By Xinyu Kang (Peking University) and Yuantong Yun (Tsinghua University)
Introduction: Low-Polygon Medievalism as Technical Mediation
Low-polygon 3D assets dominate medieval-themed games, with 47% of Steam titles (2020-2023) using this optimization. While broadening accessibility, this approach creates historical homogenization as game engine constraints (vertex limits, texture budgets) reduce architectural complexity to calculable data. This study argues for “parametric disenchantment,” where historically significant structures lose material and cultural specificity—flying buttresses become triangular meshes, cathedral spires pixelated silhouettes. Through Kittler’s media materialism lens, engine limitations (not creative intent) determine medieval space representation, producing a standardized “pan-medieval” aesthetic erasing regional diversity. Asset marketplaces worsen this, with Unreal Marketplace showing a 17:1 ratio of Western European to non-European templates. The research advances two frameworks: “Polygonality Politics,” where technical reduction functions as an epistemological filter privileging engine-compatible geometries, and “Negative Historicity,” where players cognitively compensate for omissions by projecting cultural references onto simplified models. The study challenges technological neutrality in historical games, revealing how polygon counts enforce digital colonialism.
Parametric Disenchantment: Engine-Mediated Historical Erasure
Game engines prioritize technical efficiency over historical fidelity, reducing complex medieval architecture to quantifiable data units—a process termed “parametric disenchantment.” Gothic flying buttresses, designed for both structural and theological purposes, become mere triangular meshes in low-poly models, divorced from material contexts. Engine constraints enforce homogenization: UV mapping flattens intricate carvings into texture sheets, shader limitations standardize material properties across regional architectures, erasing cultural distinctions. This generates a pan-medieval aesthetic prioritizing engine compatibility. Asset stores accelerate this trend by incentivizing reusable, topology-optimized models. Kittler’s axiom manifests here: engine parameters act as nonhuman actors predetermining representational possibilities. When developers prioritize frame rate over structural authenticity, technical imperatives replace historical intentionality. This disenchantment is epistemological, naturalizing a decontextualized medievalism where technical feasibility dictates historical imagination.
Negative Historicity: Cognitive Insurgency of Players
“Negative historicity” describes how players cognitively compensate for low-poly medieval architecture’s representational limitations. When technical constraints reduce complex structures to simplified geometries, a perceptual void emerges between game representation and material history. Players project extratextual cultural knowledge onto minimalist models. Empirical observations from Stronghold (2001) show schematic architectural renditions catalyze associative cognition, mobilizing cultural memories from films, literature, or education. Eye-tracking experiments indicate prolonged visual fixation on abstracted contours rather than detailed textures, suggesting geometric primitives activate historical imagination neural pathways. Post-play interviews confirm participants unconsciously graft narrative layers onto polygonal forms. This challenges conventional historical simulation paradigms: while high-fidelity graphics create hermeneutic passivity, low-poly abstraction necessitates imaginative labor. Players become active co-constructors of subjective historicity, democratizing interpretive agency and resisting monolithic historical discourses encoded in asset pipelines.
Conclusion: The Dialectics of Low-Poly Medievalism
This study demonstrates that low-polygon modeling in game engines operates as a techno-historical mediator, simultaneously enabling and constraining narratives of medieval architecture. Through the framework of parametric disenchantment, we have documented how engine-imposed technical parameters—polygon budgets, texture resolution limits, and modular asset requirements—systematically reduce architectural complexity to computationally tractable data. This process divorces structures from their material histories, producing a homogenized “pan-medieval” aesthetic that erases regional specificities. Crucially, these technical constraints function not as neutral tools but as nonhuman actors that naturalize Eurocentric historical narratives through asset production pipelines.
The concept of negative historicity reveals how players cognitively compensate for this technical reductionism. Empirical evidence indicates that minimalist representations trigger associative thinking, prompting users to project extratextual cultural memories onto abstracted forms. This participatory gap-filling transforms players into co-constructors of historical meaning, challenging developer-authored narratives. However, this compensation mechanism cannot resolve the political economy underlying asset ecosystems. Our analysis of marketplace supply chains exposes how capital-driven optimization metrics privilege Western architectural templates, reinforcing cultural hegemony through technical standards.
These findings necessitate a fundamental reorientation in game production praxis. We propose critical asset design as an interventionist framework with three operational principles: (1) exposing technical mediation through visible LOD transitions and mesh simplification processes, (2) embedding historical metadata within assets to preserve contextual knowledge, and (3) decentralizing production through open-source tools that resist standardized aesthetic templates. This approach repositions games not as passive historical simulations but as contested terrains where technological infrastructures, economic imperatives, and cognitive practices continuously negotiate the past.
Ultimately, low-poly medievalism embodies a dialectic: its technical limitations perpetuate historical erasures while simultaneously creating spaces for democratic narrative engagement. Acknowledging this duality is essential for developing ethically grounded practices in historical game design.
Thank you so much for your paper Xinyu and Yuantong. It’s really interesting how the ‘pan-medieval aesthetic’ assets become so widespread even in games that are not making general claims to medievalism, like the minecraft screenshot here – how much do you think that this broader prevalence is because of the practical ease of reusing already-simplified styles, and how much is it a reflection of the general cultural desire for/favouring of medievalism (even if subconscious)?
Thanks for the paper! You seem to outline two trends – low fidelity graphics being common, and a lack of non-European elements in design space – but I’m not entirely sure these are as closely connected as your paper argues. What if any is the case that higher fidelity graphics would in any sense reduce this problem? Wouldn’t a wider adoption of high-detail graphics make it if anything harder for smaller scale developers to build representations of non-European topics?
I think you also overstate the optimisation side – that might have been the case when Stronghold came out, but in modern asset design and use, the reason I as a developer might choose to use low-poly graphics isn’t that machines can’t run in high poly, the chance that I’d build a game where that would matter for a modern PC is minimal. Something designed to run Baldur’s Gate 3 and Elden Ring is not going to bat an eyelid at whether I use low poly or mid-high poly graphics. Instead, a large reason for the adoption of these styles in indie developer circles is visual consistency: low poly objects are much easier to mix and match between packs than high poly.
I do think that your call for more focus on asset availability outside Eurocentric assets is interesting and important, and something I’ve been thinking about a lot myself: I’d be interested if you have any practical thoughts on how that could be practically made to happen.
I would add: modern asset design and use *in the Global North*. The constraints Xinyu and Yuantong observe here are still pretty evident (and relevant) in less privileged contexts.
ARISE’s games, for example, are constrained to low-poly because they are developed as educational products, and many schools, particular outside the largest urban centers, do not have any sort of infrastructure to run them. It’s not uncommon that the only hardware available to students are old smart phones. Most serious game developers in Brazil suffer a similar conundrum: we must either deal with low poly asset packs or put out a product that will only be played by a tiny elite (which sorts of defeats the purpose of the serious game in question in most situations).
Okay, yes, very much a good and fair point – what sort of age and spec of hardware/OSes are you generally building for as your lower end, out of curiosity? And thinking about the impacts on asset availability and indie development, is there a major local market for the sorts of games that run on old-smartphone-level devices (as opposed to what you’re doing with ARISE which obviously isn’t operating commercially), or is this a segment that doesn’t support very many developers?
I asked Alex Martire about the specs, and he told me they generally go with what Google Play Store tells them. Apparently, you cannot submit a game if it’s too heavy.
We did have a game with a PC-only version a few years ago, “Sambaquis”. You can see the mininum/recommended specs here: https://arise.furg.br/jogos/sambaquis-uma-historia-antes-do-brasil
I haven’t yet finished, or thoroughly tested, my own game, “Galar”, to know how light one can go with the hardware. I’ll let you know once I have a full-ish prototype ready (the current one is very bare-bones!)
As for the market, there is a huge demand for light smartphone games, and it’s directly correlated to income. However, whether that translates to a healthy development *environment* I’m not so sure… Brazilian games have to compete with ‘blockbuster’ mobile games like Pokemon Go, so it’s likely they make up a relatively small market share. (Although, according to Matthew Ball’s latest “State of the Game Industry”, that seems to be the case internationally as well).
Just wanted to thank you for bringing attention to this issue. I’m part of a Brazilian development team that makes serious archaeogames (ARISE), and the problems you identified match a lot of the issues we face in a day to day basis.
We need our games to be low-poly to make them available to the largest possible audience (and also because we design them to be distributed for free, so we don’t make any money out of them aside from the initial funding). However, we’re always between a rock and a hard place, having to resort to asset packs (because building everything by hand is time and manpower consuming), yet also being held to a high standard because our games are ‘serious’ ones, with claims of historical accuracy and ties to actual research practice.
Do you have any particular examples of the critical asset design practices you outlined at the end of your paper? Projects that, in your opinion, could be used as models for other games?