Is it possible to create a prison management game without trivializing or misrepresenting the issue of mass incarceration? As video games mature and tackle more serious topics, players and developers should be aware of the values embedded in their systems.

Prison Architect is an upcoming game by Introversion Software, a British independent company. Dubbing themselves "the last of the bedroom programmers," Introversion played a key role in the renaissance of independent game development, producing a string of critically acclaimed titles and paving the way for digital distribution of third-party games on Steam.

Among their previous releases is one of my favorite games ever: Defcon, a spine-chilling, eerily beautiful multiplayer real-time strategy game in which players engage in a Cold-war era nuclear conflict. Each Defcon game culminates in a slow-motion Mutually Assured Destruction scenario. Whoever suffers the least amount of megadeaths is the winner.

Prison Architect is also tackling a dark subject, a subject that deserves special attention and defies any 'it's just a game' kind of dismissal.


As the name suggests, the player is in charge of designing (but also managing) a private penitentiary. The gameplay is reminiscent of sim games from the '90s, most notably Bullfrog's Theme Park and Theme Hospital: a mix of construction, zoning, research, resource and staff management.

Prison Architect is available as a paid alpha, a model of crowd-funding that rewards pre-orders with access to early versions of the game. More than 300K units have been sold for a total of over $10 million.

Why talk about an unreleased game?

One of the reasons developers pre-release games is to create a community around their titles and involve players in the development process. After hearing concerns about the very existence of a prison simulator, I decided to play it extensively and to start a conversation (in the comment section below) about the political and ethical implications of the game.


I've been tackling controversial topics in my games before, and I believe there's nothing inherently wrong in the choice of a subject. It's the way a subject is treated, the way a real-world phenomenon is translated into a playable model that is susceptible of criticism.

It goes without saying that all games, simulations in particular, are simplifications of existing systems. They can only aim to capture a limited set of features. But what gets included and what's left out of a model is decided by the designers, and ultimately determines what a game 'says', regardless of the author's intentions.


Introversion's co-founder Chris Delay is aware that they are dealing with serious matter, and, in an interview with Eurogamer, promised a nuanced product: "We didn't want to pick a tough topic like prisons and just make something completely cursory and surface level stupid."

In the same interview he also admits that, in the U.K., incarceration is not a major issue with huge class and racial implications as it is in the U.S.. Thus, as Brits, they may see the world in a different way.

There is no doubt that the prison system in America is beyond exceptional, housing 25% of the world's inmates (with the US representing only 5% of the world's population) according to figures from the American Civil Liberties Union. Introversion decided, nevertheless, to set their game in a nation modeled after the U.S., using dollars as currency, bright orange suits for inmates, license plate workshops, electric chairs and so on. The choice makes sense considering the size of U.S. market and the visibility of American prisons in media, but with great markets come great responsibilities.


Before it ships, can we help Introversion make a not completely cursory and stupid simulation?

What's the big deal with the prison system anyway?

When I first moved to the U.S., I couldn't understand why everybody within activist and progressive circles was so obsessed with prisons. It turns out, things are fucked up in a big way around here. Incarceration is out of control and has more than doubled in the last 20 years, according to numbers crunched by the advocacy group The Sentencing Project. The America justice system disproportionately and purposely targets minorities, which make up about 60% of the population behind the bars, according to the Center for American Progress. A skyrocketing prison population created a dysfunctional public-private industrial complex syphoning taxpayer money that could be used from crime prevention and education.


It's important to understand that mass incarceration is not just a direct outcome of economic inequality and social segregation, but also the result of political opportunism and white fear - i.e. the bipartisan "War on Drugs". There are cultural reasons and biases that have made a certain view on crime more successful than others (and subsequently produced the current legislation). The last thing we need is games or other cultural artifacts that cement this view instead of challenging it.

There are plenty of resources online and offline dissecting mass incarceration. A recent and accessible one is the excellent 2012 documentary The House I Live In.


Available on Netflix, iTunes, and on other distribution platforms.

A system critique

What I'd like to outline here is a series of issues related to the high-level design of Prison Architect. I don't intend to comment on bugs, user experience, and mere technical aspects.


Not having access to source code and design documents, my analysis is from the limited perspective of a player trying to understand the algorithm heuristically. Therefore, some details may be inaccurate. Everything here refers to the most recent version, alpha 16, released in December 2013.

Riots, riots, riots

In Prison Architect brawls and riots happen all the time, sometimes as soon as the inmates enter the building. Brawls often end up with inmates laying unconscious in pools of blood, injured guards, and damaged facilities. Riots block entire sectors of the prison, usually causing the death of guards, staff and, even more commonly, inmates themselves. For some reasons prisoners don't take control of the structure to express their grievances, force change or escape. They kind of just break things or punch or stab each other to death.


Disorderly conduct is an obvious way to provide feedback to the player when prisoner's needs are neglected, but even after mastering all the procedures and spending all the money available (about $90K, which is a lot in the game) for a state-of-the-art facility with very low population, I had to repress daily skirmishes in blood.

Simulations need to exaggerate feedback to prompt adjustments, and I certainly don't expect my inmates to enjoy their residency. But the continuous, frustrating, over-the-top violence suggests that we are dealing with an irrational, murderous, and suicidal horde that deserves no sympathy. Making the extraordinary ordinary (riots don't really happen that frequently anymore) is not only mystifying but also makes for a less subtle gameplay. The signal becomes noise and it becomes difficult to address needs by 'thoughtful' planning.


Moreover, other types of feedback could be implemented: hunger strikes, self harm, human rights and federal inspectors examining the facility, can add variety and extra level of challenge.

Solitary Confinement

Solitary confinement is one of the most controversial disciplinary procedures today. It is considered a form of psychological torture, a violation of human rights, and a largely ineffective measure. In 2013, 29,000 prisoners in the state of California organized a hunger strike to protest solitary confinement.


Even architects are taking a stance against the construction of inhumane cells as explained in this exceptional 99% Invisible podcast.


Isolation exercise yard, Security Housing Unit, Pelican Bay State Prison | Crescent City, California 2006 - by Richard Ross from Architecture of Authority

Solitary cells don't have any structural requirement in Prison Architect but the measure appears to kick in automatically, every time an inmate assaults staff or other prisoners— which is basically all the time. It would be hypocritical to not include a such a widespread measure in a prison simulation, but making it automatic and somewhat effective is a problematic choice.


Update, 1/21: As a few players have pointed out, it is indeed possible to disable solitary confinement after hiring a guard chief and researching prison policies. I somehow missed that feature, which is actually neat.

The War on Drugs never existed

In the United States, prisoners incarcerated on a drug charge comprise roughly half of the federal prison population. The number of drug offenders in state prisons has increased thirteen-fold since 1980, according to the Sentencing Project. Up until very recently, political administrations have been eager to raise minimum sentences for low-level drug offenders, de facto creating today's regime of mass incarceration.


To make things worse, drug use has been punished in a racially discriminatory way, punishing the possession of crack cocaine (used predominantly in poor African American communities) 100 times harsher than powdered cocaine (used predominantly by wealthy whites). The measure has been repealed in 2010 but the disparity remains, now with a ratio of 18 to 1.


In the world of Prison Architect, there is almost no trace of drug offenses. All inmates have a randomly generated criminal record and distinct features, including racial features which, according to the designers, are linked to gang affiliation and disorderly conduct. However, most of the prisoners are convicted of violent crimes, and even among minimum security level inmates, possession charges are incredibly rare. It could be that the randomness is not weighted. That is, each type of crime may have the same chance of being drawn.

No one expects a documentary-like accuracy in a game, but designers should be held accountable when they provide such a grossly distorted representation. Luckily, this is an issue that can be addressed with a few lines of code and no impact on the gameplay.


Punish or Rehabilitate?

The dramatic time compression in Prison Architect lets the player see the release of inmates and, in future versions, even ask for parole. But there's still no feedback regarding the degree of rehabilitation of prisoners, nor facilities providing education and development. In other words, there is an implicit assumption that prisons are just meant to contain and punish criminals without preparing them to "contribute to society" after their release.

The issue of recidivism is actually hinted at in the criminal records: most inmates have been convicted multiple times, suggesting that the system overall doesn't quite work.


Rehabilitation and reentry programs could be implemented without radical changes in the gameplay: educational or skill building facilities could work like any other room and could be an additional option in the "Regime" planner. The degree of rehabilitation and recidivism could contribute to the general evaluation. which, right now, is expressed exclusively in monetary terms, security breaches, and incidents.

"Arbeit macht frei"

At this stage of Prison Architect money can be made in three ways: through one-off grants rewarding the construction of certain facilities, daily federal and prisoner grants depending on the number of inmates, and prisoners' labour. The grants provide substantial sums in advance. They can be applied for at any time and don't impose penalties or deadlines for the completion of their to-do lists.


As a result, the current version of the game has a sandbox-like vibe, with plenty of grant money to play with at the beginning. Once the one-off money runs out, the prison that doesn't rely on license plate workshops becomes quickly unsustainable.

Prison labor is a big issue, concerning both conservatives and progressives. Prisoners are paid pennies per hour and they are in endless supply. The Department of Defense relies on it, and private companies are employing it at increasing rates, driven by the efforts of corporate megalobby ALEC.


Prison labor has been referred as a new form of slavery and as a publicly subsidized source of unemployment. While it is profitable for third party contractors, it would be misleading to think about it as the economic engine of the prison complex. Public and private prisons are far from being financially self-sustained.

In New York City a prisoner costs taxpayers more than $160K a year, according to a study by the state's Independent Budget Office. That is as much as four years of tuition at an Ivy League university. The comparison between public education and incarceration funding is apt since crime is by many seen as a failure of the educational system. For-profit prisons even, reportedly, manage to get public funding for empty cells.

In Prison Architect you get from $50 to $150 dollars a day for each prisoner which, in the game economy, is barely enough to keep an understaffed facility running, not to mention expansion or repairs after riots. The only way to make the penitentiary financially sustainable is to create license plate workshops, for which there is unlimited demand. This aspect is not only obfuscating the public costs of the prison system but also turning the game into something fundamentally different. Devoted players gather on a dedicated reddit channel to share strategies that usually involve the creation of sprawling labor camps.


The possibility of creating different types of prison, including corporate gulags, is definitely the most meaningful feature of Prison Architect, since it forces players to negotiate between the profit motive, the risk of rebellion, and their own ethical standards.

Another powerful tension, common in most sim games, is the entropy that arises from larger and more complex systems. Expansion may make sense economically but makes micromanaging harder. Inmates cease to be people and become numbers. The temptation of saving at the expenses of the prisoners' well being lurks behind every choice.


The problem is that, if prisoner labor is the only dominant strategy, if players don't get significant benefits from merely increasing the population, and if prisoners riot all the time anyway, such a compelling ethical role-play cannot take place, and the game becomes just an exercise in optimization and scaling.

The world outside

Simulations have boundaries, not only geographical ones—where the game map ends—but also conceptual—the scope of what is simulated. In Prison Architect there is a vertical road carrying the inflow of resources (like building materials, objects, inmates) and the outflow (garbage, dead bodies, liberated prisoners). It's implied that there is an external world where bricks or criminals are 'made'. It's a world the player has no control over so it doesn't need to be portrayed in the game. Since prisoners don't pay for their accommodation, money is also coming from the world outside, in the form of federal or special grants, suggesting a complex legal and political system.


Prison Architect displays an impressive granularity when it comes to the physical features of the building, requiring the player to position every single chair and shower grate, but the overarching system that keeps your prison running is depicted with very broad strokes.

Players are so busy micromanaging and streamlining the internal flows of bodies and objects that they may as well forget they are running a prison. The larger context is so far removed that Prison Architect could be easily adapted to simulate any total institution.

As Michel Foucault famously asked: "Is it surprising that prisons resemble factories, schools, barracks, hospitals, which all resemble prisons?" Yet, the social functions of these institutions and the legal and economic frameworks encompassing them are very different. Prison Architect could gain a lot of depth, replay value, and sense of purpose by implementing meaningful links with the external world. Laws could change affecting the number and type of inmates. Players could hire a lobbyist to push for new grants and policies to increase incarceration (yes, this is happening too). They could even have the option to bribe judges to funnel more delinquents to their establishment as it happened in the infamous Kids for Cash scandal.


With progressive reforms threatening their profits, the biggest opportunity for private prisons comes from the detention of undocumented immigrants, as shown in the The Atlantic, an aspect that could be easily implemented as another branch in the game's "Bureaucracy" tree. Communities in distressed rural areas are often enticed to build large prisons as a form of economic development, a phenomenon referred as 'prison towns', often with disastrous long-term effects. Prison Architect could hint at this relationship as well, either during the game or at the beginning, in the level selection.


Even in this alpha stage, Prison Architect is moddable. Graphics, maps, and grants can be customized with a bit of coding. This is a great direction: some of the issues above could conceivably be addressed by tweaking configuration files.


However, default settings are a rhetorical stance: they represent how things are supposed to work. One could replace sprites in Donkey Kong to subvert gender stereotypes but this doesn't absolve the developers who relied on them in the first place, nor change the experience of the vast majority of people who only got played to the original game.

My hope is to see more variables exposed to the players and accessible tools for modding. Mods can extend the lifecycle of a product and occasionally spin off into full-fledged games (that is the case of Counter-Strike and The Stanley Parable), but, more importantly, enable a positive proliferation of perspectives stemming from the same core engine.

Moreso than simply playing games, making and hacking games is a great way to investigate the world around us. It forces us to compare the digital models of our games with the mental models in our heads.


Prison Architect could easily become a powerful tool to see incarceration through multiple lenses, possibly even prefiguring alternatives to the existing dysfunctional system.

UPDATE: The game's developers have posted a thoughtful video response to this piece. You can see it here.


Paolo Pedercini is a professor at Carnegie Mellon University's School of Art. As Molleindustria, he develops games addressing issues of social and environmental justice (McDonald's videogame, Oiligarchy, Phone Story), religion (Faith Fighter), labor and alienation (Every Day the Same Dream, Unmanned). You can troll him on Twitter @molleindustria.

Top photo via Shutterstock/Oneword