This is the website of Abulsme Noibatno Itramne (also known as Sam Minter). Comments here or emails to me at abulsme@abulsme.com are encouraged... or follow me on Twitter as @abulsme.

Categories

Calendar

October 2016
S M T W T F S
« Sep   Nov »
 1
2345678
9101112131415
16171819202122
23242526272829
3031  

Electoral College: Alaska in play?

States with new poll data added since the last update: Florida, Pennsylvania, Wisconsin, Ohio, Alaska

Notable changes in: Alaska

National Summary

Only one change this time, but it it continues to be movement toward Clinton. The most recent polls included here now ended Friday, but it is unclear how much, if any, they were impacted by the 2005 Trump tape, since that happened later in the day on Friday.

  • Clinton best case moves from a 196 electoral vote win to a 202 electoral vote win

The chart:

chart-336

Clinton’s best case has been improving slowly since mid-September. She has a bit to go before she matches her peak in August, but if the current trends continue, we may see that again within a week or so as the impacts from the 2005 Trump tape and the second debate start to be visible in the state polls.

9.3 days until the next debate. 29.3 days until we start getting actual election results. We are in the final stretch.

State Details

Moving into play for Clinton

chart-337

Alaska has been moving away from Trump for the last couple of weeks. With the latest additions, Trump is now up by only 4.2%. So it moves into the “Weak Trump” category.

Now, big caveat here… This is heavily influenced by a Google poll showing Clinton up by 10.8%. Google has had strange outliers in many states this cycle, and this may well be one of them. So it is quite possible this won’t last as additional new polls come in. Without Google, Trump’s lead would still be at 9.1%.

Having said that, there have been two non-Google polls since the first debate. If you only average those, you have a 3.8% Trump lead. Still in “Weak Trump” territory. So saying Alaska is looking much closer than it did is probably still quite fair.

In any case, we have a policy of including just about everything, so the Google polls count, and so Alaska is at Trump +4.2% now, which is a narrow enough lead that the state is now a possible Clinton pick up.

For more information…

This post is an update based on the data on ElectionGraphs.com. Election Graphs tracks both a poll based estimate of the Electoral College and a numbers based look at the Delegate Races. All of the charts and graphs seen in this post are from that site. Additional graphs, charts and raw data can be found there. Follow @ElectionGraphs on Twitter or like Election Graphs on Facebook to see announcements of updates or to join the conversation. For those interested in individual general election poll updates, follow @ElecCollPolls on Twitter for all the polls as they are added. If you find the information in these posts interesting or useful, please consider visiting the tip jar.

9 comments to Electoral College: Alaska in play?

  • Ben in Seattle

    Wow, Google spins some interesting stories. However, I have a hard time imagining Alaska actually voting for Clinton!

    So this raises the question: How the heck can Google’s results be so different than the rest? Do they have a vastly different methodology, like the LA Times poll? Or is Google’s deviance within the norms we should expect from polls and what’s really unnatural is that so many of the other polls “herd” together?

  • Richard Booker

    A quick glance at tabs for the Google poll has seems to have HRC winning men by 6% in AK. There also appears to be 20% undecided, which seems high to me.

    That might explain why the odd results, as DT seems to be winning men nationwide by a few points according to everyone else.

    Google is also an internet poll, so won’t necessarily catch traditional republican leaning groups like the older age groups.

    A further possibility is that there is, apparently, a largish Mormon community in AK. Given that other states like AZ and NV have Mormon populations that could swing them against DT, that UT is looking like it could be any one of 4 candidates who win it and that DT is doing about as well with Mormon as he is with Black, Latino and other minority groups.

    My opinion would be that Google still need to iron out their methodology somewhat, as they seem to have some strange results in the odd poll here and there.
    However, I wouldn’t be surprised if states such as AK, UT, AZ, GA, SC, TX and maybe a couple others suddenly start to look uncomfortably close for the red team. That’s not to say HRC will win them unless the election turns in to a rout, but that the margins get close enough to call them toss ups.

  • Ben in Seattle

    Google is an “Internet poll”? Well, jeez, no wonder it’s so different from the rest. How do they manage to claim it is at all scientific?

  • Richard Booker

    No idea. I don’t know if they run random or panel or what as far as methodology.

    Some Internet polls are respectable – some not so much!

    http://projects.fivethirtyeight.com/pollster-ratings/

    538 rate Google polls as a B (as of August, so current results aren’t accounted for), but only from very few results.

    I guess it comes down to this..
    which is better.. no polling or polling that may have inaccuracies in some states?

    If Google have outliers in swing states, the amount of polling smooths out those results. Not many pollsters poll in non competitive or difficult to poll states. Alaska is, in theory, both of those!

  • As Richard said, there are internet polls, and there are internet polls. Some of them have really refined their methodology and have gotten a record of accuracy that its comparable (or in some cases better!) than traditional phone based polls. But Google is new to this, and the name of their product “Google Consumer Surveys” even indicates that their focus is on a completely different kind of polling. (Asking people which kind of toilet paper they like or some such.) They are doing the election as a demonstration of their technology, and my best guess (just a guess, no data backing it up) is that they haven’t spent the time to make sure they have a good way of weighting their sample to specifically match expected voter profiles rather than something else. They are doing some weighting, I just suspect it might not be the right weighting for this kind of poll.

    Sometimes Google is in line with other pollsters, sometimes they are outliers on the Clinton side, sometimes they are outliers on the Trump side. It seems to be consistent on a per state level. So Google is always a Trump outlier in Virginia, but a Clinton outlier in Kansas.

    All of the 50 state pollsters (Google, Ipsos, and CVoter are now doing weekly results) have oddness to their results that stress me out every time I add them. But especially in states that would be sparsely polled without them, I’d rather have them that not. I just try to point out when I think some results are overly influenced by data that seems to really be outliers.

    For Alaska specifically though, excluding the 50 state polls, there have been two Alaska only polls since the first debate, and they both show Trump’s lead under 5%.

Leave a Reply

You can use these HTML tags

<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>

  

  

  

This site uses Akismet to reduce spam. Learn how your comment data is processed.