Election Model Update 10/18/2016

[Loki Liesmith]

Search my journal pages for previous updates

Model Description:

Update with drift included 8/6/2016

Original Model Writeup 8/3/2016

#############################################
# INSTANTANEOUS MODEL
#############################################

Instantaneous Probability of a Clinton Win: 88.53 %
Mean Electoral Votes Clinton Currently Winning: 319.52
Median Electoral Votes Clinton Currently Winning: 323.0
Maximum Likelihood Electoral Vote Scenario for Clinton: 327

#############################################
# PROJECTING RESULTS TO NOVEMBER
#############################################

Probability of a Clinton Win if Current Trends Continue: 91.16 %
Mean Electoral Votes Clinton Projected to Win: 326.12
Median Electoral Votes Clinton Projected to Win: 329.0
Maximum Likelihood Projected Electoral Vote Scenario for Clinton: 329

Predicted National Point Spread (Clinton = +, Trump = -) 6.87

Comments: Drift and Instantaneous model continue to converge. Clinton national lead approaching that estimated by national surveys.

[unblock]

1. at the risk of complacency, i'm now focusing on things other than hillary's win percentage

like 6.87% estimated point spread (anything over 5.0% is rhetorically meaningful)

will she get over 50% outright?

and will we take the senate, ideally outright but 50-50 is theoretically good enough.

[Loki Liesmith]

2. Sadly

No Senate model and no time to make one at this point. Maybe I'll work on one after the election and try it out in 2018.

As for the national margin, My number work only with statewide polling spreads, so it can't project the actual outcome, only the final spread. However, if the spread is >6 I suspect Clinton will have to be over 50 by default.