How confident can we be in the Fraud Bot's ability to sniff out fraudulent votes? Is there a margin of error?
I've inspected the cookies that get set, and don't think that by themselves there is enough data. Especially if someone is clearing their cookies every time. Now, if you are tracking their IP, timezone, frequency patterns and date/time of votes, (or some other set of super secret variables), that might be enough to computationally determine fraudulent votes.
Fraud Bot is designed to work without IP or cookies. Of course, the exact functionality is not revealed. The error tolerance becomes smaller the more votes are cast. I can say that it works well. However, it cannot achieve one hundred percent security. If you are planning something like a legally secure vote/election, you should use our »invitation required« feature.
Thanks. That's what I needed to know.
Can we possible to query the voting time of voters?
You can see the voting time in "Evaluation" -> "Participants"