My last post caused a bit of a stir. Camps seem to come down on two sides:

A) It’s okay if your poll is hackable. You just need to ensure every single last vote can be counted (even multiples from the same IP, which might be different people, or the same person masquerading as multiple individuals).

B) It’s okay if there is a possibility that a tiny percentage of voters might be turned away because several people on their IP have already voted. You at least know that your poll cannot be hacked.

The real point of the post was to poke fun at CNBC and their pulling of the Republican poll wherein Ron Paul was dominating at 75% out of 7,000+ votes.

Questions for CNBC:

* What percentage of votes were coming from the same IP or same IP subnet? If a large chunk of those votes (say 25-50%, or about 3,000) came from the same IP, then the poll was clearly “hacked” (or stuffed, however you want to call it).

This is easily determined if the poll app was written to log relevant data about each vote collected. (IP, User-Agent, etc)

And of course … it’s always possible that co-workers or students in a dorm (who might share the same IP or IP subnet) get together and collectively vote (once each) and cause 10-20 votes to come from the same IP block.

However, If I want to have a reasonable poll that obtains the results of a wide cross-section of the Internet population (hundreds of millions of eligible IP Addresses), then I have no qualms whatsoever with limiting, say, a single IP address to 10 votes.

It’s almost like… I can see people wanting to stuff ballots from the same IP, and objecting to my post because that will thwart their plans!

Sorry, Charlie — polls do not exist for your sole pleasure of stuffing the ballot — they are there to take the opinion of thousands of random strangers on the Internets! (not you, a few friends and your poll-stuffing code bots)