Hoosier4Liberty
Libertarian Republican
- Oct 14, 2013
- 465
- 87
- 78
The following was my first analysis of the US Senate.
http://www.usmessageboard.com/politics/337481-gop-has-60-chance-of-taking-back-senate-in-2014-a.html
I've explained how the analysis works in the prior thread.
I have made a few slight updates to my model:
1. The following races have received ratings changes: KY(lowered McConnell's chances of winning due to bad poll numbers), GA(increased GOP chances of winning due to Dem problems in runoff + unlikelihood of Broun clinching the nomination due to poor fundraising), AR(Cotton gets higher chance due to poor Pryor polling that's been consistently against him), CO(Cory Gardner's entrance turned this into a real pickup opportunity for the GOP, giving 50% chance of win there).
A few things of note:
1. I changed from ~1500 simulations to over 16000, minimizing random error.
2. The GOP's chances of gaining more than 21 seats increased significantly, particularly due to Cory Gardner's entrance into the race.
EDIT: I made a mistake on my original Excel document. I changed my probabilities in Column B, but I forgot to change them in Column C(I should've done static references using dollar signs; I will change that).
I've corrected my probabilities and posted new images.
Note: Even though the odds are 10% higher than before, this isn't actually that huge of a difference. 9 times out of 10, the results would still be the same as if it were a 58% chance.
I still consider anything below 75% or above 25% as a "toss up" in my book. Maybe not a perfectly-weighted coin, but still a tossup. Anything could change here, and I do admit that a few of my projections may be generous toward the GOP.
Personally, I'd say the actual GOP chances are around 60-64% , assuming that my estimates are a bit rosy(again, for people who think I should change the probabilities, tell me what you would make them and I'll run a simulation.
http://www.usmessageboard.com/politics/337481-gop-has-60-chance-of-taking-back-senate-in-2014-a.html
I've explained how the analysis works in the prior thread.
I have made a few slight updates to my model:
1. The following races have received ratings changes: KY(lowered McConnell's chances of winning due to bad poll numbers), GA(increased GOP chances of winning due to Dem problems in runoff + unlikelihood of Broun clinching the nomination due to poor fundraising), AR(Cotton gets higher chance due to poor Pryor polling that's been consistently against him), CO(Cory Gardner's entrance turned this into a real pickup opportunity for the GOP, giving 50% chance of win there).
A few things of note:
1. I changed from ~1500 simulations to over 16000, minimizing random error.
2. The GOP's chances of gaining more than 21 seats increased significantly, particularly due to Cory Gardner's entrance into the race.
EDIT: I made a mistake on my original Excel document. I changed my probabilities in Column B, but I forgot to change them in Column C(I should've done static references using dollar signs; I will change that).
I've corrected my probabilities and posted new images.
Note: Even though the odds are 10% higher than before, this isn't actually that huge of a difference. 9 times out of 10, the results would still be the same as if it were a 58% chance.
I still consider anything below 75% or above 25% as a "toss up" in my book. Maybe not a perfectly-weighted coin, but still a tossup. Anything could change here, and I do admit that a few of my projections may be generous toward the GOP.
Personally, I'd say the actual GOP chances are around 60-64% , assuming that my estimates are a bit rosy(again, for people who think I should change the probabilities, tell me what you would make them and I'll run a simulation.
Attachments
Last edited: