 41.2% 40.7% 40.1% 39.6% 39.0% 44% 42.8% 42.3% 41.7% 41.2% 40.6% 40% 44.4% 43.9% 43.3% 42.8% 42.2% Margin 48% 4.0% 5.1% 6.2% 7.3% 8.4% 44% 0.8% 1.9% 3.0% 4.1% 5.2% 40% -2.4% -1.3% -0.2% 0.9% 2.0% Vote Margin (000) 48% 4,730 6,068 7,406 8,744 10,081 44% 908 2,246 3,583 4,921 6,259 40% -2,915 -1,577 -239 1,099 2,437
 9-poll average Vote Share Electoral Vote Clinton Trump Clinton Trump Total 41.7% 44.7% 202 336 AK 29.6% 49.9% 0 3 AL 36.7% 51.4% 0 9 AR 38.6% 49.0% 0 6 AZ 36.3% 47.9% 0 11 CA 44.7% 41.3% 55 0 CO 37.6% 46.8% 0 9 CT 42.6% 40.7% 7 0 DC 66.6% 23.7% 3 0 DE 46.9% 40.0% 3 0 FL 41.2% 45.2% 0 29 GA 39.8% 48.0% 0 16 HI 46.4% 42.1% 4 0 IA 37.9% 46.4% 0 6 ID 32.1% 54.9% 0 4 IL 45.3% 42.7% 20 0 IN 38.6% 49.0% 0 11 KS 32.4% 52.7% 0 6 KY 47.9% 42.2% 8 0 LA 36.6% 46.0% 0 8 MA 43.8% 37.4% 11 0 MD 51.0% 36.9% 10 0 ME 39.2% 44.3% 0 4 MI 43.5% 44.3% 0 16 MN 43.1% 45.1% 0 10 MO 39.7% 48.4% 0 10 MS 38.8% 49.4% 0 6 MT 35.3% 52.8% 0 3 NC 43.5% 42.6% 15 0 ND 37.6% 50.4% 0 3 NE 34.8% 52.4% 0 5 NH 36.2% 46.9% 0 4 NJ 40.9% 41.4% 0 14 NM 45.8% 41.4% 5 0 NV 41.7% 44.7% 0 6 NY 48.6% 37.9% 29 0 OH 41.0% 47.1% 0 18 OK 42.1% 46.8% 0 7 OR 41.6% 43.6% 0 7 PA 46.3% 42.6% 20 0 RI 47.0% 35.5% 4 0 SC 39.7% 48.4% 0 9 SD 36.6% 50.8% 0 3 TN 37.1% 50.7% 0 11 TX 39.2% 47.9% 0 38 UT 30.3% 57.8% 0 6 VA 40.5% 47.4% 0 13 VT 46.1% 41.2% 3 0 WA 42.5% 47.0% 0 12 WI 42.2% 46.1% 0 10 WV 47.7% 39.8% 5 0 WY 25.8% 62.5% 0 3