Election Model Update

[Loki Liesmith]

9/25 Update

9/24 Update

9/21 Update

9/22 Update

9/1/2016 update

8/16/2016 update

Update with drift included 8/6/2016

Original Model Writeup 8/3/2016

Model projections:


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

Instantaneous Probability of a Clinton Win: 76.73 %
Mean Electoral Votes Clinton Currently Winning: 297.65
Median Electoral Votes Clinton Currently Winning: 300.0
Maximum Likelihood Electoral Vote Scenario for Clinto: 296

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

Probability of a Clinton Win if Current Trends Continue: 62.97 %
Mean Electoral Votes Clinton Projected to Win: 284.22
Median Electoral Votes Clinton Projected to Win: 285.0
Maximum Likelihood Projected Electoral Vote Scenario: 280

Predicted National Point Spread (Clinton = +, Trump = -) 3.73

Analysis: New batch of Ipsos-Reuters polls bumped up the instantaneous forecast. Long range more or less the same (this is due to the variance of the states from which I'm sampling. Lean R states have been more consistent than lean D states. Still this long range number is better than what I ran (but did not post) yesterday.

Eventually both forecasts should converge.

[Doctor Jack]

1. You do good work

Although your username doesn't fill be with confidence.

[Loki Liesmith]

3. Lies, damn lies, and statistics my friend

Just different ways of making the truth more interesting.

[MyNameIsKhan]

2. Great job Loki, I like math myself