Scott Yang's Playground

An interesting article in 27^th June issue of Southern Courier (article can be viewed on-line with appropriate plugin). Article titled “Hillsong poll queried” on page 9, reporter questioned the result of the poll related to Hillsong buying the former RTA property at Rosebery, and how would that affect the neighbouring residents (corresponding article see Southern Courier issue 13^th June page 10). Outcome of the poll is overwhelming — 79% felt Hillsong will benefit the community. However when they actually dig into the statistics and HTTP logs, its result seems to be a bit sus.

From the article:

Further investigation revealed that 116 people had voted more than once, which amounted to 886 out o fthe total 1170 votes.

The monitoring system does not track any personally identifiable informationbut counts
unique internet protocol (IP) addresses.

The Courier traced 55 votes back to the IP address used by the Hillsong Church. The majority of the remaining votes were from Optus Internet and TelstraBig Pond IP addresses.

We gotta give applause to the analyst who did not just believe whatever the computer says the result is. Or maybe suggest that a proper poll application should at least track users base on IP address and/or cookies. Either way, we also know that there are certainly very zealous bunch of people at that church. Don’t put up Google AdSense banners on your own website, otherwise your zealous members might get you banned because of click frauds! :)

Can’t really blame them really. If there is a poll somewhere asking whether University Church is good on the UNSW campus and benefiting its community, I’ll also make sure everyone I know click on Definitely, multiple times, every day.