This post has lots of graphs, but no shocking conclusions.

Oct 13, 2009 19:24

I. Methodology
Data was collected by gapingheadwound from The Inglourious Basterds Kink Meme, 100_scalps, and un_amico (the Fredrick/Shosanna comm). Only FPF was surveyed, not RPF. I did not collect data from basterds, iheartcinemaah, or operation_kino, on the basis that most of the FPF posted to any of those communities is usually reposted to 100_scalps.

I collected information on authors, pairings, posting dates, and genres (m/m, m/f, gen, theoretically f/f, although I have yet to see any) for all stories posted where I was collecting data.

Each part of a multi-part story was considered one fic. This is primarily due to my focus of examining fanfiction trends over time.

Multiple stories posted to a single entry (example: nkrathbone's prompt posts) were counted as multiple stories.

Stories involving threesomes (example A/B/C) were classified as containing three pairings A/B, B/C, and C/A for the purposes of organizing pairing popularity trends, even though a story labeled "A/B, B/C, C/A" would likely have very different dynamics than one labeled "A/B/C".

Further methodology information will be included under each section.

No hypotheses were made; information was collected primarily to look at the pretty graphs. Graphs were made with OpenOffice.org Calc.

II. An overview of fandom traffic

Are we posting more or less fic than we were to start off with? Here you go:

Fig. 2.1 - Number of stories posted per week


Fig 2.2 - Number of authors posting per week


Fig 2.3 - Average number of stories posted per author per week


So although we are seeing a slow increase in the number of stories posted after a drop-off early on, we are also seeing a noticeable increase in authors who post very frequently: almost 2 stories a week on average compared to just over 1 on average before the peak.

I also kept track of how many authors posted a story for the first time each week. How many "new" people are coming out of the woodwork to post? It seems a healthier fandom would have more people posting for the first time in addition to regular writers we already know and love.

Fig 2.4 - Authors posting for the first time per week


While this information is good in that it means we are seeing more people becoming active in the fandom every week (and that that number is not slowly diminishing yet), it is bad in that the fanfiction output has remained somewhat steady over the past few weeks. Although lots of new people wrote fic this past week, many writers who were writing the first few weeks did not post anything at all. Hopefully this means everyone's off writing long-fic and not that people get sick of Inglourious Basterds after a week or two. (Even though the sad option is probably the trufax one, I AM BEING OPTIMISTIC.)

III. Trends in Fic-type diversity (Gen/het/slash, pairing representation)

First, let's look at what I termed "pairing diversity." That is, how many different pairings are represented in each week's batch of stories. I used the formula
Pairings Represented divided by total non-gen fic
or
PR/(TF-gen)

The closer to 0, the smaller selection of pairings written about that week, the closer to 1, the more pairings written about.

Fig 3.1 Pairing Diversity


The general trend in the fandom is toward a less diverse selection of pairings in fanfiction in IB fandom.

Now, let's look at the distribution of fic among the genres of gen, m/m, and m/f. (f/f was not included as none has been written and posted. m/m/f threesomes have been included in both the m/m and m/f categories.)

Fig 3.2 Genre distribution



Shocking conclusion: Fangirls like m/m fanfiction. Fangirls also have OTPs. Even QT fangirls.

IV. Data on trends in specific pairings

Fig 4.1 Popularity of all pairings appearing in fic as of Oct. 12.


Fig 4.2 Popularity of het pairings appearing in fic as of Oct. 12.


Here, have a table full of the raw numbers. This is the number of stories posted to either the Kink Meme, 100_scalps, or un_amico for each pairing:

1. Donowitz/Utivich 39
2. Aldo/Donny 16
3. gen (noromo) 14
4. Fredrick/Shosanna 9
5. Aldo/Stiglitz 6
T-6. Landa/Von Hammersmark 5
T-6. Aldo/Landa 5
T-8. Aldo/Utivich 4
T-8. Hicox/Stiglitz 4
T-10. Hellstrom/Stiglitz 3, Landa/Hellstrom 3, Donowitz/Stiglitz 3, Landa/Utivich 3, Landa/Shosanna 3
T-15. Stiglitz/Utivich 2, Hicox/Von Hammersmark 2, Omar/Donny 2, Donowitz/Von Hammersmark 2, Aldo/Von Hammersmark 2, Donowitz/Shosanna 2
T-21. Wicki/Utivich, Utivich/Omar, Hellstrom/Zoller, Landa/OFC, Landa/Stiglitz, Landa/Hermann, Landa/Hicox, Stiglitz/Von Hammersmark, Shosanna/Marcel, Donowitz/Wicki, Donowitz/Zoller, Hellstrom/Wicki, Donowitz/Landa, Landa/Zoller, Hellstrom/Hicox, Aldo/Wicki, Utivich/all the basterds

Fig 4.3 Characters we like to ship, a graph


I like to call this one "No wonder Eli Roth was so delighted to find the kink meme." I also like to call this one "shhh. pretend I can spell Deiter."

The numbers aren't really big enough to track pairing popularity over time for anything except Donowitz/Utivich. So go look at that graph instead. Because it's funny.

eli roth hypotheses, science!

Previous post Next post
Up