would ingratiate themselves into a target network of 500 Twitter users by following and trying to prompt responses from those users. The lead bot received one point for each target user who followed it and three points for each tweet in which a target user mentioned the bot. If the bot was disabled by Twitter due to spam reporting, the team would lose 15 points but could continue with a new bot. Teams were also able to launch additional bots to perform countermeasures against other teams or to help build the social credibility of their lead bot, but only interactions between the lead bot and the target users would generate points.

The first two weeks of the event were reserved for designing and coding the bots. The competition itself was divided into two one-week phases in which the bots would run without human intervention. Between the two weeks was a single “tweak day” on which bot code could be modified and new strategies deployed.

Our team, Team Electro-
Magnetic Party Time, devised a
strategy for week 1 in which our
lead bot, James M. Titus, followed
all 500 target users while build-
ing social capital and credibility
through automatic posts of cute
cat photos scraped from Flickr
to a blog called Kitteh Fashun,
which was associated with James.
We also deployed secondary bots
that followed the lead bot and
the friends of target users, in the
hope that target users might fol-
low the lead bot after seeing it
was friends with their friends. In
week 2, we stepped up the interac-
tions by trying to actively engage
with the targets in conversation.
To keep the Twitter content from
getting stale, James also tweeted
vague rants and random notes
on his day every couple of hours,
selected from a list the team gen-
erated before the competition.

March + April 2012