Rome is Burning: A Case Study in RSS Fraud

There is a contest being run at Daily Blog Tips right now called “Blogging Idol”. The contest is pretty simple: Get as many RSS subscribers as you can during the month of July.

For full disclosure, I’m entered in the contest and I’m currently in about 4th place, so I do have an incentive in making sure that the contest is run fair. The website I have in the contest is Everything-Everywhere.com.

Odds are, even if what I have to say in this post is accepted, I wont win. I’m cool with that, but I sure as hell don’t want a cheater to win. The current leader of the contest is Rome of Romeuy.com, and Rome is a cheater. EVERYTHING about his RSS stats reeks of fraud, and there is NOTHING which points to a legitimate increase in subscribers. Everything about what he has done defies statistical probability to the point where it should be taken are prima facia evidence of fraud. If this were a court of law, the statistical evidence would be enough for conviction. Everyone has been tiptoeing around saying outright that he is a fake, but I’ll step up and say it. Romeuy is a Fraud.

The Basics
If it looks like a duck, and walks like a duck and quacks like a duck, it is probably a duck.

Romeuy.com is a gaming site started a few months ago. There are very few examples I can think of where a blog has been launched and has managed to get over 1000 subscribers in a short period of time. The only examples I can think of are when well-known bloggers or celebrities launch new blogs and the new blog is heavily promoted on other websites. In other words, there needs to be traffic, buzz and links before there are subscribers. I dare say it is tautological.

Romeuy.com is a gaming site without any particular focus. It doesn’t focus on a niche like World of Warcraft, which might drive a particular group of fans to the site. Moreover, game sites are a dime a dozen. I know because I used to run one (Stomped.com) seven years ago.

Links

The site has few incoming links. If you go to Google and search for link:romeuy.com you will get a whopping FIVE links. Oddly enough, all five links come from the same website. A make money online website. Even more oddly, that make money online website has the EXACT same layout as Romeuy.com. In fact, the make money online site in question is owned and operated by the very same person behind Romeuy.com.

Does operating a MMO site make someone guilty? No. Does it raise suspicion? Oh yell yes. In fact, all of the fake RSS numbers I’ve seen in the past, with the exact same statistical signatures as the Romeuy.com site, have been MMO sites. Even many MMO sites operators who are legit will tell you that the niche has gotten a bad reputation because of the number of shady blog operators.

For the purposes of this discussion, Romeuy should be thought of as a Make Money Online site, not a gaming site. There is another MMO site he is associated with as well. He seems to have more irons in the MMO world, than in the gaming world. (I believe his MMO site also has fraudulent RSS numbers, but he didn’t enter that in a contest. MMO sites with fake RSS would raise a flag much quicker)

So, this guy is getting record levels of RSS subscribers with ZERO links coming from other gaming sites. How about Technorati? They show 77 links, with the majority coming from other MMO websites. Most of the links are also generic links to the site, not links to specific articles, which you’d expect if there was some sort of Digg or Slashdot effect going on.

Moreover, if you take out links dealing with the Blogging Idol contest, there has been no real increase in links during the time which his RSS numbers have been exploding. Certainly, his site doesn’t exist to the rest of the gaming world.

As another bit of evidence, his Google Page Rank is 0.

Traffic

I will hold it out as an axiom that RSS subscribers is, in general, a reflection of traffic. To subscribe to a site, people have to first visit the site. Moreover, not everyone who visits a site will subscribe. Many people don’t like RSS or just prefer to visit a site on a regular basis.

In addition to the amazing increase in RSS subscribers without incoming links, we are also to believe that he has done it without an increase in traffic. If you look at his Feedburner data, you see the remarkable fact that his RSS subscribers is several times larger than the number of hits he has incoming. Not visits mind you, but hits (or as feedburner is most certainly counting it, page views). That means that number of actual people going to the site each day is even less.

Go and look at the Feedburner data for any site which you believe actually has a lot of readers. You will find the Feedburner hits to be greater than the subscribers. Again, this pretty much has to be true by definition. People have to come to the site to subscribe. I’m going to get into the numbers with even more detail, but I believe that when subscribers > hits, by Feedburners own data, that should be prima facia evidence of fraud.

Check his Alexa data. His Alexa ranking is 273,312, which very odd for a site with so many readers. If you look at the traffic numbers, you’ll see no spike in traffic during his increase in subscribers. In fact you’ll see no growth at all.

I realize that Alexa is far from perfect, but it is usually in the ballpark, at least by an order of magnitude. Even if Alexa isn’t evidence by itself, it certainly doesn’t contradict any of the other fishy things about this, and is another bit of evidence which goes on the pile.

He has an Entrecard widget on his site as well. His popularity is 17, which means over the last 5 days, he’s gotten 17 clicks on his widget. Not 17 people mind you, but 17 clicks, 5 of which could have been his other MMO site. Traffic isn’t coming from Entrecard.

Also, compare his comments with other gaming sites with supposedly similar number of subscribers. You’ll see anywhere from 20-100 comments per post. Certainly some will have big spikes. He doesn’t.

Hard Data

If you look at any Feedburner subscriber chart, it shows behavior like the stock market. The overall trend might be up or down or flat, but with any short period of time you will see fluctuations in the data. This is due to how Feedburner counts subscribers. The act of subscribing is nothing more than telling your RSS reader to go to a particular XML file. If your RSS reader doesn’t poll the XML file, Feedburner wont count it. This is why you see dips in subscribers all the time. Weekends or holidays, some people might not check their RSS.

This is why any trend which shows a monotonically increasing number of subscribers should be suspect at face. That is just how the system functions in the real world.

The dips are usually within bounds however. The biggest dip I’ve seen my RSS numbers take in one day (other than the February ’08 Feedburner problems which had every site show a dip in subscribers) was 15%. Even a 10% is usually pretty big and rare. I had a big dip about a week ago which was 9%. This is also a statistical signature of any Feedburner graph. The graph wont be perfectly smooth, but it shouldn’t have wild swings either. Just like the stock market goes up and down on a daily basis, you wont see a 90% drop in the stock market in one day, only see it go back up 90% the next. Historic one day drops in the stock market are in the neighborhood of 10% (the 1987 Black Monday drop was 22%) . While it is possible, the wilder the swing, the greater the improbability of it happening.

Lets look at the Romeuy.com Feedburner graph. There are multiple things about this which should strike you as fishy.

1) There are more subscribers than hits. As stated above, and by other bloggers who have investigated RSS fraud, its pretty much evidence of fraud by itself. (I should note that in almost every article on this subject you can find where someone investigates a case of RSS fraud, it shows up with the same signatures you see here).

2) Except for a few MAJOR swings, the RSS numbers have been monotonically increasing since June 12. When they go up, they never go back down. (more on the decreases later). Not only are his subscribers not visiting the site, and not finding it via links, but they are visiting every day, and in increasing numbers every day.

3) The biggest one day increase was on June 28, when the subscribers when from 463 on June 27 to 713 on the 28th. What happened on June 28th? The final $3,000 prize for the Blogging Idol contest was announced.

4) Since June 28, you will notice two days with enormous drops in subscribers.

June 28 – 713 subscribers
June 29 – 48 subscribers
June 30 – 759 subscribers

July 5 – 934 subscribers
July 6 – 49 subscribers
July 7 – 1000 Subscribers

The first drop was a 93.6% daily drop, and the second was a 95.1% drop. Such drops are really unheard of. It would make more sense if they dropped to zero. The webserver might have been down or something. But they didn’t drop to zero…..

AMAZINGLY, they dropped down to almost the exact same number each time: 48 and 49.

What could explain this?

The Smoking Gun

The fact that 95% of the subscribers are failing in unison means that they are using the exact same system, whatever that is (probably email. That is how most RSS scams work. Get bogus email addresses). The fact that they check every day without fail and the numbers are monotonically increasing shows they aren’t human beings. They don’t show the stochastic behavior which humans systems display (stock markets and every other legit Feedburner graph being good examples)

The extreme dips plus the monotonic increasing subscribers pretty much shows that there is some sort of automated system behind his amazing increase in RSS subscribers.

All other anecdotal evidence (Alexa, Technorati, Google, Entrecard, Feedburner hits) points to this site being nothing special and not one which has an enormous number of followers. NOTHING points to the contrary.

Conclusion

This stinks to high heaven and it should be obvious to anyone who is in anyway familiar with RSS numbers.

Romeuy.com is a mediocre gaming site without 50-60 real RSS subscribers, the rest of which have been purchased from a service like RSSxplosion.com (I wont provide a link to them. Cut and paste it yourself).

With a $3,000 prize, even if he spent $1 per subscriber, he’d still make a profit.

Most people like to play nice and don’t want to cause a fuss and point fingers. Everyone has been dancing around this, but no one has been willing to come out and say it. ALL evidence points to fraud. NO evidence points to real subscriber growth. I think the burden of proof is on the other side now.