Astroturfing

A couple of days ago, one of Belkin’s reps got caught trying to hire people to write good reviews of their products. Given the spate of fake reviews and other games (see my earlier post about Eidos manipulating journalists), it’s not only clear that companies have an incentive to mess with reviews and create fake “user” reviews, but it’s happening. Like everyone else, I don’t like these types of tactics. I’m a consumer just like everyone else, so I don’t like it from that perspective. It’s dishonest and manipulative. Secondly, I don’t like it as a game developer because it undermines trust in the whole game-review process. I usually check reviews before buying a game or product, and if I think some of the reviews are fake, I’ll have less trust in positive reviews. As a result, I’ll be less likely to accept positive reviews – which leads to the reaction that I’m less likely to buy. In other words, fake reviews have the ultimate effect of hurting sales of good products. And, if you’re a company who doesn’t create fake reviews, well, users don’t know it and your reviews end up being lower than numbers for comparable products (because they’re padding their numbers).

It sort of makes me wonder if companies would go out of their way to make fake reviews of their competitors’ products.

Some other examples of fake “reviews”:

I happened to see some videos on GameTrailers.com for a game called “Rat Race” a few months ago. I think many of those votes are fake. First, I think they’re fake because they are remarkably positive for some bad video clips (could anyone really believe user votes are in the 7.1-8.5 range)? And secondly, if you compare the number of votes and comments to the number of views, they are remarkably incongruent. Specifically, they have five videos up. Here’s the numbers:

Video   Number of Views   Number of Votes   Number of Comments
1 78,413 598 199
2 45,600 513 173
3 40,094 474 171
4 21,617 501 189
5 30,785 509 196

Notice anything odd with those numbers? How about this: video #1 has almost four times as many views as video #4, but all the videos have almost the same number of votes and comments. Hmmm. Okay, I can’t expect the ratio of views-to-votes to be exactly the same, but it just seems suspicious.

A while back, Penny Arcade also had a comic about getting paid to write fake user reviews. I can’t find the exact post, but Gabe and Tycho said they discovered the existence of third-party companies who hire people to travel around the internet writing fake reviews for companies that hire them.

And, there was also that Amazon glitch a few years ago that accidentally showed the reviewer’s real names on the website. Authors were caught writing glowing reviews of their own books, but pretending to be someone else.

And, of course, there was a whole bunch of shadiness around Kane and Lynch. For example:

When viewing the official Kane & lynch website, a flash ad comes up with two 5-star reviews which… well… which don’t exist:

GameSpy did not say “It’s the best emulation of being in the midst of a Michael Mann movie we’ve ever seen” in their review of the game. They said that in their E3 2007 coverage. In other words, a preview. They also did not give the game five stars. They gave it three.

As for Game Informer, same deal. The highlighted quote does not appear in the review of the game. Nor do they give it five stars. Game Informer don’t even score in stars. They gave it a 7/10. (Source)

And I’ve heard that movies have a similar phenomena: there are certain people you can always count on to give a positive review. The movie companies know who they are, and then use them in their advertising, so they can say stuff like “a roller-coaster thrill ride! five stars!”. At least in that case, the “review” is part of the advertising, so we know not to trust it.

Leave a Reply

Your email address will not be published.

(optional)

Please note: if your comment doesn't appear right away, it's probably because it was automatically put into the moderation queue.