Archive for January 22nd, 2011

GWS Giants on Twitter: Follower location versus Tweet location

Posted by Laura on Saturday, 22 January, 2011

As an American, I have an obsession with the geography of sport.  City is fundamental to my identity.  I’m from the Chicago area and my sporting loyalties are based on that.  My sport team rivals are based on geography.  The Chicago Bears are going to play the Green Bay Packers in an NFL playoff game.  This geography based rivalry is more important then winning the Super Bowl.  We cannot lose to our hated neighbors to the north! (They wear cheese on their heads for pete’s sake!)

This concept of geography of sport is why I constantly want to look at location over other Australian based sporting numbers.  I’m that ugly American sort that some Australians don’t like because I feel like if I explain to you my reasons as to why geography should matter more to Australian sport, you’ll finally get it!  (And if you don’t get it, I clearly need to explain it to you more.)

Greater Western Sydney is one of the places where I occassionally get obsessed with geography, even if I don’t post about it.  Today, I decided that I should post about it and went about getting the data to do that.  First, I got a list of all of GWS related tweets.  These tweets came from archives I had on TwapperKeeper.  Beyond the GWS specific Twitter lists, I pulled the #AFL tagged tweets and then filtered out all the tweets that did not mention GWS or Giants.  I ended up with a list of around 996 tweets or so when all the lists were combined.  I then compared the user names with city/state/location information I had in order to attempt to match the tweets with a location.  The list I have to match is pretty huge, with Twitter user information from well over 1 million different user names.  When user names were matched up, I had 621 tweets that I had a location for.  (That’s pretty good as I normally get around 10 to 40% unknowns when looking at some follower lists.)  I mapped these on the map below.

After that, I got the follower information for @GWS_Giants.  They had 1,923 followers.  I had Australian city information for 1,057 of them. (In both cases, I removed all non-Australian cities.) I added these to the map below.

The purpose of this map?  It is to determine regional interest in the team.  My supposition, as that annoying ugly American who believes regional barracking is key to a club’s success, is that for the team to succeed, most of their support has to be based in New South Wales and the ACT.  I was pleasantly surprised at the number one city for Tweet location.  That appears to signal the team has done a decent job of attracting local interest in terms of getting locals to Tweet about the team.  Then I looked at where followers came from… and the problems seemed to be a bit bigger.  The GWS Giants’s market doesn’t really factor into their top Australian cities for attracting followers: Only 6 of their top 20 cities are in New South Wales or the ACT.  Ouch.  Other states just completely over represent.  To me, this signals problems in that their current target audience isn’t as large as the less relevant audience.  (Translation: Who cares if some one in Hobart follows the team unless the Hobart follower attends matches and buys merchandise?)

View full map

So a question: Do you think that the geography of the GWS Giants Twitter followers and Tweeters matters? Is this data relevant in analyzing how successful the club has been in garnering local support? Does the geographic spread surprise others?

Related Posts: