Apparently a bunch of states (23?) are using a software system called Interstate Crosscheck to look for "double voters".<p>> Election officials in more than two dozen states have compiled lists of citizens whom they allege could be registered in more than one state – thus potentially able to cast multiple ballots – and eligible to be purged from the voter rolls.<p>The problem is that it often only uses a persons name as a singular data point. So if a person votes with the same first/last name as another person in another state, it's possible that vote could be wiped out. It was even matching names even though there were differences in middle names or had Jr/Sr at the end.<p>The journalist, Greg Palast, who investigated this back in 2014 has been doing radio circuits again recently saying that it's still being used in a bunch of states. Not sure about the validity of this since there hasn't been much reporting elsewhere on this. He seems to be the only one talking about it. And googling 'Interstate Crosscheck' only brings up his articles and democrat superpac websites.<p><a href="http://projects.aljazeera.com/2014/double-voters/index.html" rel="nofollow">http://projects.aljazeera.com/2014/double-voters/index.html</a><p><a href="http://www.rollingstone.com/politics/features/the-gops-stealth-war-against-voters-w435890" rel="nofollow">http://www.rollingstone.com/politics/features/the-gops-steal...</a>
It's pretty but what is this really telling us? There's a 126% increase in searchers for "Voter intimidation" in Hoback, Wyoming. OK, so does that mean that someone's being intimidated? Planning to intimidate? Curious about the news? Does a "126% increase" mean there were 7 searches today instead of just 4?
I just had to vote "provisional" myself (Palm Beach County).<p>I have a feeling there is going to be a lot of fraud reports against the Florida Board of Elections. We registered almost 2 weeks before our Oct 11 deadline, yet neither my wife nor I were in their system - we received no mail or anything from the BoE.<p>My vote, I'm absolutely certain now, will be thrown in the trash. Even if it isn't, the election will be over by the time it's "counted".<p>Think I'm simply spreading FUD? Take a look at the 2004 Florida general election... this state's election board is wholly corrupted.
A 100% increase in searches for "voter intimidation" in my city (Seattle) is interesting. I'd love additional context on who is searching for that (i.e. is it the intimidators or the fearful?)<p>Nonetheless, this seems like the beginning of an interesting tool. What would it take to do some sort of fuzzy matching on related searches, like broken voting machines for voting machine problems? I suppose you could wait for a related term to breach a threshold and begin tracking it with related terms.
So far all of these just look like population maps of the US, skewed by time zone.<p>Basically if it's a heavily populated area that is awake, then the dots are big.<p>Maybe as the day goes on it will level out to more interesting insights?
Google did something similar for the flu without any significant result[0]. Did things improved ?<p>[0] <a href="https://www.wired.com/2015/10/can-learn-epic-failure-google-flu-trends/" rel="nofollow">https://www.wired.com/2015/10/can-learn-epic-failure-google-...</a>
I'm a bit peeved polls didn't open until 9AM in New Rockford ND when the state (and a fair number of farmers) said they would open at 7AM. Wonder what that would be on Google's chart.
Sortof OT: I was also interested in election trends. While Google is focusing on real-time search trends regarding voting, I looked at meme trends leading up to the election. There are interesting trends in the amount of attention different candidate memes received over time. It will be interesting to see whether the attention received by candidates is translated into votes, today.<p>Here's the working paper: <a href="http://iandennismiller.github.io/election-memes" rel="nofollow">http://iandennismiller.github.io/election-memes</a><p>EDIT: quick link to viz: <a href="http://imgur.com/S6nHNLT" rel="nofollow">http://imgur.com/S6nHNLT</a>
This is pretty, but it's all statistical outliers; every highlighted spot is a small city where random fluctuations above a low baseline are most likely to generate seemingly-impressive spikes in query volume.
This is cool, but seems to me to be very especially vulnerable to the Observer Effect. Any fluctuation in these search terms that shows up here will cause people to try and find out more by searching for those terms themselves. What may have started out as random noise gets fed through a feedback loop and amplified, while not necessarily being signal.<p>I saw an increase in searches for voter intimidation near where I live. I immediately went to DuckDuckGo, and "!n voter intimidation". Now I'm part of the problem, apparently :)
What's the visualisation library they're using here? Looks great (I'm <i>guessing</i> D3, but it's hard to break through the webpack file to figure it out).
I wish this were normalized by electoral college size, it would go part-way to normalizing by population but also be normalized with respect to impact on outcome.
I love the American elections! The data, the coverage, the websites, the data-driven-campaigning... all of it.<p>I mean, I'm glad I don't live there and have to choose between these disasters of candidates and see either of them hand over the control of the world to Asia... but I do love the elections as a platform!
Particularly intersting is the 'inactive voter status' band along the cotton belt (a region of African Americans in the southern sates where the voting rights act was recently repealed). Yet no such searches in KS, CO, NE, WY etc.
While we're on the subject of voter intimidation, there has been constant intimidation of Trump supporters throughout this whole campaign[0]. And at the same time, Democrats had a deliberate campaign to incite Trump supporters to violence[1] although this required provoking them by infiltrating their private events, while Trump supporters were harassed and abused on the streets.<p>Progressives have this insane argument that goes that violence and intimidation against Trump supporters is actually ok and not contrary to our deepest values, because it is done by private citizens and not the government.<p>[0] <a href="https://www.youtube.com/watch?v=VUBpRexwiPg" rel="nofollow">https://www.youtube.com/watch?v=VUBpRexwiPg</a><p>[1] <a href="http://www.breitbart.com/big-government/2016/10/17/exclusive-okeefe-video-sting-exposes-bird-dogging-democrats-effort-to-incite-violence-at-trump-rallies/" rel="nofollow">http://www.breitbart.com/big-government/2016/10/17/exclusive...</a>
This is part of the larger Electionland project, involving over 1000 journalists & journalism students.
<a href="https://projects.propublica.org/electionland/" rel="nofollow">https://projects.propublica.org/electionland/</a>
As someone trying to inspect dots, it's very annoying to me that the detail windows pop up and can block other dots. I'm not sure how you'd stop this but I found this very clunky to use. I'm also not sure if it offers anything useful...
Geographic profile maps which are basically just population maps. <a href="https://xkcd.com/1138/" rel="nofollow">https://xkcd.com/1138/</a>
I just really want to know how many estimated ppl voted for each candidate in each state. Couldn't they just extrapolate that based on search/browsing history (IE. if someone visited BreitBart, they're probably voting for Trump, if someone visited 538, most likely Clinton)