Welcome to DU!
The truly grassroots left-of-center political community where regular people, not algorithms, drive the discussions and set the standards.
Join the community:
Create a free account
Support DU (and get rid of ads!):
Become a Star Member
Latest Breaking News
General Discussion
The DU Lounge
All Forums
Issue Forums
Culture Forums
Alliance Forums
Region Forums
Support Forums
Help & Search
Random Walk-President Obama Has A 95.5% Chance Of Winning Ohio
infernoman wrote on Oct 30, 2012 23:37:50 #
Updating my own trimmed-mean random walk model of Ohio. Just to recap:
---
I include all live polls (python+Mathematica), trim the top and bottom poll (in terms of O-R differential) for each target day, and add a random walk model on top. The most recent days incorporate 10-11 polls each; basically, all polls with sampling started & completed in the 12 days before the given day.
No assumptions about ground game relative strength, LV models being too restrictive or not restrictive enough, any Bayesian priors, etc. Just the dirt-simple base random walk.
With 7.18 days left, current diff of 3.083%, and a daily stdDev of 1.81965, a random walk model shows only a 4.5% change of Obama losing OH.
Updating my own trimmed-mean random walk model of Ohio. Just to recap:
---
I include all live polls (python+Mathematica), trim the top and bottom poll (in terms of O-R differential) for each target day, and add a random walk model on top. The most recent days incorporate 10-11 polls each; basically, all polls with sampling started & completed in the 12 days before the given day.
No assumptions about ground game relative strength, LV models being too restrictive or not restrictive enough, any Bayesian priors, etc. Just the dirt-simple base random walk.
With 7.18 days left, current diff of 3.083%, and a daily stdDev of 1.81965, a random walk model shows only a 4.5% change of Obama losing OH.
3 replies
= new reply since forum marked as read
= new reply since forum marked as read
