Disclaimer: the data are incomplete. This post is a midway tuning of my method and its results.

Introduction

My project On the Shoulders of Cloud Giants (I changed the name) is analysing the citation practices within the tabletop role-playing game publishing industry. For this purpose, :

• I selected and analysed TTRPG;
• I designed a database of citations within Wikidata and queried by SPARQL; 
• I started analyzing the data with R and R Studio.

Selection

The TTRPG selection is based on:

• All the TTRPG that have a Wikipedia page (estimated coverage: 95%) ;
• Plus all the TTRPG i heard about in my periodicals the last 10 years (blogs, podcasts, magazines) (estimated coverage: 80%) ; 
• Plus all the TTRPG cited by a TTRPG selected before.

The selection is moderately biased because the notoriety is externally attributed by a third party (Wikipedia contributors and periodical editors). Each TTRPG selected is validated by the fact, they all have a reference in the database RPGGeek (almost 99.9%).

Counting & Recording

Each mention of citation was saved as a PNG file nameofgame-pXX.png (where XX is the page the snapshot was taken).
I also recorded :

• the colophons, the credit pages, the acknowledgement page, the forewords, the afterwords, etc.;
• the bibliographies, the filmographies, the mentions of artworks;
• and epigraphs (all or some of them).

Database

I choose Wikidata, because it was simple, it was open and it was fulfilling most of my needs. Wikidata can be queried by R Studio. The properties used were:

• cites work P2860
• publisher P123
• movement P135
• author P50
• publication date P577
• RPGGeek ID P7226

Unfortunately, there was no property for : has a bibliography and has epigraph

Midway assessment

• May-december 2019 : 13 000 edits.
• 300 items created.
• 800+ links created between 140+ TTRPG.
  

SPARQL 

SPARQL is a language to query Wikidata and retrieve data. The main query is here.

R Studio Analysis

R and R Studio are very famous in the university around me. Its open, its quite simple to start, there is a lot of community support, its fashionable… so why not using it to process my data ?
I deposited the code of my R Studio project here on GitHub https://github.com/pmartinolli/OtSoCG

To have a look on a very drafty and preliminary graph displaying the data, please click on the picture below: