Maps of Johnson vote in each state

[amdcpus]

Looks about right. I'm surprised you caught his strength in Indiana, if he gets 5% nationally, he'll probably easily get 10% in Indiana. In 2012, when he got .99% of the vote nationally, he got slightly over 2% of the vote in Indiana.

[amdcpus]

Looks about right. I'm surprised you caught his strength in Indiana, if he gets 5% nationally, he'll probably easily get 10% in Indiana. In 2012, when he got .99% of the vote nationally, he got slightly over 2% of the vote in Indiana.

I don't really think we can extrapolate from the small universe of hardcore libertarians who voted Johnson in 2012 to his potentially much larger base of support this year.  In particular, I think he will do better in the Northeast than you are assuming and not as well in places like Missouri and Indiana that are on the whole very amenable to Trump.

I'd be willing to bet Indiana is one of his top 10 best states in the % of vote received in the state. Also,  the Libertarians got 4% in the gubernatorial race that year, which was their 4th best gubernatorial percentage out of all the states that year.

[amdcpus]

Also, what does the 538 model say about Johnson's strongest and weakest states?

I went through the polls-only model state-by-state (tedious, but I couldn't find all the data in one place) and here are the current forecasts for Johnson percentage in each state.

1   NM   18.3
2   AK   14.6
3   UT   14.2
4   MT   13.9
5   SD   13.7
6   WY   12.8
7   ND   12.3
8   CO   11.8
9   NE   11.5

44   ID   6.5
44   NJ   6.5
44   TN   6.5
47   HI   6.0
48   AL   5.9
48   DC   5.9
50   CA   5.5
51   MS   3.8

The Idaho number is wrong, especially now after Survey Monkey found 19% of their Idaho voters supporting Johnson. 538 has Idaho's polling average for Johnson to be 14.5% and their projected vote share at 12.8%, which would put Idaho as shared 6th on the list with Wyoming, instead of the 44th place it finds itself on above.

Sigh.  The model has updated since I pulled the numbers, and the addition of that poll did jump it all the way to 12.8%!  It's quite a dramatic spike in the graph at http://projects.fivethirtyeight.com/2016-election-forecast/idaho/. 

Well, that was a snapshot of the model, as of about 7pm EDT on September 6, 2016.  If anyone wants to go through the exercise and update all the numbers, feel free.

Yep he had a big jump in Idaho numbers today as WaPo/SurveyMonkey released today which had him at 19% and a Dan Jones Idaho poll released today that him at 13%, an increase of 8% from their poll a month previous.

[amdcpus]

Looks about right. I'm surprised you caught his strength in Indiana, if he gets 5% nationally, he'll probably easily get 10% in Indiana. In 2012, when he got .99% of the vote nationally, he got slightly over 2% of the vote in Indiana.

I don't really think we can extrapolate from the small universe of hardcore libertarians who voted Johnson in 2012 to his potentially much larger base of support this year.  In particular, I think he will do better in the Northeast than you are assuming and not as well in places like Missouri and Indiana that are on the whole very amenable to Trump.

I'd be willing to bet Indiana is one of his top 10 best states in the % of vote received in the state. Also,  the Libertarians got 4% in the gubernatorial race that year, which was their 4th best gubernatorial percentage out of all the states that year.

Was that when they had a quasi-celebrity Survivor castaway as their nominee?  That could explain some of that.

Yes it was lol.