CO-04 (Special Election) See Full Big Line

(R) Greg Lopez

(R) Trisha Calvarese

90%

10%

President (To Win Colorado) See Full Big Line

(D) Joe Biden*

(R) Donald Trump

80%

20%↓

CO-01 (Denver) See Full Big Line

(D) Diana DeGette*

90%

CO-02 (Boulder-ish) See Full Big Line

(D) Joe Neguse*

90%

CO-03 (West & Southern CO) See Full Big Line

(D) Adam Frisch

(R) Jeff Hurd

(R) Ron Hanks

40%

30%

20%

CO-04 (Northeast-ish Colorado) See Full Big Line

(R) Lauren Boebert

(R) Deborah Flora

(R) J. Sonnenberg

30%↑

15%↑

10%↓

CO-05 (Colorado Springs) See Full Big Line

(R) Dave Williams

(R) Jeff Crank

50%↓

50%↑

CO-06 (Aurora) See Full Big Line

(D) Jason Crow*

90%

CO-07 (Jefferson County) See Full Big Line

(D) Brittany Pettersen

85%↑

 

CO-08 (Northern Colo.) See Full Big Line

(D) Yadira Caraveo

(R) Gabe Evans

(R) Janak Joshi

60%↑

35%↓

30%↑

State Senate Majority See Full Big Line

DEMOCRATS

REPUBLICANS

80%

20%

State House Majority See Full Big Line

DEMOCRATS

REPUBLICANS

95%

5%

Generic selectors
Exact matches only
Search in title
Search in content
Post Type Selectors
January 19, 2015 10:44 AM UTC

Busted: Local Conservative Blog Systematically Inflates Stats

  • 50 Comments
  • by: Colorado Pols

We interrupt our normal coverage of Colorado politics to notify our readers of something of…well, at least some importance, brought to our attention about a conservative blog that has been operating in our state for several years. Colorado Peak Politics has received occasional press attention as a source of Republican-leaning opinion and blog scuttlebutt, and has more or less replaced a hodge podge of GOP-leaning blog properties that have come and gone through the election cycles. Face the State, the People's Press Collective, and others you may have heard of previously occupied a niche now pretty much the exclusive domain of Peak Politics.

Peak Politics represents itself as highly influential, frequently comparing its influence to that of this blog, and sometimes even suggesting that they have "surpassed" us as the leading blog in Colorado politics. Now, we have no interest in getting into some kind of pissing match with other blogs, so we haven't seen any reason to engage these claims. But we'll admit that we have been impressed by some of the visible statistics for social media sharing on Peak Politics. For example, they recently posted an item about the Jefferson County school board drama that looks like it had a lot of Facebook shares.

peakpolssharecount

For a blog focused on Colorado politics, from our experience this would be a pretty solid indicator of social media "virality." Extrapolating the number of people who would have viewed all of these supposed shared instances, that number certainly indicates a respectable total readership for this blog post.

The only problem is, that number is fake.

We were directed this past weekend to a Facebook developer resource tool. This tool returns a simple text response via the Facebook "Application Program Interface" (API) to queries about any given URL on the web. Facebook, as you can expect, keeps very good track of the number of times a web page is shared within their system. In your browser address window, you can enter https://graph.facebook.com/?id=, then paste any URL on the web into the space after ?id=. Facebook will return a text page with the number of Facebook shares and comments to this URL.

Following this procedure for the URL for the above Colorado Peak Politics post reveals that it has been shared a total of 61 times on Facebook–about 17% of the number publicly indicated on their website. To be as generous as we can, we'll admit that the correct count of 61 isn't horrible either–but it is a small fraction of the total number they claim.

Once we started running other Colorado Peak Politics blog posts through this Facebook API checker, we found what appears to be systematic inflation of their indicated number of Facebook shares. A post yesterday attacking local Democratic consultant Laura Chapin shows 67 Facebook shares as of this writing. The Facebook API says it has only been shared four times. Another post about Gov. John Hickenlooper's State of the State address claims 24 Facebook shares, Facebook itself says there were exactly two. To varying degrees that we haven't found a pattern to explain, basically every single blog post on Colorado Peak Politics has been misrepresented in this way.

Sources tell us that what they are doing may not be technically complicated. Because they are using a third-party "share button" application, it may be as simple as the share buttons recording clicks for "shares" that are never completed through the API for Facebook–or Twitter, presumably, since the Tweet counts for their pages show similar signs of inflation. The best theory we have been given, in short, is that somebody associated with Colorado Peak Politics is clicking on the share buttons for their posts repeatedly to inflate their displayed count. No matter what the excuse is, we have been assured that the share count returned by Facebook directly via this query is correct.

Bottom line: in over a decade of operation, we've been able to interact constructively with local conservative blogs, middle-road blogs, and liberal blogs alike. We consider an influential and diverse blogosphere to be a critical piece of the total range of news and opinion available to voters and citizens.

But being dishonest about that influence doesn't help anyone. So let's please cut that out.

Comments

50 thoughts on “Busted: Local Conservative Blog Systematically Inflates Stats

  1. I always suspected their social media numbers were bogus. I think I can prove it's true for Twitter, too: if you click their Tweet button, the tweet it gives you includes @copeakpolitics. It should be easy therefore to see the 80 tweets they claim exist for the post in the image above, but I can't find them.

    I don't think Astroturf does this justice. I doubt it's illegal but it's fraudulent as hell.

      1. For those that haven't been following the dynamic scoring issue …

        Republicans in Congress Just Made it Easier to Cut Taxes

         

        House Democratic Whip Steny Hoyer said the rule change politicizes the accounting process. “What it means is the Republicans will be able to hide the true costs of tax cuts behind a debunked mantra that tax cuts pay for themselves. They do not,” he said on the House floor Tuesday afternoon, pointing out that deficit hawks should be up in arms, too. “This provision will allow them to explode the deficit, as they did the last time they were in charge.”

        Several Democrats quoted Bruce Bartlett, a former Reagan and George H.W. Bush administration official who condemned “dynamic scoring” as a sneaky and conniving political move. (emphasis mine)

  2. The number returned by the graph.facebook.com API call is just not supposed to be the same number as the number next to the Facebook Like button. The API call returns just the number of shares, but the number next to the Like button is shares + likes + comments + some others. See the Facebook documentation.

          1. All you've done is provide an additional way that the numbers may be going up, and provided no evidence at all to decide how much the two possibilities are affecting the number. Note that it's simply a fact that the Share/Like buttons aren't reporting what they seem to be reporting, and that's as documented by Facebook.

  3. I think I figured out how they are faking these numbers.

    http://support.sharethis.com/customer/portal/questions/914415-counter-increments-without-really-sharing

    QUESTION: We are noticing that when you click the Facebook share button, but don't complete the share, the page still increments the counter up 1, and the new value is visible to other users – basically inflating the number of shares.

    Could there be an implementation issue? It is happening on multiple sites with newer and older share this modules.

    ANSWER: This is the normal behavior of ShareThis counter buttons. If you wish to implement native counter buttons (available right now for Facebook, LinkedIn and Stumbleupon), you can make changes in your ShareThis code as shown below –
    <script type="text/javascript">stLight.options({ publisher: 'PUBKEY', nativeCount: true });</script>

    For pros and cons of native implementation, please refer our FAQs page – 
    http://support.sharethis.com/customer/portal/articles/517333-analytics-faqs#counts

    Regards,
    The ShareThis Support Team,
    support@sharethis.com

    Peak Politics uses the ShareThis sharing plugin, which WOULD use the Facebook API but it has its own counter (see above). But if you sit in front of your 'puter and click their Share button over and over again it runs up the ShareThis counter even though the share isn't really happening.

    If it was just once in awhile, or at least not the vast majority of their counts as it appears to be, it could be dismissed as an accident. But there's NO WAY it's an accident when Facebook says the real number of shares is a tiny fraction.

    NOTE I am not the person who gave this to Pols. But Peak Politics is totally fucking nailed if you ask me.

    1. Also, go look at Peak Politics' Facebook page. They don't have shit for likes, shares or comments. These fuckers are caught red handed. I'm amazed it took this long for somebody to figure this out, it's completely obvious.

      1. They're using ShareThis with the URLs of their blog pages, and not referencing their Facebook page. So there's no reason at all to expect matching counts on their Facebook page. It just means that most people are interacting with their site through the blog content, and not through Facebook. So, still no evidence at all of anything nefarious.

      1. That depends. It's a convenient solution for WordPress sites like ours, and we also use the ShareThis plugin. We do not utilize the counter function, not because we knew it was inaccurate but because it cluttered the layout of our blog when we tried it.

        We certainly won't use their counter now that we know it is so easily abused. The tech support message quoted above is exactly how the issue was represented to us by our source for this story, and settles the question in our minds.

        1. Though, ShareThis have abuse filters which run at the end of the day, to remove multiple clicks received from the same machine. It's not so easy to hypothetically run up the numbers as it might seem.

        2. >30% of ColoradoPols traffic comes from India.

          Country

          Percent of Visitors

          Rank in Country

          United States Flag  United States

          58.9%

          123,500

          India Flag  India

          32.5%

          81,194

           

          http://www.alexa.com/siteinfo/coloradopols.com

          Are they really interested in Colorado Politics in India or is Pols paying for traffic to goose its numbers?

          If you are going to complain about someone cooking the traffic books, it seems an explanation is in order.  Paid links, perhaps?

            1. dusty…..oh nevermind. He doesn't need permission – none of us do. I have the "style" button on my task bar, can shut him out at will, but he's been a bit more interesting lately, so I'll leave it open for now. Pols seems to have his a nerve, though, for a CPP writer like AC, and I'm enjoying that.

              As I understand it, a lot of cell phone and internet traffic is routed through servers which may be in India. So if folks are accessing through a mobile device Plenty of people are more knowledgeable about this than myself,

                1. Whatever. I actually like when people outside the US read my stuff. I track hits with bitly. Apparently, I have some fans in the UK and China.  It doesn't make me feel un-American or anything. I don't understand enough about the India server thing to even comment on it anymore, so I won't.

                  coloradopols facebook page sucks. It was last updated 2 years ago or so. It's apparently not a priority.  They should either keep it updated, (I offered to help), or take it down. 

                  So, hard to compare apples to oranges on FB likes and shares, although my own social network shares pols articles frequently.

                   

                  1. MJ,  They don't read your stiff.  They create bogus writings that mention ColoradoPols and have a link back to the site.  A few clicks on a few links makes the supercomputer in the sky think people are actually reading this in India.  

                    If people were interested about Colorado politics in India and were actually reading about it you would expect similar traffic on CPP.  They have zero traffic from India.  It is just ColoradoPols doing the same kind of stuff it is calling out CPP for doing.

                    1.  artificially pump up its numbers.  I do the same thing on behalf of a corporate web site.

                      You just admitted below that you are a professional liar…

                      why don't you just stfu and go away? No one cares what a liar thinks…

                    2. Ah well, I'll stick with my bitly stats for info on who's reading my posts and clicking bitly links. They seem pretty accurate. I do know that it's time I stopped reading your "stuff".

                      Other than knee-jerk defense of any idiotic thing Ken Buck does or says, and anything CPP or Redstate ever posts, you have little to contribute, and I can feel  brain cells dying when I get trappped in your snark battles. 

                      Thanks for reminding me why Pcat created the "NoPolsTrolls" style. 

          1. Alexa estimates internet traffic based only users who have installed the Alexa toolbar. The toolbar seems to be relatively very popular in (e.g.) India, and that dramatically throws off the numbers. For example, Alexa says that 2.3% of denverpost.com's traffic comes from India, and only 1.5% from Canada! That's not believable. I'd guess that the India numbers are off by a factor of 1000.

            1. Paccer, What it suggests is Pols buys SEO services from India and has links built there in order to artificially pump up its numbers.  I do the same thing on behalf of a corporate web site.  So Pols buys traffic from India and CPP tries to pump up its numbers from facebook.  Pardon the lack of outrage.

          2. We don't have an explanation for what Alexa shows about "traffic from India," and we are getting our hosting provider on the case to figure it out. Our best guess is that this is related to the spam we have to clear off the site most mornings, but we'll wait to hear from our providers to be sure. If there is this much clearly illegitimate traffic on our server, we want it stopped since our servers have trouble keeping up with the load as it is.

            And of course, we're not publicly displaying "traffic from India" to pretend we have relevance in Colorado. Thanks for bringing this issue to our attention, but as some kind of defense for what Peak Politics has been doing with their publicly displayed share stats…not so much. Sorry.

              1. I see you haven't been too busy to spend lots of time on this thread. But still nothing at all to say on the Brownback thread. Gee…. could it be because the Borg's got nothing?

            1. So you're saying that not even spammers can find a good reason to visit PeakGobbletyGook???? . . . 

              . . . Jesus, how sucky can a site get??!!??

            2. If the number of calls I get from Gary and Kevin and Jason with Indian accents telling me I have a Windows problem that needs fixing is any indication, sounds like a likely theory.  The record is three of those the other morning in an hour and a half. Probably accounted for more than half of the landline calls I got that day.

          3. All blogs are shameless self promoters. This is the pot calling the kettle black, nothing more.

            I remember when Colorado Pols was the only blog in town. I guess they're upset that they have competition.

  4. This is why I rarely read the comments here anymore (this one peaked my interest because of the SEO angle):

    1. Everyone has a case of the vapors over the awful, horrific, contemptable, sleazy, … – actually very standard practices.
    2. A couple of conservatives post some valid points and in reply get raw venom spewed back at them.
    3. Pols gets called out for some of their (very common) SEO actions, like links from India, and no one even owns up to it.

    Every successful (and most unsuccessful) websites does everything it can to boost its numbers, traffic, and SEO. Everything listed above is SOP, although the links from India are probably hurting Pol's SEO.

    Keep in mind that when your worldview is this hateful, it degrades your own happiness.

     

    1. You're saying it's normal for blogs to click on the share buttons a hundreds times to make it look like they have more traffic?

      You web masters are a devious lot then, Dave.

    2. We've been using the same 2 outlets to track our statistics for about 8 years now, so we can be confident in our traffic figures. In 2014, 97.08% of Colorado Pols traffic came from within the United States.

      We don't use Alexa, and they don't have access to Colorado Pols — they are just making guesses about our traffic sources.

      1. That's quite possible. If you don't have their Javascript in your pages then all they can use is info from people with their toolbar (a very small number) and try to find comparable websites (a big guess).

Leave a Comment

Recent Comments


Posts about

Donald Trump
SEE MORE

Posts about

Rep. Lauren Boebert
SEE MORE

Posts about

Rep. Yadira Caraveo
SEE MORE

Posts about

Colorado House
SEE MORE

Posts about

Colorado Senate
SEE MORE

187 readers online now

Newsletter

Subscribe to our monthly newsletter to stay in the loop with regular updates!