No, Google Isn't 'Pushing' People To Vote For Bernie Sanders
from the push-it-real-good dept
Every political season, the conspiracy theories come out to play. I personally like the one that suggests that all the high ranking world leaders are actually lizard-people from the planet Bilderburg, or however that one goes. But the slightly more believable conspiracy theories can be fun, too, such as the one that says Donald Trump is a Clinton plant to ensure Hillary wins the Presidency on the theory that not enough American voters are actual crazy enough to elect Donald Trump as president. Which... yeah.
Anyway, the fun aside, it's disappointing to see the mainstream media get in on this conspiracy theory action, such as when the Washington Post runs an article with the headline: How Google Is Pushing You To Vote For Bernie Sanders. The actual article doesn't quite live up to that headline, of course, since accusing Google of trying to get Bernie Sanders elected through search results doesn't make any sense on any number of levels. Still, the piece is about an analysis done by Slate that appears to get things exactly wrong in the most un-scientific way possible.
In an analysis published by Slate, Daniel Trielli, Sean Mussenden and Nicholas Diakopoulos tallied the top 10 search results for each of 16 Democratic and Republican presidential candidates. Then they had a crowdsourced team of reviewers rate whether each of these results painted the candidate in a positive, negative or neutral light.Ok, so through the method of getting some people out there to Google a candidate's name and then decide for themselves whether a search result is positive or negative, the general trend is that Democrats get one better search result than Republicans. For anyone outside of the group of my conservative friends that gets their GOP knickers in a twist at the slightest hint of liberal bias anywhere, this is a statistical nothing, particularly given the muted sample size of results. It also seems to think many people are going around and inputting presidential candidates into Google and then making up their minds based on search results. An assumption for which no evidence is provided, mind you.
For instance, if the top search result was a candidate's own Facebook or campaign page, that would qualify as an overwhelmingly positive result. But if a news story critical of a candidate came up in the top 10, it would be coded as negative. In they end, they found that "Democrats fared better than Republicans when it came to supportive and positive sites within the first page of results. Democrats had, on average, seven favorable search results in those top 10, whereas GOP candidates had only 5.9."
Even the article notes that nothing about this indicates that Google is trying to influence voters.
Does this mean Google is in the tank for Democrats? No. The authors say that the outcome is "an organic, emergent result constituted from a complex prism of quantification involving hundreds of signals and increasingly complex and opaque artificial intelligence." In other words, we don't know a whole lot about Google's algorithm. But we do know that pages that get a lot of traffic or are frequently linked to on other websites tend to do better in the rankings.Left out of this conversation is the fact that Sanders, agree with him or not, is a candidate with a populist message. That helps him in the search results as well, as many diverse groups of people will be discussing him and his campaign, likely all linking to one another and so on. But regardless of that fact, the point is that the premise of the article is exactly wrong.
That helps explain why Democratic presidential candidate Sen. Bernie Sanders fares so favorably in Google searches -- nine of his top 10 results were rated "very pro" in the analysis. The senator from Vermont has a strong social media presence and a lot of followers on multiple platforms. So his Facebook, YouTube and Twitter accounts all come up high in the search results, contributing to an overall positive result.
Google search results aren't influencing voters; voters are influencing Google's search results. Coupled with some personalized results, Google search results are all about the most commonly read, linked, used, links out there. The more positive links about Sanders, or any other candidate, are popular, the higher they appear in the results. Same with negative links.
So, while it's tempting to think the big evil corporation is trying to get us to vote for the socialist (because that apparently makes any kind of sense), the truth is more mundane.
Thank you for reading this Techdirt post. With so many things competing for everyone’s attention these days, we really appreciate you giving us your time. We work hard every day to put quality content out there for our community.
Techdirt is one of the few remaining truly independent media outlets. We do not have a giant corporation behind us, and we rely heavily on our community to support us, in an age when advertisers are increasingly uninterested in sponsoring small, independent sites — especially a site like ours that is unwilling to pull punches in its reporting and analysis.
While other websites have resorted to paywalls, registration requirements, and increasingly annoying/intrusive advertising, we have always kept Techdirt open and available to anyone. But in order to continue doing so, we need your support. We offer a variety of ways for our readers to support us, from direct donations to special subscriptions and cool merchandise — and every little bit helps. Thank you.
–The Techdirt Team
Filed Under: bernie sanders, conspiracy theories, presidential candidates, search, search results
Companies: google
Reader Comments
The First Word
“Subscribe: RSS
View by: Time | Thread
Clickbait
Did WaPo hire the same people to make up titles for their stories that scuzzy clickbait articles use?
[ link to this | view in thread ]
Only the Best
[ link to this | view in thread ]
Ignorati
Any results from this type of "research" reveal more about the people doing the searches than any bias inside Google.
[ link to this | view in thread ]
Of course at this point it could arguably be Google's responsibility to make sure that their results aren't being played with.
Google also tends to sort results based on the people searching for them and their past history of pages they tend to visit. So naturally a republican searching for a republican candidate will get positive results while that same person searching for a democratic candidate will get negative results.
[ link to this | view in thread ]
Re:
It would be interesting to see if the results of these 'studies' change from location to location, computer to computer, and to what extent and what factors may influence the results and how.
[ link to this | view in thread ]
Why can't it be both? The possibility of a feedback loop is not zero.
[ link to this | view in thread ]
[ link to this | view in thread ]
Re:
[ link to this | view in thread ]
https://www.youtube.com/watch?v=slkQohGDQCI
[ link to this | view in thread ]
[ link to this | view in thread ]
Hmm
Probably not. But if they were, how would we know?
[ link to this | view in thread ]
Re:
and I'm all outta bubblegum.
[ link to this | view in thread ]
Re:
Like if Hitler was running against Gandhi, then there would be some real bias. All those pesky facts causing articles about Hitler to be negative. Shame on you Google.
[ link to this | view in thread ]
Re: Hmm
[ link to this | view in thread ]
Undercurrent
For The 2016 Presidential cycle, the Republic candidates so far are mostly nut jobs. Trump is a populist racist who gets tons of bad press and bad social media. The rest of them are equally lame and are all pretty much the butt end of jokes (did you here the one about Jeb's Exclamation mark walking into a bar?). It's what people are talking about, I guess. The Democrats are pretty much all on the lower key side of things, so they don't get the type of negatives on social media, blogs, and the like.
It reflects poorly however on Google because their results are more of a popularity contest rather than an attempt to get you the best results. The best results in searching for Candidate would be their official site, their facebook, their twitter, and so on. Having "Donald Trump: UK 'should be thanking me' for Muslim comments but Muhammad Ali condemns politician" as the first result when searching for Trump seems to be a failure in SERPs. Google can be seen to be leaning because of how they rank things.
Google has become a giant echo chamber for a world way more interested in style over substance, rumor over fact, and opinion over information. It's a feedback loop that just amplifies the very best and very worst, and doesn't seem to serve the people as well as it once did.
[ link to this | view in thread ]
We also know that the search history of the user is taken into account, their geographical location, and many other factors that the researchers themselves could be inserting into the results. I'm sure that's why they crowdsourced things, but it opens up a huge variability even on search on the same computer by different users.
We also know (at least in my impression as a foreigner) that the "left" tends to attract people who would know how to correctly SEO a site, which means that their candidates might have their own stories returned more positively. While the "right"'s positive spins may be overshadowed by coverage of whatever batshit thing was said today to compete with Trump's insanity.
That's certainly a generalisation, but this study is clearly useless as anything other than chumming the troll waters with another anti-Google/anti-Democrat headline.
[ link to this | view in thread ]
Re: Undercurrent
"It reflects poorly however on Google because their results are more of a popularity contest rather than an attempt to get you the best results."
Erm, yeah, that's how it works. If you're merely searching for a candidate's name, what else would you want it to return to you other than the top results associated with that candidate? If you're after something more specific (such as, say, their Facebook page), then you need to tell Google what you're searching for by including the desired keyword.
"Having "Donald Trump: UK 'should be thanking me' for Muslim comments but Muhammad Ali condemns politician" as the first result when searching for Trump seems to be a failure in SERPs."
No, that's the top news story of the day relating to the subject searched for. What, specifically, is a failure on Google's part by showing the top news story as a result of a search of their news section?
Also, unless your layout is different to mine, you're also lying (what a surprise!). That news story is the second result returned (and the first news story). The first actual result above the news section is Trump's Wikipedia page, then his official Twitter feed appears after a small block of news stories. Following that are his Facebook, official biography, a Forbes profile, etc. The first critical result outside of the news stories are a couple of Guardian stories at 9th and 10th position.
It's not Google's fault that his (by your own admission) batshit racist rants are the things that are the most popular news results, but they sure as hell aren't the top general search results, barring some major differences introduced by your own search history.
"Google has become a giant echo chamber"
No, that's your underpowered mind, which seems incapable of addressing reality when your preferred fantasy will do.
[ link to this | view in thread ]
Re: Undercurrent
http://snag.gy/upQLY.jpg
[ link to this | view in thread ]
[ link to this | view in thread ]
WHY is it "DISAPPOINTING to see the mainstream media get in on this CONSPIRACY THEORY action" ?
are you a coincidentialist or you haven`t got the news?
Yes the WAPO article is asinine because they totally ignore how the internet and google work, not because they are arguing political campaign and big money could collude to influence voters.
[ link to this | view in thread ]
Re:
Wapo ignores the fact that
Google gives you the result of what it thinks YOU WANT TO find.
That is why people do this stupid game in the parties where everybody gets diferent search results in their cellphones when looking for scary political news
[ link to this | view in thread ]
ignore the fact that YES OF COURSE there are companies that offer the service of "management" of your google results and probably most of the campaigns have one of this arrangements...
even they could hire the same companies to fsck the google results of the competition?? DON`T THEY?
[ link to this | view in thread ]
Re: Re: Undercurrent
In searching for a candidate's name, I would expect to see something like Wiki near the top, the official site, maybe the party official site, etc. The news is "next", not first. If I wanted news, I would search for, say "Donald Trump news".
"It's not Google's fault that his (by your own admission) batshit racist rants are the things that are the most popular news results, "
See, that is where you miss part of my point. Should Google provide the most visited sites / most linked sites as the best results, or should they return something else that would be a better result? Social signals are not for finding the right answer, otherwise we wouldn't need a site like Snopes.
"No, that's your underpowered mind, which seems incapable of addressing reality when your preferred fantasy will do"
As always, fast to the insult and slow to figure out that not everything works only one way, the way you expect it. Sorry to prove you wrong (AGAIN). It's why I rarely answer you anymore.
[ link to this | view in thread ]
Re: Re: Re: Undercurrent
So, you admit that you assertions above, based on the results you personally saw, were your usual high pile of bullshit? That your idiotic rant was based on something that only you saw? Good to know.
"In searching for a candidate's name, I would expect to see something like Wiki near the top, the official site, maybe the party official site, etc."
Then, what in your personal history prevented you from seeing this in the way I did? Because that's exactly what I saw, outside of the news results.
"The news is "next", not first"
So, as I saw it then. You seem to be whining like a child over Google not having the page layout you want, which is pretty stupid. Would them having the option to turn off news results make you happy, since you're incapable of parsing the results with your own brain?
"Should Google provide the most visited sites / most linked sites as the best results, or should they return something else that would be a better result?"
How do Google determine what is a "better result" if you only specify someone's name? You object to them using the most accessed and/or linked sites, so what criteria do you want them to use?
"As always, fast to the insult and slow to figure out that not everything works only one way"
So that's obviously why I've offered several comments, including the one you replied to, where I directly address multiple possibilities. Your inability to address reality without lying your ass off is noted, yet again.
[ link to this | view in thread ]
Re: Undercurrent
Chomsky is overrated, putting it mildly.
"It reflects poorly however on Google because their results are more of a popularity contest rather than an attempt to get you the best results."
Your obscure definition of "best results" proves exactly my point. Who is to say your "best results" are any more appropriate than anyone else's "best results"? You want to be the puppeteer to pull the very strings you deplore.
It's the same mentality that is currently coming from here in the UK in regards to Leveson, in regards to Scottish nats lashing at the BBC for "bias", in regards to GamerGate thinking the approprate response to SJWs thinking gaming media is biased against their interests is to say gaming journalism is biased against THEIR interests, Corbynistas claiming every media criticism is a "smear", and now (predictably, as Google is naturally a target for conspiracists) that Google is the big, evil, moustached organisation pulling all the strings of our collective brainwashed unconscious or something.
[ link to this | view in thread ]
Crazy Americans
There were enough to elect Ronald Reagan!
https://www.youtube.com/watch?v=jmWLJmbytkk
Enjoy!
[ link to this | view in thread ]
User bias
[ link to this | view in thread ]
Re: Re: Re: Re: Undercurrent
No, but thanks for trying troll and shit me. My results are likely the same for many people, as are yours.
"whining like a child"
Thanks for the personal attack. That added so much to the discussion.
" where I directly address multiple possibilities."
No, you asserted that your version of Google is correct, and mine is "bullshit". Stop wasting my time old man, I'll get off your lawn.
[ link to this | view in thread ]
Re: Undercurrent
This is where you lose me and likely everyone else reading this. For a site like Google that is trying to return the best search results for the masses, the most popular ARE the best. If you're saying it would be better if Google returned the most INFORMATIVE results, sounds great, except who gets to decide which are the most informative? Popularity is really the only objective game in town for a search engine, with a nod to weighting certain results that ought to show up near the top, which is why for most search items Wikipedia DOES indeed end up near the top of the list (at least for me, I understand that others may have a different result).
"Google has become a giant echo chamber for a world way more interested in style over substance, rumor over fact, and opinion over information."
Now that REALLY doesn't make any sense. Google is returning search results, not feeding people information they want. Tailoring search results to past search history might have SOME effect in paring the results down, but claiming that a person who leans one way or another on any issue will only see those results in their search field is pretending to live in a world that doesn't exist.
[ link to this | view in thread ]
Without getting into the actual merits of the bias claim here, I'd have to take issue with you dismissing it as a "statistical nothing". This is ranked on a scale of 1 to 10, and so the difference between 5.9 and 7 is 11%. That's huge, particularly when applied to politics. (Just look at how many major races are won by a margin smaller than 11%!)
Just saying...
[ link to this | view in thread ]
Re: Re: Undercurrent
See, that's where we disagree. The most popular are not the best, they are only the most popular (ie, most discussed or active on social media). By standards of popularity, Rick Astley should be the top response for anything to do with music, internet memes, tricks, and all sorts of other things, because 161 million views say he is pretty much the best of music, right? His video is popular, but it's not the best result for many of those terms.
" Google is returning search results, not feeding people information they want. "
It's a feedback loop. It's not about personalization, rather it's about ranking results based on social media mentions and how many facebook likes they get. It's no longer about considering what is really on the page and it's actual relevance, replaces by a system that all but ignores content for a heavy dose of "what people are talking about". So if more people link and talk about Trump being an idiot instead of linking and talking about his campaign site, then idiot generally gets a better ranking - even if the idiot page has little to do with Trump.
[ link to this | view in thread ]
Re: Re: Re: Undercurrent
You missed the point; we DON'T disagree here. I never said the most popular was the best objectively, only that it was the best in terms of the business of search results. As I said, for a search engine, popularity is the only game in town. Google can't try to make a value judgement on the "goodness" of search results, or it will lose users who may not want what they/you/I think are the "best" results.
[ link to this | view in thread ]
Re: Crazy Americans
Ronald Reagan v Walter Mondale: Reagan the incumbent was succeeding in foreign policy and many domestic policies though the public was pissed about his actions with the air traffic controllers. Mondale, however, offered nothing concrete that would be better than Reagan, and being Vice President under Jimmy Carter didn't help.
[ link to this | view in thread ]
Re: User bias
[ link to this | view in thread ]
[ link to this | view in thread ]
http://www.latimes.com/nation/politics/la-na-trump-focus-group-20151210-story.html (For your own sanity, turn on your adblocker before visiting that page.)
[ link to this | view in thread ]
High ranking world leaders
[ link to this | view in thread ]
Re: Re: Re: Undercurrent
[ link to this | view in thread ]
Re:
Unless the margin of error is, say, 12%. ;-) This "study" sounds so bad I don't know how you would even calculate a margin of error. 12% sounds way too low.
[ link to this | view in thread ]
Slate
[ link to this | view in thread ]
Re: Re: Re: Re: Re: Undercurrent
Yep. Hence my point. You went on a moronic rant based on your subjective result, yet the mere fact that other results happen for other people make your point... well, pointless. You're whining at something which even you admit changes from person to person, and trying to make some sort of attack on Google for returning results that match a person's desired outcome. Pretty stupid, no?
"Thanks for the personal attack. That added so much to the discussion."
An accurate observation, nonetheless. As ever, you focus on where I've offended you with these observations and avoided answering any of the questions I've posed to you. Such as - how do you want Google to operate, since you outright reject the algorithms that have made them such a demanded and useful service? I know it's easier to pull the kindergarten act, but you could at least act like you passed puberty, surely?
"Stop wasting my time old man, I'll get off your lawn."
That's OK. When you get past whatever strange adolescent phase you're going through that makes you act like this, I'll welcome you into the community that's capable of addressing the real world.
[ link to this | view in thread ]
Re: Re: Re: Undercurrent
Heh, you can't even get your popular culture correct. Sadly, outside of trolling morons on the internet 5 years ago (your peer group, I understand), the most popular results today would be Adele and Justin Beiber. That might pain people who prefer their music differently, but it seems that another segment of objective reality has passed you by.
"It's not about personalization, rather it's about ranking results based on social media mentions and how many facebook likes they get"
This is where you cite the fact that Facebook likes have a direct effect on Google's search algorithm. Because, yet again, you're full of shit.
"So if more people link and talk about Trump being an idiot instead of linking and talking about his campaign site, then idiot generally gets a better ranking - even if the idiot page has little to do with Trump."
Unless you look at the results that were returned for the other people you refuse to admit exist, where Trump's official pages were returns for 8 out of 10 results outside of the news articles. Plus, you still haven't explained why Google's top news results should not be the actual top news results, nor why your inability to look at the results outside of the news section magically makes the other results disappear..
[ link to this | view in thread ]
This from the paper
It really is the end of an era. There is WPBJB and WPAJB. (Washington Post Before and After Jeff Bezos) It is sad. They used to be journalists, now they are just another group of poo flinging monkeys.
I thought it was funny a Saudi prince weighed in on Donald Trump. If Americans wanted a prince's advice, they'd ask Jeb Bush wouldn't they?
BTW, what evidence do you have that Donald Trump ISN'T a false flag operation? Doing that much damage to the Republican party on accident? If so maybe we should elect him just to keep the DOJ and Congress busy with impeachment for the next 8. Like a piece of paper left on a table that says on both sides: "how do you keep an idiot busy, turn over for the answer."
[ link to this | view in thread ]
Re: Re: Crazy Americans
Mostly things outside his control.
So the mood was that Reagan couldn't be any worse.
Outside the US most of us thought Reagan was a joke! See video I linked to above.
Nothing that happened changed our minds.
see this : https://www.youtube.com/watch?v=s2n87YKSjrA
At best he was an actor reading somebody else's script.
Personally he was a romantic who was easily swayed by whoever talked to him last. This is amply demonstrated by the walk in the woods with Gorbachev.
Turns out Reagan was better at foreign policy
Luckier maybe, but better?
Iran Contra anyone?
Jimmy Carter was actually one of the best presidents you ever had - his record since leaving office underlines that. Unfortunately he was about the unluckiest president you ever had.
[ link to this | view in thread ]
Re: Re: Hmm
That's what I thought.
[ link to this | view in thread ]
Re: Re: Re: Re: Undercurrent
They love stuff like that.
[ link to this | view in thread ]
Re: This from the paper
Please read that sentence again. The man (and his message) is popular. THAT is the problem. That all the GOP have to offer is recycled failed policies based on their own notions of religious propriety and outdated, disproved fiscal ideology is THEIR OWN failing. Basically, they're crap and they're out of ideas. What would the new guy be like? The old one. This is what happens when you pander to the base, ramping up the fear factor instead of working for the service of the country as a whole.
This is not an endorsement of t'other lot. They're not much better.
If so maybe we should elect him just to keep the DOJ and Congress busy with impeachment for the next 8. Like a piece of paper left on a table that says on both sides: "how do you keep an idiot busy, turn over for the answer."
"He's awful but he'll have to do." Really? Break out of the Red V Blue teams game and vote for someone else. You need to build consensus enough to get the numbers to get your party or president of choice in but it has happened before and will happen again if you don't get caught up in worrying about the other side.
[ link to this | view in thread ]
Re: Re: Re: Re: Re: Undercurrent
Far better than answering simple repeated questions here such as "if you're so offended by Google's results, how would you like them to be calculated", I'm sure.
[ link to this | view in thread ]