The Edge recently held a political poll on whether Anwar Ibrahim should quit as the Opposition leader–But when the editor begun to see that the one-week survey attracted 12,736 responses and the responses were overwhelmingly one-sided, she smelt something fishy.
Upon further checking with the IT team, they found that 6,354 of the responses came from one IP address, and about 1,700 came from several IP addresses within the same building. Another 2,000 responses came from seven different IP addresses.
Now to put that into laymans terms–imagine you conducted a mail in poll, where people would submit their votes via regular snail mail, and you conduct your polls on a regular basis getting between 2,000 to 4,000 responses every time. Now imagine you start a new poll, and suddenly you get 12,000+ responses, with more than 6,000 of those responses coming from just one address.
This reminds of a conversation I had with the Digi WWWOW team sometime back, apparently in the first installment of the DigiWWWOW awards, the award was given purely to the person who had the most votes, and so some brilliant techie figured out he could buy the votes at a cost of far less than the prizes given out–and so he won the WWWOW awards, and his blog didn’t even last the year.
Unlike everyone else, I can safely say, these aren’t cybertroopers. You see for all those responses to come from just one address, and the rate at which they were arriving–this had to be automated rather than manual. Which meant this was more botnet army than cybertroopers.
Unless of course, there’s a huge market for people to do low-grade clerical work, by clicking links and voting in polls. I still doubt these are cybertroopers, the quickest most efficient way to get these kind of results to through some automated scripting from a single source, rather than having an army of people clicking and refreshing links.
Even Anil Netto, who posted a similar poll on his blog noticed some peculiar patterns:
An identical poll on my blog showed that it was being manipulated when the ‘Yes, he should quit’ responses were increasing at a rate of seven per minute. The first 300 responses had indicated that some 90 per cent felt he should not quit but that fell rapidly to around 55 per cent. After I raised the alarm about the possibility of the poll being rigged, the “Yes, he should quit” responses slowed and eventually the “No, he should not quit” responses outnumbered the ‘Yeses’ by over 60:40.
This band of cybertroopers will stop at nothing. It is all par for the course for them. But they are not fooling anyone.
Just visit any truly controversial political blog in Malaysia (I won’t link them here, but they’re not hard to find), and you’ll find allegations of cybertrooper thrown at nearly every blog comment, and they’re two versions, either you’re an UMNO cybertrooper or a member of the Red Bean Army (the colloquial term for a DAP cybertrooper).
Think of the cybertrooper as someone whose willing to wear a Manchester United Jersey with ROONEY on the back–to a LIVERPOOL tour match. The colloquial term here is ‘troll’, which basically means you’re there just to cause trouble. In my experience, the debate going on in these blog comments do nothing more than just provide an avenue for people to state their thoughts, which is usually on the extreme end of the spectrum (no such thing as moderation here). They do not change anybody’s mind. I’ve never seen people in the comment section go, “I was wrong, thanks for pointing that out”, that just doesn’t happen. No one is changing anyone’s mind on these blogs and everyone is just having a go at each other–which is fine, because these people go to these sites with that as their intention.
However, trolling a blog post is one thing–trying to influence a poll result on a independent news site is another. That’s just blatant lying.
While these guys are cybertroopers, partly because they’re taking part in the conversation, there’s also a botnet army out there looking to bias polls in their favor and it’s the botnet we should be worried about–not the cybertroopers.