City May 22, 2009 9:04 AM

Weekly Poll: Unprecedented "Turnout"?

Weekly Poll: Unprecedented “Turnout”?
As you may have noticed, our weekly poll has received a tremendous amount of participation.  As of this writing, we've received just over 76,000 votes.  We're flattered, really.

But the pace and style of 'voting' has raised some questions.  At 4pm yesterday, there were approximately 11,000 votes.  At 11pm last night, there were about 36,000 votes.  And this morning, we're at 76,000 votes.  Now, our traffic has been growing steadily, but it's unlikely we  had 40,000 unique visitors overnight all of whom chose to vote for either Byron Brown or Mickey Kearns. 

Instead, it appears some folks have figured out how to game the system.  Multiple browsers, clearing caches - that kind of thing.  This isn't exactly War Games level hacking, but it's something people would need to spend time and effort to manipulate.  We're working on uncovering where many of these "votes" have come from and we hope to be able to share some insight over the next few days. 


View image

Comments

Leave a comment

it will be particularly interesting to see if any of the problem IP's originate within City Hall.

Score: 0 ( 0 votes ) Vote up Vote down Report this comment

Particularly since both candidates work in City Hall and have factions within those walls.


I certainly hope the "Gang of Five," with greater force due to the sheer numbers in each of their districts, do not have a stronghold. And, going further, no one else, as well.

replied to Sean Brodfuehrer
Score: 0 ( 0 votes ) Vote up Vote down Report this comment

That baby looks stunned by the power of democracy in action.

Score: 0 ( 0 votes ) Vote up Vote down Report this comment

The lead swung back and forth between the candidates yesterday. BRO should have mentioned that. Seems both camps may be stuffing the ballot box.


Funny that Would Not Vote actually comes in 2nd. But that could be stuffing, too.

Score: 0 ( 0 votes ) Vote up Vote down Report this comment

I voted for will not vote as I don't live in the city limits. But i don't think that answer provides for that level of specificity. I take that answer to mean i refuse to vote for either candidate.

replied to Dagner
Score: 0 ( 0 votes ) Vote up Vote down Report this comment

I have been watching this poll with interest and was surprised to see that Brown captured all of the 50,000 votes that were logged in last night. Great job Casey and crew. You guys did a great job. Hope you didn't lose too much sleep.

Score: 0 ( 0 votes ) Vote up Vote down Report this comment

The pro-Brown overnight stuffing came after clear stuffing for Kearns yesterday.

replied to R.T.
Score: 0 ( 0 votes ) Vote up Vote down Report this comment

Suggestion to BRO: provide a means of commenting on the "weekly" surveys. This situation highlights the need, but it existed before, such as when your selection of best (or favorite or whatver)Streets was very Elmwood Village centric and left out many great streets. Or always provide an "other" choice.

Score: 0 ( 0 votes ) Vote up Vote down Report this comment

I supposed there's no point in having the poll up there if it is so easy to hack and is constantly inaccurate.

Score: 0 ( 0 votes ) Vote up Vote down Report this comment

Drifter, I think you have the wrong perspective on this. The poll is used to gather options in a relaxed fashion, the thing that is pointless is hacking it. Measures could be put in place to safeguard gaming the polls a bit more, but why bother? The results are just for general information, unless the city decides to forgo an actual election and use the weekly poll on buffalorising to determine the winner this year.

.

The really amazing thing is that both sides took the time to automate a voting mechanism.

.

Also, the poll is not inaccurate, it is a simple tally.

Score: 0 ( 0 votes ) Vote up Vote down Report this comment

If the IP addresses are captured when folks vote in these polls (and if they're not, they should be), I'd suggest dumping any stats available (IP address, time of vote, etc.) into a simple document and posting it where people can "crowdsource" it, looking for any irregularities or patterns -- like ArtVoice did with some recent info they obtained through FOIL.

Score: 0 ( 0 votes ) Vote up Vote down Report this comment

FOIL is meant to allow the public to view government agency records, server access logs are not public information. BR won't expose it's user traffic data to the public but the data is being analyzed.

replied to RaChaCha
Score: 0 ( 0 votes ) Vote up Vote down Report this comment

Unicorn, glad to hear the data is being analyzed.


Please don't misunderstand what I said about FOIL -- I wasn't suggesting that in this case. My reference was to a situation earlier this year where ArtVoice posted some raw data that had been requested from City Hall (subpoenaed by Common Council, as it turns out, not FOIL'd) online, so that folks who were interested could pore over it and sort through it, etc. looking for anything of interest. In other words, "Crowdsource" it. Details here:
http://blogs.artvoice.com/avdaily/2009/04/02/the-library-berc-bura-bnrc-documents/

replied to Unicorn Ninja
Score: 0 ( 0 votes ) Vote up Vote down Report this comment


How about one sign-in, one vote?

Score: 0 ( 0 votes ) Vote up Vote down Report this comment

A fourth option of 'neither' or 'other candidate' would have been helpful as well. I plan on voting, but not for Kearns or Brown, unless they are the only candidates on the ticket.

Score: 0 ( 0 votes ) Vote up Vote down Report this comment

This type of thing happens with pretty much any online poll, so I don't pay much attention to the results. I often get email or Facebook messages urging me to go vote in this or that poll.

Score: 0 ( 0 votes ) Vote up Vote down Report this comment

There is as much reason to assume that only one party did this. Unless the timing was off, aka a lot of only one vote on one day and the opposite the next day.


Whatever program was used to hack this probably had variables in it to vote for all three categories to make it seem likely that it might possibly be legit.


if 50k votes came in for one person only it seems obvious. If you spread them out it seems a little believable. The fact that Brown 'won' with 56% (just over 50% of all total votes) makes it seem likely someone in his camp did most of the hacking.


If both were hacking then why would either side have stopped when they did.

Score: 0 ( 0 votes ) Vote up Vote down Report this comment

You people cannot be serious. Do you actually think Byron Browns people or Mickey's supporters give a crap about a poll on BRO?

That number is completely fabricated from the inside to "generate" interest.

The total votes number can be set to say whatever BRO wants it to say.

Don't be stupid.

Score: 0 ( 0 votes ) Vote up Vote down Report this comment

More votes were cast in this poll than the 2005 mayoral election (73,073). Congrats.

Score: 0 ( 0 votes ) Vote up Vote down Report this comment

Leave a comment

Buffalo Rising Poll