Historia Ludens

The Playing Historian

Alexander von Lünen, Katherine J. Lewis, Benjamin Litherland, Pat Cullum, Alexander von Lünen, Katherine J. Lewis, Benjamin Litherland, Pat Cullum

• Publisher: Routledge
• Year: 2019
• ISBN:
  9781000693317
• Edition: 1
• Topic: History
• Subtopic: World History
• Index: History

This book aims to further a debate about aspects of "playing" and "gaming" in connection with history. Reaching out to academics, professionals and students alike, it pursues a dedicated interdisciplinary approach. Rather than only focusing on how professionals could learn from academics in history, the book also ponders the question of what academics can learn from gaming and playing for their own practice, such as gamification for teaching, or using "play" as a paradigm for novel approaches into historical scholarship. "Playing" and "gaming" are thus understood as a broad cultural phenomenon that cross-pollinates the theory and practice of history and gaming alike.

Part 1
History of Gaming

1 A Quantitative Study of Historical Video Games (1981–2015)

Yannick Rochat

Introduction

Human history is often used as a narrative background in video games, whether game developers want to share their views on the past or need a context to host an assemblage of interactive elements and game mechanics. Some of these historical video games, ‘those digital games that make meaning out of the past’ (Chapman 2016a), are amongst the most popular video games ever created, e.g. The Oregon Trail (MECC, 1975), the Civilization series (MicroProse, Infogrames, 2K Games, 1991–2018), Age of Empires I and II (Microsoft, 1997, 1999), and the Assassin’s Creed franchise (Ubisoft, 2007–2018). In this chapter, I provide a survey of a large number of historical video games released between 1981 and 2015. For this purpose, I use as proxy a database composed of 1,690 video games released in that interval.¹ It was created—and is still maintained—by an online community called HistoriaGames.² I discuss in this chapter the potential biases of working with a database built externally. Eventually, I provide a descriptive analysis of this set of historical video games with a focus on the distribution of video games over release dates, platforms, genres, historical periods, and historical events.

A Crowdsourced Effort

Librarians, archivists, and scientists are preserving and documenting video game heritage, most of the time with limited means. To pursue this task, they have to face the commercial and technical obsolescence of video game hardware and software.³ Therefore, information on older video games is usually hard to find today in public archives and libraries, whether one is looking for issues of old magazines or willing to borrow a video game on its original cartridge in order to experience it in the technical conditions of the time. More often than not when studying video game heritage, scientists have to turn to the work of online communities, for example, to study a platform (Altice 2015; Ippolito 2016), to study the history of video games (e.g. the lists and categorisations of video games on Wikipedia⁴ or thematic databases like the initiative of the FEMICOM Museum⁵), to study video games through paratexts (see e.g. the digitised versions of computer and video game magazines on the Internet Archive⁶ and on the Abandonware Magazine community website⁷), or to get an overview on the video game industry (e.g. MobyGames,⁸ a crowdsourced database referencing 121,923 video games as of 30 September 2018⁹). Amateur (game) historians started early to work on the preservation of video game heritage. We can expect in the future a collaboration with scientists on these resources. Nevertheless, today we strongly depend on such crowdsourced efforts. We have to take them into account when we study large sets of video games.

Data Gathering

Building a database requires identifying a category of objects—here, video games—and then recording information on these—that is to say, collecting metadata. Later in the process, requests are formulated in order to isolate subsets corresponding to given criteria. In the case of a database compiled for a given study, these requests must be thought of in advance so that they are relevant to confirm or contradict hypotheses. Hopefully, that process would happen before the gathering of data even begins. However, this was not possible in the context of this study. Therefore, I describe the origins of the database and how it is maintained.

Video game databases compiled by online communities often follow self-imposed rules, such as criteria (hopefully well-defined) that help the community decide which video games must be collected and how to organise metadata (what variables to collect, what levels to choose, and optionally how to build the data model). These public databases let both researchers and a general audience access their content easily via a web interface or an application programming interface (API). They usually invite the visitor to take part in the community and contribute.

In the present context, where the author of this research is not the originator of the database, we need first to understand how the database was compiled in order to determine a perimeter for the study, that is to say: what are the questions the database can (help to) answer, and what are its blind spots. The most important potential drawbacks in such a case are the lack of control on the criteria used to include a video game in the database and the choice of variables and levels. I discuss later the concepts of ‘video game genres’ and ‘historical periods’, especially how the levels (the values they can take) were determined by the owners of the database.¹⁰

Distant Playing

With a sufficiently large and dense database, we can (perhaps) provide answers to theoretical questions by analysing subsets of its elements. In the case, for example, of a corpus of about a hundred games, it is hardly conceivable to play all of them. Not playing some or most of them is the most realistic option for the researcher here, thus reducing on one side the scope of an analysis conducted with respect to game content but increasing on the other side the range of results based on metadata. In the field of digital humanities, this situation has led to the distant reading approach (Moretti 2013; Ganascia 2015; Underwood 2017), where questions like ‘how are the genders of characters distributed over time’ (Underwood, Hamman, and Lee 2018) are answered based on corpora of sizes much larger than what a scholar is capable of reading during his or her whole life.

There are few works in game studies dealing with databases in order to examine large sets of video games. For a study that surveyed first-person shooter (FPS) games released between 1991 and 2009, Michael Hitchens (2011) compiled a database of 566 video games (of which ‘63 games were played in whole’). In his analysis, he reviewed the platforms on which these video games were released, the distribution of FPS games over time, and the gender distribution of the avatars. The database was compiled with the help of a variety of online resources. In a paper published recently, Jérémie Pelletier-Gagnon (2018) compiled ‘2053 JRPG reviews gathered from 10 different online journalistic outlets posted between 1992 and 2014’ in order to analyse the emergence and use of the Japanese role-playing games (JRPG) label. For this purpose, he used topic modelling methods. Eventually, in a study on video games representing the First World War, Adam Chapman isolated 58 video games (Chapman 2016b) to ‘[explore] the relation of WWI popular collective memory to historical video games’.

Whether it is a database of video games or a database of texts about video games, these initiatives pave the way to studies on video games allowing researchers to uncover and validate results at the scale of a whole genre or platform—that is to say distant playing.¹¹

A Database of Historical Video games

I use the HistoriaGames database as a proxy for the analysis of historical video games, with potential biases. Notice first that I will not attempt to define what a ‘historical video game’ is. Most aspects of this question have already been discussed with success by Adam Chapman (2016a), among others. Here, we rely on a database and a harvesting process that we must understand first. The HistoriaGames database was used in studies on historical video games in French language recently (Facchini 2016; Natale 2017; Maillard 2018; Marti 2018). The database is large, well-documented, and follows the trends in numbers of MobyGames, one of the largest databases of video games.¹² The proportion of historical video games to the number of video games released in total is constant over time, based on data from MobyGames and HistoriaGames. However, there are at least two possible biases here. First, in recent years, MobyGames included video games for mobile platforms outside of mobile video game consoles, while HistoriaGames does not. Second, since HistoriaGames was built in part on data of MobyGames, it might have inherited its biases, that is to say: recent video games, video games on popular platforms, and video games developed or released in North America, Europe, and to some extent Japan, have a higher probability to be listed.

Among the criteria used to compile the database, the originators decided not to attempt to identify all video games released for smartphones and tablets, even if they met their criteria. Therefore, for the sake of consistency, I decided to exclude from the database all the versions of games available on iOS or Android. This is the case, for example, with Assassin’s Creed: Altair’s Chronicles (Ubisoft, 2008), which was released for Nintendo DS, iOS, and Android: only the Nintendo DS version was kept in this study.

The website HistoriaGames was displaying 1,783 titles at the very beginning of 2016, when a dump of the database for this research was obtained.¹³ A video game or a video game extension¹⁴ released on multiple platforms was considered as a single entry and the remake of a video game was considered as a video game in its own right (e.g. Sid Meier’s Pirates!, released in 1987 by MicroProse and its remake released in 2004 by Atari). The oldest video games in the HistoriaGames database date from 1981.¹⁵ The most recent video games at the time of gathering data were video games announced for 2016 and video games already available in early access, with no planned release date. They were excluded from this study.

Based on these criteria, 1,452 video games and 238 video game extensions released between 1981 and 2015 were extracted from the HistoriaGames database. Collector’s edition or gold edition reissues, considered too similar to the basic versions, are not included. However, remakes are included, since they often contain original work and revive the interest for the video game, thus reaching a new audience. Each release date in the database corresponds to the earliest release date among all the platforms on which a video game is available.

The video games compiled in this database were released on 66 different platforms. By including all releases, this equates to 2,466 different versions of video games and 312 different versions of video game extensions. Thus, on average, a video game comes out on 1.7 different supports, with a maximum of 9 different supports for Wolfenstein 3D (Apogee Software, 1992).¹⁶ Without surprise, the most frequent platform is the family of Windows operating systems (1,075 video games and 225 video game extensions released between 1989 and 2015), followed by the Mac operating systems¹⁷ (193 video games and 45 video game extensions released between 1985 and 2015) and the iterations of MS-DOS (140 video games and 7 video game extensions released between 1981 and 1998). The console with the most historical video games in the database is the Xbox 360 from Microsoft, with 117 video games and 12 video game extensions. Numbers for video games released on computers are difficult to compare with numbers on gaming consoles due to the dependency of these on locked hardware, and thus shorter operating cycles. Therefore, in this analysis, the platforms are organised in three classes: computer OSes (‘computers’),¹⁸ home video game consoles (‘home consoles’),¹⁹ and handheld video game consoles (‘mobile consoles’).²⁰ The distribution of platforms into these three classes can be consulted online.²¹ This sorting choice allows comparison of the various groups of platforms.

There are 341 publishers identified in the database. Among these, Matrix Games, a company specialised in war games, released 102 video games;²² followed by Ubisoft, developer and editor of the Assassin’s Creed franchise with 100 video games, and Paradox Interactive, editors of the Europa Universalis, Crusader Kings, and Hearts of Iron series, with 99 video games. As for the development studios, Koei is the most prolific with 50 v...