Video Game Development as Public History: Practical Reflections on Making a Video Game for Historical Public Engagement

As the separate fields of digital public history and historical game studies have continued to expand in recent years, attention has increasingly been directed to the potential of video games as a method of public history. Scholars from both fields have found that the video game format is uniquely suited to fulfilling public history’s goal of sharing authority between history professionals and their publics in collaborative and co-creative ways. Public historians have been particularly drawn to the capacity of video games to cede narrative control to players through the agency of meaningful choices and to create spaces of historical co-creation via online forums and modding communities, while historical game scholars emphasize their value for simulating complex historical processes. However, despite an ever-increasing flow of scholarship on analyzing and teaching with commercial video games, historians themselves have been slow to adopt game development as a method of practicing public history.¹ This reluctance is in contrast to disciplines such as public health and communications studies, where indie scholarly video games are a norm for public engagement and knowledge mobilization, and there are countless systematic literature reviews on best game development practices. Historians are at risk of falling dismally behind the times.²

In this paper, I offer insight into the practicalities of producing an indie video game as public history by reflecting on my experience of designing and developing Ab Uno Sanguine, a game based on my PhD research on nineteenth-century British imperial humanitarianism.³ I write this paper as someone with zero dedicated funding and zero game design experience to demonstrate how historians can combine the wealth of available scholarship on game design with modern indie game development tools to harness the video game medium as a method of public history. I also write from a British perspective, where “doing public history has become one of the popular means by which university humanities scholars can demonstrate community engagement,” and I indeed designed Ab Uno Sanguine as a method of public engagement to contribute toward the impact requirements of my dissertation.⁴ As such, the following discussion is principally directed toward researchers considering public history video games as an accessible means of public engagement. Nevertheless, this paper is for all public history practitioners curious about the practicalities of game production, whether that be graduate students planning an impact statement for their thesis, faculty planning a knowledge mobilization scheme on their grant proposal, or museum staff planning a new public history project.

Of course, developing a video game requires a certain level of technical skills that not all public historians will feel they possess, yet the “ludic turn” and the rise of the “gamic mode of history” have make it essential for historians to understand how history games are made.⁵ Moreover, with the increasingly digital nature of public life, especially since the mass movement online following COVID-19 lockdowns in 2020–21, historians are in dire need of effective ways to engage our publics in virtual spaces. My discussion of the development of Ab Uno Sanguine is not the first to explore this issue. Kevin Kee and John Bachynski’s analysis of making Outbreak and Laura Zucconi et al.’s examination of making Pox and the City are aimed towards helping future historian-developers in designing their own games.⁶ Theoretical frameworks have also been proposed to provide best practices and design principles for the production of historical games, particularly those by Keven Kee et al., Julian Bazile, and Dawn Spring.⁷ Chih-Chieh Yang has even set forth a heuristic process for historical game design to provide historians with a game development roadmap.⁸ All of this scholarship has generated valuable insight into many aspects of historical game design, and I incorporate elements from each of these works throughout this paper. However, although these historians have dedicated themselves to establishing historical game design as a fixture of the discipline, none of them provide any insight into the practical aspects of development, such as level structure, coding, asset creation, or scaffolded delivery. This paper, therefore, focuses on these practical considerations. Where reasonable, I expand on my technical process so that those inclined can follow along, and I also point toward less technically demanding options as well as resources for those who would like to begin learning themselves.

This paper progresses through four sections. First, I describe my game, Ab Uno Sanguine, as it exists at the time of writing this reflection piece, including its underlying scholarly evidence base and arguments, gameplay narrative and mechanics, and technical foundations. Second, I situate the choice of video games as a medium for public history within the growing public history scholarship on games, shared authority, and digital co-creation, and then survey scholarship from game studies and historical game studies to provide an overview of the theoretical foundations behind planning a historical video game. Third, I utilize scholarship on the development of serious games to walk through my decision-making process during the entire ADDIE production model: Analysis, Design, Development, Implementation, and Evaluation.⁹ Finally, I reflect on two key takeaways from my experience of developing the first version of Ab Uno Sanguine—the importance of limiting asset requirements, and the challenges of game impact evaluation—and point toward how they can be addressed in future game iterations.

In its current iteration, Ab Uno Sanguine is a management/simulation style game based in late nineteenth-century London, in which the player takes control of the British Aborigines’ Protection Society (APS). The Aborigines’ Protection Society operated between 1838 and 1909 as a political lobbying group, gathering intelligence on Indigenous rights abuses from a network of colonial informants and pushing for change by raising public awareness and petitioning the British government.¹⁰ My dissertation, based out of University College London (UCL) and funded by UCL and the Social Sciences and Humanities Council of Canada, uses letters written to the Society by settlers, Indigenous peoples, and missionaries living in Canada, Australia, New Zealand, and South Africa between 1860 and 1890 to argue that the connections forged between colonial subjects and the Society reveal some of the contingencies of settler colonialism on the limitations of imperial humanitarian networks. For example, I showed that settler acts of violence towards Indigenous nations were able to go unchallenged by British authorities for reasons such as evidentiary requirements of official government communications, newspaper editors’ disinterest, and the incompatibility of political activism with missionary society fundraising campaigns. This last argument is what I focus on in Ab Uno Sanguine, demonstrating to players some of the financial and social pressures that limited the ability of late Victorian activists to effectively challenge Indigenous rights abuses.

As secretary of this Society, players have two tasks. First, they must review intelligence received in letters from the colonies and make a simple decision for each letter: to either accept the letter and take action in response to it or reject the letter and do nothing. Each decision has consequences for the Society’s finances: good decisions align with social expectations and attract supporters, while bad decisions transgress social expectations and repel supporters. The second task is to allocate the funds received between the Society’s primary expenses: office rental, publication fees, delegation costs, and member activities. Victory is achieved by preventing bankruptcy and maintaining the Society’s lobbying efforts after a set amount of time in the face of rising racism and declining concern about Indigenous peoples, but the nature of victory is dependent on player decisions.

The game is broken into four levels, with each level containing increasingly complicated social environments that the player must navigate to attract enough supporters to keep the Society solvent. In the first level, the player begins in the 1860s, a period in which the Society was often in poor financial and political standing. As such, the player must carefully read each letter to ensure that they do not accept any letters that contain petitions or requests for funding that the Society lacked the capacity to facilitate. In the second level, the player encounters criticism for supporting interference in foreign territories, so players must ensure that they only accept letters sent from British territories by comparing postage stamps to an atlas of the British Empire. In the third level, the Society has accumulated enough resources to support petitions and funding requests, but only if accompanied by valid evidence, requiring players to evaluate the evidentiary basis of each letter by cross-analysis of letters with newspaper clippings. And in the final level, the player encounters criticism for working with informants of questionable character, requiring players to ascertain informant credibility by searching historical records for evidence of informant immorality. These individual rules compound and evolve across each level so that it becomes increasingly difficult to identify which letters will attract supporters and which will repel them.

Between each level, players open the Society’s ledger to allocate funds and must decide to increase or decrease spending on rent, publication, delegations, and member activities. Each of these decisions contributes to maintaining four key markers of the Society’s health: reputation, public awareness, political influence, and member satisfaction. For example, upgrading publication from writing letters to the editor of The Times to printing a quarterly journal increases yearly expenses from eight pounds to seventy-five pounds in return for potentially boosting public awareness from 15 percent to 65 percent. There is a range of options for each expense category, but it is typically impossible to afford the best option for each category, thus requiring players to devise their own strategies to keep the Society operational. If any of the key markers falls to zero percent, it cannot be recovered, but the game continues. The game ends if all four key markers drop to zero percent or if the Society’s funds drop below zero pounds. When the game ends, players get different endings depending on their success at balancing the budget and how many key markers they succeeded in preserving.

Although the technological decision-making behind the production of Ab Uno Sanguine is discussed in detail below, it is useful to briefly outline here the game’s basic specifications and assets. Ab Uno Sanguine was made in Twine 2.3.16, using the SugarCube 2.36.1 story format. SugarCube is the “medium” difficulty Twine story format, allowing those with some programming knowledge to customize its functionality beyond that of the basic Harlowe format while being much more accessible to beginners than the advanced Snowman format. The game incorporates the following external libraries: the JavaScript library Interact.js 1.10.11, which enables players to click and drag objects in the game space; the CSS library Animate.css 4.1.1, which provides pre-built animations for manipulating game objects; and the Twine meter macro set made by Chapel 1.0.1, which provides pre-built progress bars for visualizing the player’s key marker levels.¹¹ Graphical assets were created using Adobe Illustrator CC 2015, and royalty-free audio assets were sourced from https://www.epidemicsound.com/. As for the historical documents encountered in the game, artifacts were sourced from a range of archives. Letters to the Aborigines’ Protection Society were consulted at the Bodleian Libraries at the University of Oxford.¹² Petitions to the British Colonial Office were consulted at the National Archives in London.¹³ And a wide range of digitized historical newspapers were consulted through Gale’s British Library Newspapers and 19^th Century UK Periodicals collections, the National Library of New Zealand’s Papers Past collections, the National Library of Australia’s Trove collections, the Bibliothèque et Archives nationales du Québec’s Numérique collections, and the University of Victoria’s British Colonist collections.¹⁴ All primary sources have been made available for players to explore at the companion website https://aps.darrenreid.ca/.

Of many debates over the future of public history, the discussion around shifting practice from something that is for or about the public to something that is with or by the public has been one of the most prominent.¹⁵ Popularized through Michael Frisch’s concept of “shared authority,” the move toward participatory and co-creative public history has generated extensive rumination on how public history practitioners can balance competing academic and public interests in history. These questions have been made more complex by the advent of Web 2.0 and its accompanying host of methods for our publics to participate in the construction of historical narratives.¹⁶ Within this context of calls to practice a digital public history that straddles the balance between academic rigor and public interest, video games have increasingly been looked to as a medium for participatory public engagement, with at least nine articles and chapters dedicated to the subject in public history handbooks and journals in the past five years.¹⁷

There are two primary ways in which video games afford shared authority between public history practitioners and their publics: the co-creation of interactive narratives through gameplay and the engagement in public historical discourses via discussions on game forums and social media. The co-creation of interactive narratives stems from the fact that game developers do not have the power to control how players will play their games. Similar to exploring a history website, which Sharon Leon classifies as participatory in that the “choices of what sections to visit and what to read, watch, or listen to determine the information, understanding, and questions that [users] take away from the experience,” players of a video game chose which characters to play, which dialogue options to select, which strategies to employ, which items to interact with, and ultimately which ending they will get.¹⁸ In their analysis of how the game Valiant Hearts: The Great War (2014) structured possible player choices around historical and archival material, Abbie Hartman, Rowan Tulloch, and Helen Young found that the game is “a particularly participatory example of public history” because it operates “like a museum that can be explored, [as] historical details and artefacts are spread throughout the gamespace and can be found (or ignored) at the player’s own pace, and reflected on, researched, and incorporated into their own narrative in a variety of ways.”¹⁹ Not only do games afford shared authority through allowing “players’ choices and input to change the outcome of the gameplay and with it the historical narratives,” but the replayability of games enables players to try and retry infinite combinations of choices and strategies, allowing them to experiment with developer-constructed arguments about historical causality, significance, and agency.²⁰

The second form of games as public history—what Hartman, Tulluch, and Young refer to as the creation of “affinity spaces”—takes place outside of actual gameplay and essentially entails players becoming public history practitioners themselves by sharing their interpretations of a game’s historical content with other players.²¹ Jeremiah McCall points to large and active affinity spaces that have developed around popular history games such as Total War: Rome 2 (2013), Civilization IV (2011), and Crusader Kings 2 (2013), while Jakub Šindelář points to the active community around Valiant Hearts (2014), each of which demonstrate strong public engagement in historical discussions.²² There are many different ways that players participate in affinity spaces, ranging from commenting on forums, to uploading Let’s Play videos on YouTube, to modding games with their own code.²³ Video games thus afford two different levels of sharing historical authority and engaging with our publics: first through gameplay itself and then later through affinity spaces where “the myriad of game focused review sites, blogs, forum and streamers…are a significant space for engagement from and among the public.”²⁴

Unfortunately, the creation of affinity spaces is largely beyond the control of game developers; they can only develop organically around games with comprehensible historical narratives that players enjoy playing enough to form communities around. The task of the historian-developer is, therefore, to focus on A) designing games that effectively incorporate the shared authority of player agency into the interactive historical narratives we wish to convey and B) ensuring that those historical narratives are clear and fun to engage with, in the hope that an affinity space may develop around them. To do this requires a strong grasp of how games convey historical narratives and how games can be made “fun.”

Understanding how games convey historical arguments and how games become fun requires delving into the theoretical underpinnings of game studies and historical games studies. Adam Chapman, one of the most well-known historical game theorists, identifies two approaches to gamic representations of history: realist simulations and conceptual simulations.²⁵ Neither approach is necessarily better than the other, but each facilitates a different style of historical engagement. Realistic simulations follow a Rankean agenda of presenting arguments about “how it actually was,” and are well-suited to facilitating historical consciousness of specific events and historical empathy with specific people.²⁶ This is how Sky LaRell Anderson, as well as Hartman, Tulloch, and Young, envision games—as “interactive museums,” where players learn about historical people, places, and things by interacting with digital versions of them.²⁷ Realist simulations try to mimic the real world as closely as possible and avoid counterfactuals as much as possible in order to provide historical facts and details.

Conversely, conceptual simulations often eschew details and facts in favor of presenting arguments about more abstract historical processes that embrace counterfactuals and are, therefore, more suited to facilitating historical thinking and reasoning about causation, complexity, and contingency.²⁸ Chapman’s understanding of conceptual simulations is based on Ian Bogost’s theory of procedural rhetoric, which holds that, within video games, “arguments are made not through the construction of words or images, but through the authorship of rules of behaviour.”²⁹ In other words, games teach you how to win or lose based on the rules developed by the game maker, and if a player wants to win a game programmed to simulate historical processes, the player needs to master the historical rules that structure the game.³⁰ An important implication of Chapman’s differentiation between realist and conceptual simulations is that although scholars like Sian Beavers argue that digital games are particularly effective at facilitating “historical perspective taking” (that is, historical empathy) and “cause and consequence” (historical reasoning), a single game generally cannot do both at the same time. Historian-developers must ensure that their adoption of either the realistic or the conceptual simulation style aligns with the primary objectives behind their games.³¹

As for ensuring that a history game is fun, this requires balancing challenges and rewards to achieve what game scholars refer to as the flow state. Flow refers to the state of mind that playing a game ideally instills in players. It is defined in many ways but generally consists of concentration on the task at hand, feeling rewarded for skillful behavior, and experiencing a loss of sense of time.³² It requires a careful balance between a game being so easy that it is boring and so hard that it is impossible. If the player cannot get into a flow state while playing a game due to either boredom or frustration, they will stop playing, and the game will have minimal public impact.

Making educational games fun has traditionally been difficult because scholars place too much emphasis on challenge (the educational part) and not enough emphasis on reward.³³ This is precisely what happened to Kee and Bachynski; they designed a historical video game to teach people about a plague outbreak in 1885 without considering the importance of flow and discovered that, in doing so, they created a game that no one wanted to play.³⁴ Some scholars question whether educational games should even try to be fun. For instance, McCall argues against the use of “fun” as a category of analysis for historical games in favor of terms like “engaging” or “immersive,” insisting that fun is not of pedagogical utility. In the context of using games in history classrooms, he argues, the purpose of the game is to facilitate learning, and “fun is simply beside the point.”³⁵ Some game scholars similarly contend that fun is less important for “serious games” since their primary purpose is to educate.³⁶ Yet, as will be discussed further in the Implement section below, public audiences are not captive audiences who can be forced to play an unfun game by an authoritative teacher or supervisor. Moreover, for most game scholars, the very definition of a game is something that is played freely or voluntarily, and Roger Callois famously asserted that play ceases to be play the moment it is forced.³⁷ If a game does not stimulate voluntary play through being fun, it may cease to be a game at all and veer into the territory of training exercises and educational simulations that few will click on and even fewer will play to completion.

Thankfully, when designing historical games, we have a wealth of inspiration for how to balance challenge and reward at our fingertips. Chih-Chieh Yang’s “heuristic process of historical game design” advocates for using exemplar games as inspiration for game mechanics that have proven to be both fun and educational, and historians have pointed to myriad successful games like Civilization, Crusader Kings, Valiant Hearts, and Assassin’s Creed that can be used as models.³⁸ We can also look at the games we enjoy playing ourselves and identify successful game mechanics that we can adapt for our purposes. We don’t have to have played all games imaginable to do this or even be a gamer at all. YouTube is overflowing with video playthroughs. By implementing game mechanics that are proven to be enjoyable, historian-developers can improve their challenge/reward balance, facilitate player transportation into the flow state, and thereby reduce the likelihood that their games will go unplayed.

Historians as well as game scholars have proposed various game production models to structure the development process, each with their strengths and weaknesses. Kee and Bachynski advocate for the “mechanics-dynamics-aesthetics” (MDA) framework, developed by games studies scholars, in which designers first determine the aesthetics (the emotional responses the game will evoke in players), then the dynamics (the overarching events within a game that will facilitate its aesthetics), and finally the mechanics (the fundamental actions within a game that will facilitate its dynamics).³⁹ The MDA model is useful for establishing a basic order of doing things (i.e., identifying mechanics to best achieve a pre-defined outcome rather than beginning with a mechanic and grasping for an outcome afterward) as well as for ensuring that each step in the game design process remains oriented around achieving a specified public history objective (that is, the aesthetic). Yet the MDA model is lacking from a public history perspective because it makes no space for our publics. For example, how do we ensure that the aesthetics of our games are aligned with the needs, perspectives, and interests of the publics we are engaging with, and how do we ensure that the dynamics of our games leave space for our publics’ shared authority in crafting interactive historical narratives? Mary Flanagan and Helen Nissenbaum propose their alternative “Value at Play” (VAP) model to better address public interests by changing the order of business from aesthetics > dynamics > mechanics to discovery > implementation > verification, where discovery entails identifying the needs of key stakeholders and incorporating societal input from the very beginning.⁴⁰ Chih-Chieh Yang further developed VAP into the following seven-step “heuristic process for historical game design”:

1. Evoke cultural awareness.

2. Explore possible challenges.

3. Analyze exemplar games.

4. Define solution spaces.

5. Facilitate concept ideation.

6. Establish evaluation criteria.

7. Interplay creative group.⁴¹

Yang’s heuristic model is much more useful for public history practitioners than Flanagan and Nissenbaum’s because of its emphasis on using the gamic mode to address social and cultural needs of our publics. Further, it helpfully breaks down the game planning process into the various questions and decisions that game makers need to address. However, the utility of this model for indie game developers is limited by Yang’s assumption that historians would be only acting as consultants on large interdisciplinary development teams. As a result, their model consists solely of conceptualization activities and completely ignores the crucial steps involved in actually building games.

More useful for indie game designers is the ADDIE model developed within the scholarship on serious games. Serious games, defined most basically as games which have “an explicit and carefully thought-out educational purpose and are not intended to be played primarily for amusement,” have become massively popular across disciplines such as education, marketing, and health care.⁴² Although typically rooted in education and game studies and disconnected from historiographical theories and methodologies, the scholarship on serious game design does contain various game design models that historian-developers can adapt to their specific purposes, with ADDIE being one of the most common models.⁴³ Under the ADDIE model, which itself was borrowed from Instructional Design scholarship, game producers Analyze the goals, audience, and theoretical basis of a game, Design the mechanics and rules of a game, Develop game assets and integrate them with the mechanics and rules to form a game prototype, Implement the game by releasing it to the intended audience, and Evaluate the game to assess how well it achieved the objectives. Because the ADDIE model is based around a workflow of basic tasks, it can largely be merged with Yang’s heuristic model to achieve the best of both worlds, with Yang’s steps 1, 2, and 7 falling under the Analyze task, steps 3–5 falling under the Design task, and step 6 falling under the Evaluate task. Yet it is the Develop and Implement tasks where the ADDIE model becomes most useful to indie game developers, and it was for this reason that I structured my game production process around ADDIE. For the remainder of this paper, I report and reflect on each phase of ADDIE in turn before highlighting my primary takeaways.

The first step in designing Ab Uno Sanguine was to identify its objectives and audience. In the context of designing a game to disseminate research findings for public engagement, this entailed identifying the various audiences of my research and determining how my research would be relevant, useful, or impactful for them. As David Dean demonstrates, there is never a single “public” that engages with history, but many “publics” with divergent interests and perspectives.⁴⁴ Exhibit designers have long understood that different publics like to be engaged in different ways and presented with different narratives, and thus typically provide multiple means of historical communication.⁴⁵ The same concept applies to public history games, which are sure to be played by people with varying personal connections to, and interest levels in, the content of the game. I, therefore, needed to identify which publics were most likely to encounter my game, which publics I would specifically target, and how I could address the needs and perspectives of each.

Using an interest-level approach to assessing my multiple publics, I identified three distinct groups of people who were most likely to play my game and who required different scales of engagement. The public with the highest level of interest are the direct stakeholders in the history of the Aborigines’ Protection Society: the descendants and communities of those who wrote letters to the Society as well as Indigenous rights activists who are the spiritual successors of those who supported the Society. At the medium level of interest are history enthusiasts—both amateur and professional—who do not have a direct connection to the Aborigines’ Protection Society, but enjoy learning about history. And at the lowest level of interest are gamers in general, people who simply want to play a fun game and have little interest in historical learning.

There are ways to build a game so that each of these three publics can engage in historical content in the ways that they prefer. For example, the game can be crafted for general gamers by designing an enjoyable base game narrative that conveys my historical arguments through gameplay alone without feeling overly “educational” or requiring an unreasonable amount of mental labor. For history enthusiasts, who are likely more desirous of an academic environment and more willing to do extra mental labor, an option can be added to the game to provide further historical context, discussion of the underlying argument, and references to further reading. And for direct stakeholders, who have a direct interest in the histories represented in the game, links can be provided to digitized versions of the letters and documents used to build the game, which can be useful for genealogical, legal, and activist purposes. By providing these three levels of opportunities for engagement, the game can further share authority with players, affording them the choice of how deeply they want to engage in historical discussion.

Nevertheless, while a game can afford different levels of participation to different publics, the MDA model emphasizes that design must be orientated around a specific aesthetic or objective, which requires identifying one target public to prioritize. According to the VAP model, identifying a target audience entails a process of “discovering” the potential relevance of one’s research to its various publics.⁴⁶ To do this, I was particularly inspired by the Antislavery Usable Past project, which aims to “translate the successes and failures of past antislavery movements into effective tools for contemporary policymakers, civil society, and the heritage community.”⁴⁷ Building on the Antislavery Usable Past’s connection between antislavery historical research and antislavery activists, I reflected upon how one of my stakeholders, modern Indigenous rights activists, might benefit from my research on the Aborigines’ Protection Society, and I landed upon the theme of developing historical reasoning skills around understanding the underlying contexts and forces that shaped activism at different points in time. By presenting players with what I have learned about how and why the Society supported some causes and not others between 1860 and 1890, and by highlighting the social and financial pressures that prevented the Society from achieving meaningful success at different points in this time period, my research can contribute to historical reasoning about the causal and contingent factors that structure Indigenous rights activism. I determined a single underlying argument for the game: that Victorian activism was shaped by its financial dependence on public donations, and these donations were, in turn, dependent on shifting social understandings of appropriate fields of activity, standards of evidence, and who was deserving of aid. The next step was to design a game that could convey this argument through the effective use of game design principles.

The next phase of game production, Design, entails constructing the narrative and mechanics of the game. This can be split into two processes: first identifying the mechanics that are most suitable for conveying an argument and then applying those mechanics to a game narrative that aligns with existing design principles for historical video games. I knew from Chapman’s delineation between realist games that facilitate historical consciousness/empathy and conceptual games that facilitate historical thinking/reasoning that a conceptual game was most appropriate for my learning objective. A useful starting place for determining the basic structure of a conceptual game is the Experiential Gaming Model (EGM), which is visualized in Figure 1 below.⁴⁸ The EGM holds that the relationship between a learning objective and gameplay is cyclical and revolves around a central challenge. Gameplay fosters experimentation and reflection on how to solve the central challenge, and each cycle of experimentation and feedback sparks new understandings, thereby pushing the player closer toward a holistic understanding of the learning objective.⁴⁹ Thus, one way to identify effective mechanics for my game was to first define a central challenge which, through experimentation and reflection, could reveal the causal and contingent factors behind Victorian activism. Next, I would find ways for cyclical experimentation, feedback, and reflection to reveal different elements of my argument.

Since my learning objective was for players to understand how financial and social pressures acted as contingent factors on the operation of Victorian activism, I decided that the central challenge of my game would be for players to take control of the Aborigines’ Protection Society and support as many Indigenous rights causes as possible in the face of increasing financial and social challenges. Active experimentation could be realized by presenting players with multiple options for responding to new Indigenous rights causes, with feedback being provided by player choices that would either damage or improve the Society’s financial and social position. To allow for cyclical experimentation and reflection, the game could be divided into multiple levels, with each level presenting new financial and social contexts that push players to reflect on the multiple causal and contingent factors shaping Victorian activism.

Having determined the overall shape of my game based on the EGM, the next step in design was to flesh out its specific mechanics. Following step 3 of Yang’s heuristic process for historical game design, I analyzed exemplar games to identify successful game mechanics that could be extrapolated to my simulation of the Aborigines’ Protection Society. Given that one of the main functions of the Society was to review and respond to letters from colonial informants, I was immediately drawn to a popular game called Papers, Please as inspiration for what my game could look like. Papers, Please is a fictional indie game released in 2013 that places players in the role of an immigration officer in a fantasy world meant to represent Soviet-era Eastern Europe. As people pass through an immigration checkpoint, the player is required to grant or refuse entry based on rules dictated by the department of immigration (see Figure 2).⁵⁰ Although it was not designed as an educational game, Papers, Please has been extensively studied as a tool for fostering pro-social morality, empathy, socio-political consciousness, and ethical self-reflection, indicating the game’s potential as an exemplar for successful public engagement with complex learning objectives.⁵¹ What really drew me to Papers, Please, however, was how the mechanic of granting or rejecting entry to a country is almost identical to the historical process by which the Aborigines’ Protection Society granted or rejected support for Indigenous rights campaigns in the settler colonies, and how the rules imposed in Papers, Please by the fictional ministry of immigration could be easily translated into the financial and societal “rules” imposed on the Aborigines’ Protection Society. I therefore decided to utilize a mechanic whereby players take on the role of secretary of the Aborigines’ Protection Society, and as petitions pass over their desk, they must consult continuously changing social norms and rules to balance the Society’s income, political influence, reputation, and member satisfaction (see Figure 3). Supporting the wrong petition at the wrong time would sacrifice one of these elements in favor of another element, all of which were based on historical research into the outcome of choices the real secretary made.

Of course, the fictional and purely commercial Papers, Please could not simply be cloned into an effective public history game, and I went through an iterative process of adjusting the format of Papers, Please to existing scholarship on historical games. There are various forms of such scholarship that historian-developers can consult, although not all are as easily applied to development. McCall’s “historical problem space” framework offers a set of questions that historian-developers can use to reflect upon their games.⁵² For example, by asking themselves, “why do the components in the historical gameworld function as they do, and what kinds of historical messages does their functioning promote,” historian-developers can ensure that their mechanics are purposeful and do not deviate from the historical arguments that they want to make. Yet the “historical problem space” was designed as an analytical framework rather than as a set of design principles, and as it is reflective rather than prescriptive, I found it more suited for post-development evaluation between game iterations than for the initial development process of the first iteration.

On the contrary, Julien Bazile’s design principles for what they call “historiographical games” establish three clear and prescriptive guidelines, and I used these design principles to ensure that my adaptation of Papers, Please adhered to the best practices of historical game design.⁵³ Bazile’s first design principle is to balance ludic coherency with historical validity, which refers to the challenge of negotiating shared authority between the player’s desire to make meaningful choices with fair results and the historian-developer’s desire to base choices and results on historical research. My game is tethered by historical validity in ways that Papers, Please, as a fictional game, is not, which required significant alterations. For example, Papers, Please has twenty possible game endings based on the choices that players make during the game, ranging from the player character’s family dying from starvation to the player character successfully leading a revolution against his country, meaning that players wield significant power to determine the course of the narrative. This is possible because Papers, Please does not have any limits as to what can happen within its narrative. Ab Uno Sanguine, on the other hand, must follow the rules of counterfactual historical representation in video games in which “scenarios do not purport to be factual, but they are inherently linked to baseline verisimilitude, which require historical reference to work.”⁵⁴ Therefore I had to reconceptualize my possible endings system based on what outcomes were historically plausible.

Based on the available evidence, I determined that there were eighteen plausible counterfactual outcomes to the Society’s activism:

• 1. APS goes bankrupt.

• 2. APS stays solvent with complete public, political, and member support.

• 3. APS stays solvent but with no political influence.

• 4. APS stays solvent but with minimal membership.

• 5. APS stays solvent but with a poor public reputation.

• 6. APS stays solvent but with no public awareness.

• 7.-18. Combinations of endings 3. through 6.

The “real” ending was a combination of endings 3, 4, and 5, with the APS needing to merge with the Anti-Slavery Society in 1909 due to its declining political and public reputations. As a result, players have the agency to manage the Society both better and worse than how it “actually happened.” And, importantly, players exercise this agency by making choices within a range of options derived from the Aborigines’ Protection Society’s historical records. For example, players can upgrade or downgrade their headquarters to save money or increase reputation, and each location they can choose from, as well as the cost of rent for each, are real locations that the Society occupied at different points in the nineteenth century. The same applies to all options that players face within the game. By mining the Society’s administrative records for meaningful choices for players to make and evidence-based plausible outcomes, I ensured that my game balanced player demands for ludic coherency with academic demands for historical validity.

Bazile’s second and third design principles both relate to balancing the amount of scholarly details presented in-game with maintaining player immersion. The second principle relates to effective in-game citation, referring to the need to ensure that history games are self-contained media for the transmission of historical arguments while, at the same time, preserving the flow of gameplay by not constantly breaking immersion with clunky footnotes. Similarly, the third principle relates to effective signposting of the historian’s presence in the game’s design, referring to the need to differentiate between elements such as original interpretations, syntheses of secondary literature, and details from primary sources. Again, this needs to be achieved without forcing players out of the state of play if they are uninterested in it, while not requiring that those who are interested leave the game to find this information.

Bazile designed these last two principles in the context of creating games as a form of academic argumentation (created by historians, for historians) and considers the display of evidence and signposting to be essential to making games that can be peer reviewed. But this also applies to public history contexts in which scholarly games need to be able to differentiate themselves from fictional games, to convince players that they are, indeed, based on rigorous historical research, and also to help players differentiate between “facts,” counterfactuals, and interpretations. Obviously, Papers, Please does not incorporate any of this as it is a fictional narrative, and so I adapted its game design by following Bazile’s suggestion to implement an optional historiographical overlay that players can toggle on and off as they wish (see Figure 4). By opening the game’s menu sidebar, players can access citations for the historical documents currently on-screen, including links to view digital versions of all sources. In the game’s current iteration, the menu does not yet have historiographical commentary or extensive signposting, but I plan to add this in future iterations. In this way, I designed Ab Uno Sanguine through a combination of game studies scholarship, a commercial exemplar game model, and historiographical game design principles.

The development phase consisted of coding the gameplay mechanics and creating the artistic assets. The first choice to make was which engine to build the game in. Choosing the right engine is one of the most important early decisions as it will impact what kind of game the designer can make, how difficult it will be, and how long it will take.⁵⁵ There are essentially three tiers of free game engines categorized by difficulty of use, each with extensive documentation to help historians on the road to game development. In the first and most user-friendly tier are engines like Twine, Bitsy, and PuzzleScript, which are designed to make small, simple games very quickly.⁵⁶ The second and moderately less user-friendly tier includes engines such as GameMaker and Construct 2, which present a steeper learning curve and can generally make more complicated games.⁵⁷ The third and most difficult tier includes Unity and Unreal Engine, which require programming knowledge and are widely used to make commercial games.⁵⁸

Any of these engines are viable options for a public history game, but I ultimately decided upon Twine for two reasons. First, Twine already has significant uptake within humanities departments as an educational tool, making it a more approachable case study to illustrate public history game development for researchers who are likely already aware of it.⁵⁹ Second, Twine games are built using HTML, CSS, and JavaScript, which are the universal building blocks of the modern internet, and many people already possess at least a basic understanding of them, making Twine very accessible for aspiring historian-developers. If you can make a basic website, you can make a Twine game. Moreover, as mentioned above, Twine comes in multiple levels of accessibility. For those with no interest in programming, the Harlowe story format allows you to avoid HTML/CSS/JavaScript entirely, and Jeremiah McCall and Gabi Kirillof offer detailed and accessible Harlowe guides for academics.⁶⁰ The SugarCube story format is better for those who want to dip their toe into basic programming, and Adam Hammond and Allison Parrish provide equally useful guides of this format designed for academics.⁶¹ Finally, the Snowman story format is available for the more adventurous developers who have a strong grasp of programming, and Jon Cinque’s tutorial provides an overview of how to use it.⁶² Moreover, there are countless videos, forums, and tutorials available online to guide historian-developers through the entire process of development.⁶³

Having chosen my game engine, I next began coding the mechanics and building the assets. It is important to note that this process will look very different for every project and that there are options for both programmers and non-programmers, graphic designers and non-graphic designers. For basic text-based games such as McCall’s Path of Honors, there is no requirement for assets of any kind (graphic, audio, or otherwise) and often only a simple mechanic of tracking player statistics, which can easily be copied from tutorials.⁶⁴ People with minimal programming experience can make more complicated game mechanics by adapting existing open-source JavaScript libraries such as Physics.js to add physics events, Anime.js to add animations, Popper.js to add tooltips, and so on.⁶⁵ Non-graphic or sound designers can use open-source assets from databases such as OpenGameArt.org without having to make their own graphics or sounds. In my case, although I began this project with no game development experience, I did have web development experience, and so opted to create my own mechanics and assets.

I began by laying out a CSS grid of how the user interface would look, with a central desk in the middle, a drawer at the bottom, and a view of a street on top. Next, I created HTML elements of the game objects: the table, the drawer, envelopes, letters, pedestrians walking along the street, and a horse-drawn postal carriage. Finally, I used JavaScript to add interactivity to the elements: clicking the drawer handle opens the drawer, right-clicking the envelope reveals the letter, and dragging the letter into the street results in the postal wagon driving across the street to pick it up. For a simple example, see Table 1 below for how I programmed envelopes that, when right-clicked, disappear and reveal the letter enclosed. The CSS places both letter and envelope at the same point in the grid and defines a “hidden” rule that can make HTML elements invisible. The HTML assigns the “hidden” rule to the letter and assigns an anchor point (such as a link) to the envelope that executes a “getLetter()” function when right-clicked. The JavaScript defines the “getLetter()” function as removing the “hidden” rule from the letter and adding it to the envelope, thus making it appear as if the envelope is being opened, to reveal the enclosed letter.

In many cases, I was able to adapt existing code rather than writing it from scratch, which would significantly increase production time. For example, one of the most basic mechanics of Ab Uno Sanguine is dragging and dropping, a behavior that can be accomplished with Interact.js, an open-source JavaScript library.⁶⁶ All I had to do was import the Interact.js library into my Twine project and configure its code for my specific needs. For example, Table 2 shows a comparison of Interact.js’s prefabricated code on the left and some small customizations I made on the right. Much like in the above example, this code simply says that when an envelope is dragged and dropped into the street, the “hidden” rule is applied to the envelope and removed from the postal carriage, and then the “animation” rule is added to the postal carriage, making it roll across the screen and appear as though the carriage collected the envelope.

Another basic mechanic is managing resources. Players have to spend money to maintain resources like “public awareness” and “member satisfaction” to finish the game. To do so, they must observe resource levels, depicted as colored-in progress meters, and these meters could be coded using open-source meter macros created specifically for Twine by Chapel.⁶⁷ See Table 3 below for how this was coded for the “public awareness” resource, which was done entirely using Twine’s macro language rather than JavaScript. After importing the macros into Twine, I first defined the variable “$awareness” as a number between 0 and 1, with 0 corresponding to zero percent and 1 to 100 percent in the progress meter. I then declared a new progress meter named awareness, assigned it the “$awareness” value, and set its color and size. Finally, I created a button that, when clicked, updates the meter with the current value of “$awareness.”

The game also consists of a substantial amount of original coding. At the same time, it is often possible to save time and trouble by searching for open-source code that can be adapted to a game’s specific purposes.

Once the basic mechanics were programmed using place-holder text and images, the final step in development was to generate assets. There are two primary types of assets in Ab Uno Sanguine: foreground documents and background props. Foreground documents include the archival letters, newspapers, and monographs that players interact with during gameplay, while background props include the streetscapes, horse-drawn carriages, and furniture that help establish the historical setting. I created both types of assets by copying real-world historical artifacts, and I used two methods of reproduction: manual SVG art and automated SVG art. SVG (or Scalable Vector Graphics) is a vector image format for creating two-dimensional digital images that are responsive—that is, images that can be manipulated without losing image quality. Vector assets are infinitely preferable to raster graphics (such as JPGs and PNGs) for game development because they have smaller file sizes, can be easily altered during development cycles, and are endlessly scalable during gameplay (for example, they can be zoomed and shrunk without altering image quality).

Manual SVG art involves using vector graphics software such as Adobe Illustrator to re-create historical images by hand with lines and shapes (see Figures 6 and 7). Doing this can be very time consuming and requires a certain level of imagination and software proficiency, but it allows you to transform historical documents and images into lightweight and responsive assets with complete customizability. Automated SVG art, on the other hand, uses image tracing tools to “vectorize” existing images (see Figures 8 and 9). This is an easy way to quickly generate highly detailed reproductions of historical images, but it comes at a cost. The amount of detail in automated SVG art means that they contain more shapes, which can create a host of problems down the line, such as longer loading times, slower framerates, and difficulty editing or customizing assets in future game iterations. To balance the benefits and drawbacks of both methods of reproduction, most assets in Ab Uno Sanguine were created manually, and only those assets that were beyond my artistic skills or time restraints were automatically generated. Once my SVG assets were generated, I added them to the game by first defining a Twine widget to house the SVG code and then “calling” the widget (i.e. using shorthand to reuse pre-defined code multiple times) on each scene of the game that it appears within (see Table 4 below for an example). With the mechanics coded and assets generated, the first iteration of Ab Uno Sanguine was ready to be released to the public.

The implementation phase involved delivering Ab Uno Sanguine to its intended audience in a format most conducive to achieving the game’s objectives. At the most basic level, this meant publishing the game on itch.io—a free hosting service for indie games—and advertising it to people interested in Indigenous rights activism and decolonization via social media. However, merely circulating a link and allowing people to play the game without educational scaffolding is a sub-optimal route to achieving objectives. Scaffolding is, of course, a staple of public history in the form of interpretive museum panels that prompt reflection on exhibit narratives, and the same concept applies to video games.⁶⁸ While various studies support the educational efficacy of historical games in controlled learning environments, John Majewski argues that only a small minority of gamers outside of such environments critically engage with the historical content of history games.⁶⁹ Lisa Gilbert challenges this finding somewhat, countering that unscaffolded play in uncontrolled environments is significantly correlated with increased historical empathy.⁷⁰ However, even Gilbert acknowledges that gamers frequently place too much trust in depictions of history in video games and as a result fail to demonstrate historical thinking and reasoning skills without scaffolds prompting them to reflect on the nature of histories they encounter in games.⁷¹ Serious games scholars, therefore, emphasize the importance of scaffolding gameplay to prompt player reflection.⁷² Some scaffolding is built directly into Ab Uno Sanguine’s gameplay. For example, when players lose the game by bankrupting the Aborigines’ Protection Society, they are prompted to think about the relationships between financial and humanitarian imperatives and their impact on activism. But this form of internal scaffolding is only triggered after playing the game for a significant amount of time and, therefore, cannot prompt players to play the game with reflection in mind from the beginning. External scaffolding, on the other hand, helps control the situation in which players engage with a game by prompting them into a reflective state of mind from the outset.⁷³

Following a brief playtesting phase in which I invited fellow PhD students to play the game and report any bugs or possible improvements, I carried out a three-week social media campaign to deliver Ab Uno Sanguine to Indigenous rights activists in a targeted and scaffolded manner. Social media posts utilized graphics comparing real historical documents and images to their in-game SVG reproductions to visually indicate the historical nature of the game, while questions connecting the game to issues in Indigenous rights activism provided scaffolding that prompted players to think historically while playing the game. Further scaffolding was provided through citations and in the game description, which explained the historical context, as well as with a short warning on the title screen indicating that the game contains histories of colonial oppression. Posts were then shared through Twitter hashtags and Facebook groups dedicated to decolonization and Indigenous rights.

Evaluation, or measuring the project’s impact against its intended outcomes, is one of the most important stages of any public history or public engagement project. Historical game scholars have long touted video games as being uniquely positioned to easily harvest data for evaluation through player participation in comment sections and game forums. Majewski, for instance, uses social media posts to analyze the historical impact of Sid Meier’s Civilization games, and McCall contends that internet forums “where players can and do share their thoughts and opinions about games with their peers offer an untapped resource for investigating how players can interact with the history embedded in games.”⁷⁴ Indeed, one of the primary ways in which researchers evaluate the moral and social impact of Papers, Please is by studying online comments and forum posts.⁷⁵ Consequently, as a PhD student with no budget to fund a survey or to compensate interviewees, I planned to take advantage of voluntary online comments to evaluate the impact of my game.

Unfortunately, the capacity for commercial games such as those studied by Majewski and McCall to generate online discussion has so far failed to translate to my small indie game. Itch.io provides basic analytics tools that are useful for evaluating how many people have played the game, and the numbers have certainly been promising. The game was played more than two hundred times during the three-week social media campaign, and that number continues to grow, with more than five hundred plays at the time of publication. This alone suggests that Ab Uno Sanguine has successfully presented my historical arguments to a substantial public audience. However, the analytics within itch.io are not capable of differentiating between different audience types, and not a single comment has yet been written indicating how players have responded to the historical content in the game despite repeated public requests for players to offer their thoughts. For my purposes, this was not too much of a problem since Ab Uno Sanguine was a personal project with no funding body to report to. But for those who would like to build a public history video game as part of the knowledge mobilization section on a major funding proposal, I would caution against relying on voluntary comments and recommend implementing a more rigorous evaluation methodology. Some methodological suggestions are offered in the following section.

Although I learned many things from the experience of making Ab Uno Sanguine, there are two main takeaways that will guide the next iteration of the game, which I hope will be useful for other historians planning to make a public history game. My first takeaway is that asset generation was by far the most time-consuming and headache-inducing part of the game creation process. I had originally anticipated coding game mechanics to be the most difficult part of the process, and there were certainly many evenings spent trying to troubleshoot mysterious bugs or make certain mechanics work with custom JavaScript, but those evenings were nothing compared to the countless weekends spent creating, editing, and re-editing the more than four hundred SVG files that make up Ab Uno Sanguine. With hindsight, there were several things I could have done to drastically reduce this colossal headache. For example, instead of making an individual SVG version of each archival letter I reproduced in the game, I could have made a single blank SVG template of a piece of paper and overlaid different texts on top of the template with CSS. I could have also designed the rules so that players, rather than briefly analyzing a large number of archival letters, could have been required to more extensively analyze a small number of letters. During the game design stage, I could also have designed a less visual-intensive, more text-based game. Those with available funds to draw on could also outsource asset production to graphic design students or freelancing services such as Fiverr. However it is done, I would strongly encourage future historian-developers to seriously consider ways to reduce the number of assets they need to create in order to expedite the game production process and significantly reduce labor requirements.

My second major takeaway relates to the difficulties I encountered conducting meaningful evaluation. As explained above, I had hoped to capitalize on the culture of leaving comments, reviews, and suggestions on video game web pages to evaluate the game’s impact, but these comments, reviews, and suggestions did not materialize. This was not the same issue that Kee and Bachynski encountered, in which lack of fun prevented people from wanting to play their game. Nor was it the issue of hitting the “digital divide” that Andrew Hurley experienced, in which he observed “abysmal” engagement in his digital public history project because his target audience had limited access to the internet.⁷⁶ On the contrary, my game attracted significant numbers of players, but it was comments rather than engagement that I could not secure. I did obtain plenty of feedback from playtesters, but my playtesters were not drawn from my target audience and, therefore, were of little relevance as to its impact. Since evaluation is a fundamental aspect of public engagement and necessary to determine whether a project has achieved its objectives, this is a rather significant problem that needs to be addressed if video games are to be accepted as a venue for public history. I see three possible solutions.

The first option would be to conduct player interviews or surveys to assess their historical consciousness and empathy before and after playing the game. This would certainly be the most effective and scientific means of evaluation, yet it is also expensive. I have previously conducted interviews and a survey to assess the impact of a public health intervention, and the cost of interviewee compensation and hiring a professional surveying company went well beyond CAD $1,000. Given that the purpose of this paper is to report on the potential for all historians, including graduate students and earlier career researchers, to easily and cheaply disseminate their research to new audiences, I do not consider interviews and surveys to be a realistic method of evaluation. Players could certainly be asked to voluntarily fill out a survey after completing the game, but as players were unwilling to write brief comments on the game’s webpage, I do not see this as a viable alternative.

A better way to conduct evaluation would be to include players in the game design process. Serious games scholars have long argued for the value of a participatory design process. Research from other disciplinary contexts shows that by involving stakeholders and target audience members in the game design process, game designers can ensure that games reflect stakeholder interests and are relevant to the specific contexts of target audiences.⁷⁷ Moreover, participatory game design moves even further toward the goal of shared historical authority, turning players into player-designers who not only shape game narratives through meaningful choices and guide historical discussions in affinity spaces, but also inform game structures themselves. At the same time, a participatory design process can also offer a built-in means of evaluation. Instead of evaluating the impact of playing a game on a large number of players, historian-developers can evaluate the impact of co-designing a game on a smaller number of participants by observing how co-designers respond to a game’s intended outcomes throughout the design process. Of course, this method of evaluation is dependent on there being enough stakeholders from the target audience interested in game design, and the process of collaboration will extend the production timeline significantly. But if participants are willing to be compensated in game design experience and being credited on an indie game, this form of evaluation may be more accessible to researchers without public engagement budgets. This is the method of evaluation that I plan to follow in the next iteration of Ab Uno Sanguine by inviting people who have played the game and are interested in game design to collaborate on revising.

An alternative method of evaluation would be to incorporate feedback mechanisms into the actual gameplay. For example, pre-gameplay and post-gameplay survey questions could be reformulated as dialogue options at the beginning and end of the game, and player dialogue decisions could be exported to an external database. This model of evaluation is proposed by Christina Tsita and Maya Satratzemi, who argue that data collected from gameplay can be used as formal evaluation of learning outcomes, and it is certainly the easiest and cheapest form of evaluation.⁷⁸ However, any data obtained by such a method would be unreliable because it was collected in a state of play; many players would simply choose whichever dialogue options they believed would help them win the game rather than choosing the options that reflect their actual opinions. There are ways to increase reliability through in-game signposting, but this method of evaluation will never be as reliable as first-hand interaction with players. As such, in-game feedback mechanisms should be used as suggestive or preliminary assessment and supplemented with additional methods such as those discussed above.

Predictably, designing and developing Ab Uno Sanguine took me longer than I anticipated. I originally planned on a production timeline of around six months, working a couple of hours every weekend, but in actuality it took closer to a year. From my perspective, it was a year well spent. To be sure, the missteps I made in terms of evaluation make it difficult to determine whether my explicit outcomes were reached. However, at the very least, there are hundreds of people now aware of the Aborigines’ Protection Society who previously had no knowledge of it. Moreover, by sharing a case study of my missteps around evaluation as well as showing that I was overly ambitious and needlessly laborious with asset creation, I hope I have helped prevent future historian-developers from making the same errors. Game production is an iterative process, with games progressing initially from alpha to beta and then receiving continual revision via updates and bug fixes. Missteps are thus not to be perceived as “failures” but as areas for future improvement. Fields such as public health and communication studies have dozens of published reflections on games for public engagement to guide and improve future game development, leading to a snowball effect of more and better games. The only way history can catch up to these disciplines is by building up our own library of errors made and lessons learned, and this paper is but my own humble contribution.