Archive for July 25th, 2010

Twitter Weekly Updates for 2010-07-25

Posted by Laura on Sunday, 25 July, 2010
  • RT @BrianGainor: 40 Ways Sports Teams Can Utilize @Foursquare #Sports20 #
  • Australian elections: The geography of Twitter: #
  • Use of #ausvotes appears to have peaked last night around 7pm: So right before #masterchef :) #
  • The extent of my ability to make animated movies. Should be titled: "Swans can't play AFL footy." #
  • RT @anthonyalsop Looking Back at Digital Sport Summit – 5 key learnings #dss10 #
  • Been sitting on it since June, posting now: Impact of Akermanis’s Comments on the Western Bulldogs’s Fanbase #
  • I problematically switch between @purplepopple and @ozziesport as I treat each like its own filter and get confused with replies. #
  • I think I found a tool which should make it easier for me to get a list of locations of where people follow an account are from. #
  • Though ug. :( Limit of 100 people. :/ #
  • Not sure how useful sentiment analysis would be for sport… Lots of neutral reporting or views buried inside links. #
  • I have about 3,000 AFL related tweets saved. Would love to share with some one who will write an interesting analysis for me to read. :) #
  • I feel like I could almost make a sentiment analysis map re:AFL footy. Just would take a bunch of days and don't have enough tweets. #
  • Rather than do what I'm supposed to do, I'm making a sentiment analysis map of AFL related tweets. #
  • Is it the nature of sport to have more happy words than sad words? #
  • AFL fan community sentiment: Are fans happy or sad? #

Powered by Twitter Tools

Related Posts:

AFL fan community sentiment: Are fans happy or sad?

Posted by Laura on Sunday, 25 July, 2010

I generally don’t believe in sentiment analysis.  I tend to think it is junk because so much of automated sentiment analysis misses contextual clues, relying too heavily on keywords.  Sentiment analysis also tends to give sentiment to bot generated posts and to posts that are neutral.  Despite this, I thought it might be a bit interesting to try to do a sentiment analysis of the AFL fan community on Twitter.  Are they happy or are they sad?  And later, where are the happy fans and where are the sad fans?

The first step in doing that was to collect a whole bunch of AFL related tweets using searchtastic.  I’m currently up around 3,400 tweets. (When/If I do the geolocation version, I’ll provide the raw data set.)  The second step was to develop a list of AFL specific sentiment related words.  In my case, I’m just going with the characteristic of happy and sad to make this easier.  My sentiment keyword list is as follows:

Happy Sad
Best Worst
Win Spoon
happy lose
excited sucks
smile awful
star sackermanis
brownlo fired
medal suspended
victory fouled
club song sad
pride upset
won fail
congrats heartbreak
Purchased lost
lucky Sold
success Desperate
champions blow it
Fit blew it
Riewolt damn
glad avoid
legend unlucky
star failed
brilliant Akermanis
cruised slammed
hope injured

The next step was to give a happy or sad label for tweets that included these terms. The last step was to count up how many Tweets were labeled Happy, Happy / Sad, Sad, No Sentiment.  I generated the following table:

Tweets % % – None
Happy 704 21% 55%
Happy / Sad 147 4% 11%
Sad 436 13% 34%
No Sentiment 2128 62%
Total 3415
Total 1287

Later, I’m hoping to match this sentiment with the geographic location of the tweets to find out where the pockets of AFL happiness are versus the pockets of AFL sadness are.

Related Posts:

Formal PhD Proposal draft

Posted by Laura on Sunday, 25 July, 2010

This is a draft of my formal PhD proposal.  Any feedback anyone has would be appreciated.  Any grammatical errors that you spot that need to be fixed?  Please point them out.

AUSTRALIAN FOOTBALL LEAGUE ONLINE: a demographic, geographic and social examination of the  AFL fan community online
The online ecosystem is changing the definition of sport fandom, how teams engage with their fans, and the potential demographic and geographic reach of sport fandom.  These three components or sport fandom are inextricably linked and are worth studying to understand what sport fandom will look like in the future.

There are currently four operational definitions of sport fandom.  Each definition originates from and is used by a different group.

Popular culture studies scholars like Jenkins (2006) offer the first definition.  They define fans as an active population who engage in activities related to an object produced by the larger popular culture.  This production includes activities like writing fan fiction, creating costumes, producing fanvids and organizing conventions.  Popular culture academics also define fans as posessing a sense of ownership of their product that is removed and distinct from the official one, that they view actors, athletes, copyrighted and trademarked materials as communally owned by fans.

Sports marketers and managers, who broadly speaking, define sports fandom around potential for spectatorship. Stewart (1983), Shilbury, D., Quick, S., & Westerbeck, H. (2003) and Sullivan (2004) talk about various aspects of this, including the goal of teams to sell tickets to matches.  They sometimes diverge from this definition to talk about spectator related behaviors that can be monetized including merchandise sales and television viewership.

Sociologists and historians offer the third definition. This group tends to define sports fandom as a form of identity and as a product of a specific culture.  Cashman (2002) looks at sport fandom in Australia as an extension of a wider Australian identity. Collins (2005) looked at sport as a component of people’s identity as it pertains to the rest of the world: Australian and inward looking or international and outward looking.

The final definition of sport fandom is provided by sport fans themselves and the media.  They define sport fandom around allegiances and in the moment activities that demonstrate these allegiances.

None of these definitions particularly work in the increasingly online-based world of Australian sport.  Fans can more easily be monetized by teams that does not rely on getting people into the stadium.  Identity continues to play a role in fandom but this is changing as the Internet allows teams to draw a more interstate and international audience.  In the moment activities can often extend out years as fans maintain large fansites and become more actively involved in organizations dedicated to but independent of their team.  How fans express allegiance also has changed.  No longer is it based on club membership and being kitted out in a club’s jumper or scarf.  Instead, fans can and do express allegiance by following their clubs on social networks, checking in on geolocation based social networks, creating message boards and fan pages, attending events organized on social networks in order to meet their fellow fans, and creating content related to their clubs to distribute across various networks like Facebook, LinkedIn, LiveJournal, Twitter, and YouTube.

The definitional change means that more fans than ever could be counted as barrackers for a club could.  This could potentially mean that a club has a different demographic population than the one historically associated with it.  This is because the demographic characteristics of the Australian community are sometimes at odds with populations described by sport historians and sociologists. The potential demographic change necessitates a benchmarking of a club’s population to help future sport historians understand the community as it exists in the period of 2010 and 2011.

Research Topic

I propose to document the demographic, geographic and social characteristics of the AFL fan community online.  This should help several groups including sport historians, sociologists and marketers better understand who fans are, and how current characteristics differ from the past.  It should also help these groups by providing benchmarks for other work done using these populations in terms of making certain groups they study are representative of the larger population.  Hopefully, this topic will help sport organizations make more informed decisions regarding which networks to market on, and help them better cater to specific segments of their fans.

Accordingly, the working title for this research is: ‘AUSTRALIAN FOOTBALL LEAGUE ONLINE: a demographic, geographic and social examination of the AFL fan community online.’

Research Questions
The proposed research will focus on the following questions:

  •  What is the best methodology for conducting an online population study?
  •  What are sport fans in an online context?
  •  What are the demographic, geographic and social characteristics of the AFL community online?
  •  Why does the online population change over time?
  •  What influence do stakeholders have in shaping the demographic and geographic characteristics of the fan community dedicated to their team?
  •  How does the AFL population compare to other Australian based sport leagues?

Relationship to Previous Research

This topic is grounded in sport sociology, sport history and sport culture.  These three areas are complimented by a framework of popular culture studies, sociology, history and other areas in social sciences.

Much of previous research involving AFL fans has focused on offline populations, and was conducted using survey research, observational work or historical work.  This is the case for Stewart (1983) where the methodology was based on around club history and observations of match attendance.  A population study done by the marketing agency Roy Morgan Research  (2009) relied on a telephone survey.  There does not appear to have been a large scale study involving documenting the actual characteristics of fans that measures the actual composition of the fan community beyond club membership or match attendance.

Most social media research uses one or more of ten methodologies identified by me.  These research types are:
1. Individual case studies for how a business uses social media and the web;
2. Search and traffic analytics analysis;
3. Sentiment analysis and reputation management;
4. Content analysis;
5. Usability studies;
6. Interaction and collaboration analysis;
7. Relationship analysis to try to determine how people interact and to identify key influencers;
8. Population studies;
9. Online target analysis of behavior and psychographics; and
10. Predictive analysis.
Some of these research methods have been used for analyzing online groups, behavior and content.  The most popular methods include case studies, content analysis, usability studies, influencer identification, reputation management, and interaction and collaboration analysis.  Based on my preliminary research, the last three are ones that the least likely to be done.

When population studies are done, they tend to short, do not detail methodology, focus on one particular site on the Internet and do not compare different populations.  This appears to be because there are few automated tools that allow for measuring population characteristics of several sites at once.  Of the tools that exist, most are focused on providing information related to other methodologies including interaction and collaboration analysis, sentiment analysis, and search and traffic analytics analysis. These tools generally do not provide demographic and geographic population related data.  Those that do offer demographic information tend to focus only on one site such as Twitter or Facebook.

The existing methods and the reliance on automating data collection around a single site acts as an intellectual and practical barrier in doing large-scale population studies across multiple sites.  My research will help provide a methodological framework for doing a population study online that compares populations across different networks and subgroups as this is largely lacking.  I believe this will also be useful in terms of laying a framework for people who want to further explore the methodology for online target analysis, psychographics and predictive analysis as it pertains to demographic and geographic characteristics.

The methodology for this research involves two parts: A population study and a series of interview with sport administrators.

The first part involves conducting a population study of the AFL fan community online.  Some of the social networks examined will include 43things, bebo, BlackPlanet, blogger, care2, ecademy, Facebook, friendster, Gaia Online, hi5, LinkedIn, LiveJournal and its clones, MySpace, orkut, Twitter, Wikia, Wikipedia, Yahoo!Groups and YouTube.  Publicly accessible data regarding gender, age, income, occupation, astrological sign, joined, last logged in, location, dating status, languages understood, etc.  This information will be complemented with social behaviors in the context of events that take place during the AFL season.  This will include the occasional monitoring of the total people belonging to a group or listing specific clubs as an interest.  It will also include gathering some data regarding total actions taken during certain time periods, volume of web traffic as measured by tools such as Compete, Alexa and Quantcast, and total blog posts as measured by tools like IceRocket.

The second part involves interviewing sport administrators.  This includes people working for the AFL and for people from other clubs and leagues including the Canberra Raiders, Canberra United and Melbourne Victory.  The interviews will have a general framework of asking four questions:
1. How do you define fandom?
2. How do you reach out to the fan community?
3. How much influence does league management provide in terms of defining fandom and how to engage in outreach?  Do they give guidance on social media policies?
4. How much do other sports, teams and leagues play a roll in development and implementation of concepts related to fan engagement and social media?
These questions will be supplemented with other questions based on answers given.  The interviews will take place in person to facilitate the ability to ask related questions.

The results of the population study will be contextualized in the thesis conclusion to help provide context for why specific communities have the demographic, geographic and social characteristics that they do.

Intended Outcomes

At the conclusion of this research, an established, repeatable methodology for conducting population studies online should be established.  Shortcomings for this type of research will also be established, providing context for the data within and for others who are trying to do similar research. Maps of where AFL fans are located will exist.  These maps will show population distribution by website and by team.  There will be benchmarks for community size that can be used by future research to examine growth patterns and population shifts.

This information will be disseminated over the Internet with copies being shared with the AFL and its clubs.

The timetable for this research will be based around the AFL season, with a majority of the data collection being done right at the beginning of the season, during the season and shortly after the season ends.  During the off season for the AFL, numbers should be stable as there is less club news and the teams will be less active in promoting their games.   Appendix A provides a more detailed breakdown of the proposed timeline.  This timeline assumes that most of the work will take place over two years, though if a possibility exists of compressing it down to one, that will be explored. Data has and will continue to be gathered from between April 2010 and October 2010, the end of the AFL season.  During the data collection periods, work will also be done on completing the literature review, methodology, introduction and getting other administrative tasks like ethics clearances to be completed.  Interviews with sport organizations will take place during the data collection period and while completing the data analysis and conclusion.

The majority of the research would be computer based and involve data mining of online resources.  Some assistance may be required in developing tools to automate data mining of publicly available information and running the applications developed to do that work.  There is an intention to supplement knowledge gained through data mining with interviews. Travel may be required to interview management and staff involved with developing a team’s online presence.  The intention would be to take a trip to Sydney, Melbourne, Adelaide and Brisbane to conduct these interviews.  The attempt would be made to compress as many interviews as possible into the fewest trips possible. Ideally, the department would assist in funding travel to talk to teams.  Funding for travel costs would also be sought from the AFL.

Approvals and Permits for Ethical Research
Approval would need to be attained from the UC Committee for Ethics in Human Research in order to conduct interviews with sport managers.

Reasons for Choosing the University of Canberra
I chose the University of Canberra after having visited the Australian Institute of Sport in July 2009 and learning how the Australian system for sport worked.  The University of Canberra’s Sport Studies has close contact with the Australian Institute of Sport, sporting federations and with the professional sports leagues in the country.  These resources are important as they allow access to sports administrators who can provide context to social, demographic and geographic patterns that research will reveal by explaining a team’s role in the online fan community.

Cashman, R. (2002). Sport in the National Imagination, Australian sport in the Federation decades. Sydney: Walla Walla Press.
Collins, T. (2005). “‘One Common Code of Football for Australia!’: The Australian Rules and Rugby League Merger Proposal of 1933.” In R. Hess, M. Nicholson, & B. Stewart (Eds.), Football Fever: Crossing Boundaries (pp. 27-38). Hawthorn: Maribyrnong Press.
Jenkins, H. (2007a) Fans, Bloggers, and Gamers: Exploring Participatory Culture. New York: New York University Press.
Jenkins, H. (2007b) “Afterword: The Future of Fandom”. In Fandom: Identities and Communities in a Mediated World. J. Gray, C. Sandvoss and C.L. Harrington, eds. pp. 357-364. New York: New York University Press.
Jenkins, H. (2006) Convergence Culture: Where Old and New Media Collide. New York: New York University Press.
Roy Morgan Research. (2009, July 19). More Than 7.6 Million AFL Supporters A Great Market for Prospective Sponsors 2008 Grand Finalists Hawthorn (Up 72,000) & Geelong (Up 99,000) Have Biggest Jumps in Support. Roy Morgan Research. Press Release. Retrieved May 31, 2010, from
Shilbury, D., Quick, S., & Westerbeck, H. (2003). Strategic Sport Marketing (2nd ed.). Crows Nest, New South Wales: Allen & Unwin.
Stewart, B. (1983). The Australian Football Business, a Spectator’s Guide to the VFL. Maryborough, Victoria: Kangaroo Press.
Sullivan, M. (2004). “Sports Marketing.” In J. Beech & S. Chadwick (Eds.), The Business of Sport Management (pp. 128-153). Essex: Pearson Education Limited.

Related Posts:

  • No Related Posts