Why Buy-A-Vote Sites Will and Do Fail
Why don’t they get it? All of the major vote-rigging sites have sold off or are in the process of being sold. A new one pops up, destined to fail and be sold (Buyer beware, there is a reason why such a seemingly profitable venture always gets sold in a couple of months). SpikeTheVote – Sold. UserSubmitter – On The Auction Block. Why is it that all of the buy-a-vote sites fail? And why will they continue to fail? It is actually quite easy to catch and marginalize these sites.
|
- Super-History Tracking
If you havent seen this yet, then you need to. Using a simple CSS+Javascript hack, it is quite easy for a website to determine your browser history, or at least check to see if you have visited certain sites. Every time you load a page, your browser changes the way that links look if you have previously been to that page (you know, it turns the blue link purple). Using javascript, a web page can inject a large number of links and check to see if your browser changes the style. If it does, it knows you have been there. Thus, in a matter of a second or two, a site like Digg, Reddit, Slashdot or any of the other web 2.0 sites can not only check to see if you have visited a site like Subvert and Profit, it can even check to see if you have been to pages that you can only reach by logging in! Digg can know not only that you have been to the bad guys’ sites, but that you have logged in there as well. It. Is. That. Easy.
Quick Example:
1. Visit Here… http://ha.ckers.org/weird/CSS-history-hack.html
2. Now, imagine, instead of just printing out the links you have visited, it prints out img src=’tracker.php?site=www.digg.com’. Now, his server can use that image call to record both your ip and that you have visited digg.com before. - Timing is Everything
There is a natural pattern for how people vote. When a story goes up, it has 30 minutes to 1 hour to get most of it’s votes, before it gets shuffled off several pages deep and won’t be seen again for a long time. When stories stop meeting this pattern (ie: votes cast over several days with different referers and no early adoption) things start to become obvious. This kind of timing issue along can destroy the effectiveness of Digg-spamming networks where voting is based solely upon when the user decides to log in. Don’t you think it is pretty obvious that some guy logs in every day, votes on 5 new stories and 1 that was submitted 13 hours ago?
- Spam Reporting Lowers Reputation of Network Users
After a while of voting on crappy enough Paid-for-Digg stories, the simple reputation algorithms will devalue user votes so heavily that they will become utterly usesless. After a while, the ineffectiveness and expensiveness, plus the pattern of low-reputation, flagged users voting heavily on their stories, will make the program a waste of cash. Maybe that is why sites like SpikeTheVote and UserSubmitter decided to cash-out so early.
- Where You Vote Matters
Most voters/diggers/redditors vote on the story headline, not visiting the unique comments page on the site and voting for it that way. These services provide direct links to the comments pages. Voters aren’t perusing through the site, they didn’t perform a search and find a story, they went directly to the comments page and voted. This kind of pattern easily sets apart vote manipulation. Yet another easy reason why these Buy-A-Vote sites will and, do, fail.
- Big Targets are Easy
A. Lawsuits. Yeah, it hasn’t come to this yet, but if one of these sites ever did succeed, it would be really difficult to avoid getting sued. Bear in mind here, they may never lose the lawsuit, but these small vote-rigging sites don’t have the war chests that a Digg or Reddit does.
B. User Revenge. People love their online communities. While there has been little precedent for it, I can easily imagine someone dropping a site like UserSubmitter into a copy of SpamVampire. Or, even easier, using the power of the Digg Community’s massive traffic to just bury a site like this into the ground with bandwidth charges. I’m not advocating it, but it definitely could happen.
The truth is that the methods of marginalizing buy-a-vote sites are both easy and numerous. This is just a small list of the tens if not hundreds of other simple algorithmic changes that a site like Digg, Reddit or Netscape can, do, and will make to reveal, purge and defeat such buy-a-vote networks.
If you really want to take advantage of the traffic that these web 2.0 communities and Social Bookmarking Sites offer, then take a look at Viral Marketing Ethics and then start writing content that is worthy of a viral campaign.
19 Comments
Trackbacks/Pingbacks
- japadamus.com » Digg rigging ultimately fails - [...] GoogleCache has a great article on why vote rigging websites ultimately fail. I think its interesting that people are…
- Why Every Buy-A-Digg Site Fails. (and always will) « News Coctail - [...] Buy-A-Digg Site Fails. (and always will) Filed under: Uncategorized — recar @ 8:17 am Why Every Buy-A-Digg Site Fails. (and…
- . .. . . .. the sux stops here .. .. .. . . .. … .. . .. … . . . .. . . .. . » Digg versus Wired and Reddit - Hot Three-Way Action - [...] is too funny … a very intelligent-sounding report just came out about how Digg can’t be corrupted. Of course,…
- pligg.com - Why Buy-A-Vote Sites Will and Do Fail... Why don’t they get it? All of the major vote-rigging sites have sold…
- Pronet Advertising - 'Subvert and Profit' Tries to Subvert Top Digger, Profit From It... Subvert and Profit is soon to become the latest…
- Subvert and Profit - Erfolgschancen? (6/7) | NetDesignz - [...] der Lage geworden ist und die Website wieder offline geht. Für Zweifler hat Russ Jones bereits 5 Gründe gesammelt,…
Great read, with some cool info. Thanks 🙂 I didn’t know about the JS history thing, it sounds like a really cool idea. Gonna have to do some further reading on it.
Once again, thanks 🙂
Digg will go down in history as “ME TOO” version 2.0. There is no such thing as wisdom of crowds this is why 90% of the post comes from 25 people. Then every 13-19 year old click “Digg this” they just automated posting ME TOO!
Look at the recent articles, they are either The BEST $x EVAR or blog spam. Digg was a great idea for about a week, but it has become much worse than Slashdot ever could have dreamed to become.
1. Wrong. The browser does this locally. You CANNOT do what you claim.
Reply: Proof of Concept: You can easily use javascript to measure something and then record it by simply having javascript print out html that calls an image on your server. ie: if the digg.com link changes color, print out [img src=’proof.php?site=reddit’]. Of course
e, proof.php isnt an image, its a php file that stores the site and your ip address into a database.
3 and 4 are just GUESS on your part as to how hte algoirthm works. Give one piece of evidence to show that digg a story from the comments page carries a lesser weight than doing it from the front page. ANY evidence will do. Of course, you have none.
Reply: It is a guess, and they probably don’t do this yet. But they could, and it would work. Which is the point of this post, that these manipulation sites will, in the long run, fail.
5A. The lawsuit would never get past the first stage, and thus would not cost the website much money in lawyers fees. You can’t sue without any merit. Digg would lose the VERY first time it got to a judge, and would likely face punitive damages for filing a baseless lawsuit for malicious reasons. Hence, the site would MAKE millions off Digg.
Reply: While digg may lose, they would certainly not face “abuse of system” charges for filing a lawsuit against a company that profits by encouraging people to violate the terms of service of your site. I won’t get into the legality of it all, that is not for this post. However, it is not nearly as egregious as the type of lawsuit that would bring about an abuse of system counter suit.
Insightful post! Last I saw, User/Submitter was being auctioned for FIFTY THOUSAND dollars. I’m not sure if it actually sold or not, but either way they still seem to be preying on digg.
All good points, except 5A. Buy-a-digg sites would *kill* to be sued by Digg, because 1) it would be tantamount to Digg saying that they are effective, and 2) it would be tons of free advertising. Bandwidth is cheap. Lawyers are cheap. Even settlements are cheap, in the face of huge increases in business.
1) You honestly think that Digg is going to add javascript to every page that exploits a browser vulnerability to check every one of its users history? That is ridiculous. Digg would have a huge backlash against it and lose a lot of reputation. And for what? Preemptively preventing a few shill stories until this vulnerability is fixed in a few weeks?
2) You have way too much stock in timing. I really doubt that Digg stories always follow a very predictable pattern. What’s the difference between me posting a forum topic encouraging fans to Digg my website versus a paid site that encourages shills to Digg my website? If all you’re looking at is the timing, it’s the same.
3) You are assuming that it is easy to track shills. That is an invalid assumption. You are assuming that shill accounts would Digg very few stories and the stories that are Dugg happen to correspond to other shill accounts… In that case, yeah, any code monkey could detect them and trivially stop the entire problem. However, a more interesting situation comes up when you are trying to fight off competent shills. What if the company has a distributed network of bots, each with a unique IP and each with an impeccable, but automated Digging history? How would you stop that?
4) This is just wrong. A large chunk of stories are promoted to the front page because the article has a fat “DIGG THIS” button on it and the websites fan base promoted it. Either way, this is trivially avoided by removing the link to the comments page and having people Digg it in the magical way you are talking about.
5a) Is it really illegal to pay someone to Digg a story? If so, then this is really your best bet, although a trivial defense is to relocate to another country.
5b) This is also trivially avoided. The only sites that are really susceptible to DoS attacks are those that are computationally expensive, e.g. each hit requires multiple SQL queries or PHP scripts to execute. If you are talking about trying to increase their bandwidth costs, that is also only a problem if they have large images or movies. A plain website is almost impossible to be DoS’d and uses no more than a speck of bandwidth.
You forgot to mention that a member of the Digg team can simply signup at the services and see which stories are being “promoted” or whatever they want to call it.
Ad1: You can turn off JavaScript
Ad2: You can pariticipate as normal digg user.
Ad3: see above
Ad4: You can disable referrers. And see above.
Ad5: Ok, this is true
actually, wired successfully abused digg by buying votes … link to that story and other tales of digg’s obvious corruption at: http://allsux.com
another assumption you’re working off of is that a site being sold = a failed site. Its very possible they owners are getting out with a quick buck! *gasp* people sometimes sell things to make money! */gasp*
Jesus, this somewhat scared the shit out of me. I saw the article about SAV on TechCrunch yesterday and as my normal curiosity usual leads me, I checked out the site. I’d never even consider using it for profit, but I’m just naturally curious about lame sites like that and I find it amusing to see their list of reasons as to why their service will work.
But now I feel paranoid about getting bumped off Digg just for checking out a site. Surely the systems you mention in your first post aren’t going to print up false positives?
I have to disagree with point 2.
Whenever I troll through the upcoming stories section, I notice all those stories (we’re talking about 8+ pages deep) have been submitted in the last 2 hours. or so, but when you look at the sidebar “Hot in [Category]” these are stories that are popular but not yet on the front page (or might not ever get there) typically have been submitted 15 hours ago, give or take a few hours.
And out of the section, you might only find 1 story worth digging. Easily pegging you into the profile that your vote has been bought.
And easy way around this would be to have early response people, submit story, notify your early response people by instant message, vote it up to “Hot in [Category]” . which might need as little as 13 votes to get into, Then once its easily visible to the masses, inform the rest of your bought votes to vote on it as well.
Digg would never institute the css hack because it’s not 100% working. Basically users would find out, it would be so obvious to find out and it would also require storing a huge database. Digg will eventually (if not already) be exploited and soon after we’ll move elsewhere.
I’m already starting to see a huge amount of digg blogger pages that make it to the frontpage and are nothing more then 1 legitimate article in an otherwise spam filled site.
1 – Just disable Javascript (that’s what I’ve done with NoScript & Firefox). And there are legal implications for a site like digg to record users browsing history.
2 – Vote buyers just alter their timing
3 – Perhaps, then just use new accounts and/or vote along with the popular stories aswell.
4 – Vote buyer just provides a unique link to the story/comment page
5 – War chests? Up till a few months ago reddit was 4 guys. And when you start suing, you’re saying that your application is flawed and you need to resort to legal measures (not technical) to stop them, which lowers the percieved reliability of your site.
You know, these top 5 reasons are bunk. For 5 detailed and hilarious counter-reasons check out:
http://allsux.com/2007/04/04/want-to-share-something-that-sucks/
If someone really wants to they’ll succeed.
Well, I recently heard from another website that seemed to work kinda different from the ones you mentioned. They dont actually buy votes but seem to have a lot of contacts in Digg and use them to promote content from their clients. I think the url was something like socialelves.com. Another good thing is that they don´t promote bad stories so I guess it avoids spam stories to become popular on digg.
Every day I am always more astounded about WWW marketing because of understanding the way in which the the next generation interact with the online world through technology. My 11 year old grand daughter just sent me a web system they had launched to aggregate cool topic for their classmates. They were interested in knowing the best way to launch advertising on the system to generate money. We are very poud.
epic fail.