Gamasutra: The Art & Business of Making Gamesspacer
The Tyranny of the Game Engine
Printer-Friendly VersionPrinter-Friendly Version
View All     RSS
April 16, 2014
arrowPress Releases
April 16, 2014
PR Newswire
View All
View All     Submit Event

If you enjoy reading this site, you might also want to check out these UBM TechWeb sites:

The Tyranny of the Game Engine
by Declan Kolakowski on 02/06/14 06:05:00 am   Featured Blogs

The following blog post, unless otherwise noted, was written by a member of Gamasutra’s community.
The thoughts and opinions expressed are those of the writer and not Gamasutra or its parent company.


This article was cross-posted on my personal blog.

Ask almost any game developer (who isn’t a tools programmer) what they think of modern game engines and the response will be overwhelmingly positive.

Developers will enthusiastically say things like:

“Game engines are democratizing development!”

“Game engines make complex tasks easier and quicker!”

“With a game engine I can make anything I want!”

Etc. etc.

While this may be true to an extent, there is a big problem with game engines that our development community seems set on ignoring:

Game engines force you to think about creating games in a particular way.

Game engines are specialised to solve a particular set of problems in the creation of games. The nature of these problems are overwhelmingly influenced by the AAA standard of representation in games. In essence all game engines are trying to solve two very limited problems:

  1. Representing a 3D space using a first or third person avatar.
  2. Representing 2D space with an animated sprite.

There are a great number of games (as we have seen) that can be created through a grounding in these two forms, however, the huge foundational assumptions game engines force creators to make about the mode of representation and the form of design in games is forcing the development commmunity into particular ways of thinking. As soon as you load up any game engine these choices have already been made for you. You are immediately presented with a base representation of your game world which is geared towards emulating the “standard mode” of game creation.

In Rules of Play Katie Salen and Eric Zimmerman note that almost all conceptualisations of games start from the point of view of an avatar moving in some form of space. This singular and limited understanding of the representative power of our medium has lead to a homogenisation of how we conceptualize games, which, in my opinion, is incredibly harmful to the creativity of our industry.

To make it clear, I do not think that it is impossible to create original games using game engines that depart from the tropes of 3D/2D representational space, but at some level, even if you can escape the tyranny of the game engine interface, you are forced to understand the process of your creation in terms of its basic functionality. The basis for your creation with a game engine will always be from these specialised modes of limited representational space and everything you try to do to escape from it will have to be conceptualised as a deviation from that. A subversion of what is made easy by the engine. A struggle to realise a vision that is opposed to the engine’s intent and design.

Humans naturally try and gravitate to the easiest way of accomplishing their goals. When everything you try to do to realise your imaginative vision of a game space is set in opposition to the interface that mediates your work, you are being manipulated into thinking more and more in terms of the dominant modes of game representation.

All of this leads to a damaging feedback loop between consumer and creator that cements these assumptions about game design ever more deeply. Consumers (including developers) buy and play the huge amount of games that use the 3D/2D representational norms thus solidifying expectations about what a game is and is not. This in turn leads to developers making more games that fit these preconceptions and cater to their perceptions of the market, thus reinforcing the consumers expectations again and feeding back into the design process once more.

How did this situation arise? One way to understand it is to look at the music industry – an even more homogenised creative sector than games or film – where this process of conceptual domination has strong parallels. Theodor Adorno theorised that if people are constantly played the music they like and never exposed to things they might not enjoy because they are unknown then the creation of sound will slowly become homogenised as all extraneous and challenging influences are removed from the creative process in a inward spiralling feedback loop designed to capitalise on marketing music to the lowest common denominator of consumer. As we can see from the state of the music industry today, he was not far wrong. One need only look at the development of software like Logic and Pro Tools to see how specialised these interfaces have become in fostering the imperialistic propagation of the 3-5 minute pop song complete with top ended vocal line, electric rhythm section and bass boosted drum beats, forcing musicians to think in terms of these culturally dominant forms.

The reason I say that the music industry is even more homogenised than games or film is because indie/alternative music from self-confessed pop hating bands is still incredibly close to the pop music formula. These musicians generally still use the standard band set up (guitar and drum heavy), the standard forms of song structure (verse/chorus), the standard harmonic progressions (I-IV-V-I) and standard subject matter (sex and relationships). What do the pop music industry and the independent music industry share at the most intimate level?  Both use same software interfaces to create their music. It is not a big stretch of the imagination to intimate that these assumptions from huge corporations about creativity are passed on to the independents through the musical interfaces they fund and control.

Returning to the games industry; we can see how the development of the game engine may be fostering a similar form of creative dominance. Over time, like genetic organisms, game engines have adapted and specialised themselves to cater to this wildly uncreative subset of game creators – a form that has been mainly fostered by the AAA industry and its control over so much of the industry’s mind-share. Each iteration of these game engines has been focused on honing the capabilities to produce a specific “correct” result while making any other forms of development increasingly tangential and difficult to access in their interfaces. This developmental homogenisation leads to game creators being unable to distinguish between “developing games” and “developing a subset of possible games” because of the tools they use reinforcing a particular understanding of how we create in our medium. This understanding has even more alarming cultural currency among the average consumer who has a very specific way of understanding what game development is and what it is not.

As the younger generation begins to make games, more and more we will find developers who have only ever used these streamlined interfaces and if we are not vigilant in reminding them of the conceptual limitations of these tools the imaginative possibilities of our medium might be abandoned in the future.

Part of the reason game engines have become so important to our development culture is that they allow creators to begin making games without having to first front load their development process by worrying about “trivial” things like how they will render their characters on screen, or how they will display menus but instead jump straight in to making the content for their game. This is certainly a big advantage if you want to actually make games rather than just end up being a very good tools programmer. I am not saying that we should cease to use game engines but we should definitely have an awareness of the assumptions that engines are instilling in our industry. Like many forms of psychological and cognitive dissonance, simple awareness can be enough to over come them.

As a prototyping tool game engines are unmatched, however, if you are trying to look and think outside the realms standard game representation I would recommend you try coding from scratch or at least thinking your framework from the bottom up and tweaking (even in a game engine) the most fundamental aspects of your game before rushing into content creation. You will be amazed how many unexpected and exciting things emerge from revolting against the tyranny of the game engine.

If you have other solutions, ideas or a completely different view I’d love to hear it in the comments.


Related Jobs

Treyarch / Activision
Treyarch / Activision — Santa Monica, California, United States

Production Coordinator (temporary) - Treyarch
Vicarious Visions / Activision
Vicarious Visions / Activision — Albany, New York, United States

Senior Software Engineer-Vicarious Visions
Telltale Games
Telltale Games — San Rafael, California, United States

Senior Platform Engineer (Xbox One & PS4) @ Telltale Games
Nintendo of America Inc.
Nintendo of America Inc. — Redmond, Washington, United States

Associate Software Support Engineer


Dmitriy Barabanschikov
profile image
I disagree with your conclusions and would like to offer a counterpoint. Modern game engine can be viewed as a tool of the trade, not a part of the creation. I'd compare it to movie cameras and film in that the engine is an essential part of the process, but, ideally, should not be noticeable in the final product.

I also disagree with your interpretation of the game engine goals and list of the problems it solves. If you get to the core of it, almost every engine does one thing: it represents a number of objects and their relations in a 2D or a 3D space. Avatars in 3D games are completely optional, that is, if you don't consider the basic camera as an avatar. Strategy games, puzzle games and many other genres make do without a concrete avatar. At the same time, sprites can be used to represent pretty much anything non-3D, not necessarily a coherent space.

Any game, however abstract, still have the need to show (or tell) something to the player. Engines help us do exactly that. I believe that we should be celebrating the end of times when anyone wanting to make a game had to go through fire and water to make it work. The tools are now plentiful and are easy to work with. Many more people from various backgrounds are now making their own voices heard in our ever-expanding field. What's not to like?

Alejandro Rodriguez
profile image
Pretty much this.

Game engines let me see what I make (thanks to cameras and rendering)
Game engines let me make things that do stuff (thanks to scripting)
Game engines let me transcend platform barriers (thanks to cross-platform compatibility)

The content I create, the nature of the games I imagine and the end result are mine to imagine and execute on. An engine used traditionally for shooters can be used to make an ant colony simulator, a sunshine machine, or a smile factory.

Interest and results are big motivators for plenty of people that would normally want nothing to do with exploring the greater nuance of tech. They want a venue through which to create.

Anyone that does it long enough, and feels the burning need to solve problems that game engines don't will do so. It's part of the process of mastery. Others will not.

Declan Kolakowski
profile image
I agree with you and I feel I acknowledge these criticisms in my article. While it is "possible" to make a vast array of different things using a game engine. The base way they are set up encourages a set of assumptions about form that many developers seem to take for granted. Obviously all games have to show some representation of space. But, if you open up Unity's 2D engine everything is pretty much geared towards producing a platformer (because that is generally considered the archetypal 2D game) which is a very limited and specific representation of that space. As you point out, the huge number of creative puzzle games play around with that space and do interesting things, but generally, the engines tacitly encourage you to create a representation of 2D space that centres around a character, similarly 3D encourages first person / 3rd person representations of a very standard vision of 3 dimensional space. Engines are an important part of creation and they do allow for easier development but they do force some uncomfortable assumptions about development. Admittedly many developers transcend these but for every one that does there is another who falls into the trap of creating something generic and unthinking and I think this can be partly attributed to what their tools are unwittingly encouraging them to do.

Thanks for your response! :)

Andrey Coutinho
profile image
While I agree that the engines do point you towards some specific scenarios, I also believe that the really creative designer minds won't settle for the "standard scenario".

In my particular case, I usually design the game concept in my mind, then I put it into paper and design documents, and only later turn to the engines to try and "make it real".

What limits the design decisions for me is my own creativity, not the engine. I will twist the engine around in any way I can for the game to match my vision. If that means using functions and features from game engines in ways that are completely unintended by the engine developers, so be it.

But I do agree with you that, if the programmer tries a more organic, bottom-up, experimental type of game design flow (programming stuff, testing and seeing what happens without planning too much beforehand), using a game engine is bound to direct his dicoveries and end up limiting them.

Artur Moreira
profile image
Good article. You get lost a bit in too much unnecessary text, but the point is there. As far as my opinion goes, I think the blog post hits some truths, but is pretty extreme. A common game engine might be adequate for most games you can think of. Plus, you see tools being modded to achieve a particular design all the time. For other kinds of experimental games, its usually easier to just roll your own.

And let's not forget many engines have low level interfaces too and give you all the flexibility you could ever want.

Your points might apply the best to beginners who struggle with the underlying concepts of an engine and just want to use visual tools to get something done. The veterans will probably not be affected by the question you raise.

Michael Joseph
profile image
re: "Theodor Adorno theorised that if people are constantly played the music they like and never exposed to things they might not enjoy because they are unknown then the creation of sound will slowly become homogenised as all extraneous and challenging influences are removed from the creative process in a inward spiralling feedback loop designed to capitalise on marketing music to the lowest common denominator of consumer."

I think this is only true so long as you can prevent that person from discovering new things. Because I don't think you can totally shape them into forever liking only a particular class of thing. What Theodor Adorno seems to be suggesting is an effective monopoly of the arts and I think that requires a monopoly of distribution because as soon as people are exposed to some good new thing, they're going to crave more of it.

Wendelin Reich
profile image
This article misses a very important point. Modern games are complex systems that require highly intricate, organized forms of collaboration between people. Even a 'purist' couldn't (or wouldn') build a game without a C/C++ compiler and a host of other tools and APIs. All such tools impose limitations on what can be done - that's how technology works.

The less you want to rely on the constraints set by other people's tech, the simpler your game will have to be (unless you're incredibly rich, fantastically smart or have an infinite amount of time).

At the same time, some of the problems that have to be solved for video games are highly recurrent, and so they're exactly the kinds of problems that *should* be solved in standardized ways. The concept of a 'shader' is a nice example here. This is an idea to which the game industry owes so incredibly much, but it does introduce certain limitations. But rather than abandoning shaders, the industry continues to expand the concept (as in the newest generation of compute shaders).

The abstractions introduced by game engines are no different. However, historically, game engines used to be tightly guarded, closed, selectively licensed and incredibly expensive beasts. The emergence of open-source engines and later Unity, UDK etc. really blew up this space and unleashed an incredible amount of creativity and variation.

The answer to this so-called 'Tyranny' is not: less use of game engines, but more and more flexible tools (a broader variety of engines, add-ons etc.).

Benjamin Quintero
profile image
...there are reasons why some companies still roll their own tech. It has nothing to do with hubris or thinking that their life is going to be easier by writing their own version of a Unreal/idTech/CryEngine/ect. Different developers aim for different design objectives. If you license an engine and touch nearly every system and rewrite major components to meet your design, these are times when you ask yourself if the fees were worth it.