Making Grimsfield — A murky journey from animation to game design

Grimsfield — my first, (and possibly last) video game — is out next week. I thought I would do a deep dive into its development to share my learnings with designers and animators or anyone else brave enough to want to make their own game.

First Trailer for Grimsfield — This is what the game looks like for a bit of context.

For years now I have been making animated short films. It’s a real thrill to watch things move, and working in 3D has always reminded me of model building which I really loved when I was young. I made the decision to make a game for several reasons.

  • Skill development. There’s a kick to learning something new, it was like that when I first learned 3d software. As I started playing with programming I got a little thrill at seeing character designs I have worked with for years. And whilst I’m not intending to become a professional games developer, I love tinkering with new things.
  • Fun. I always wanted the animated films I made to feel like you’re on a theme park dark ride, being pulled through a story and having that story unfold in front of you bit by bit. Although Grimsfield did not end up in line with this design ethos, it was what I set out to do.
  • Audience. The animation community is small and very narrowly focused. Personal films are an amazing thing, but it does sometimes feel as if you’re building things just for a small set of people. I think games have a wider range of appeal, and I am not afraid to say that I would like my stupid little stories to try and reach as many people as possible.
  • Love of the medium. This is probably the main reason. It’s becoming clear that these games are still in their infancy. I have always been drawn to the more experimental vignette-style projects that exist, and so wanted to have a stab at building my own.
Grimsfield features a self aggrandising menu page.

If you work with 3D, you know about Unity — mostly just through word of mouth, things on Twitter, the splash screen on minigames you like. So I knew from its reputation that Unity was probably where I was gonna end up. So I downloaded and opened up the program. At a glance, it looks pretty much like a 3D app which is a familiar environment to me. However, I did not really know how anything worked…

Doing tutorials is a totally boring but pretty useful way of being introduced to a new tool or process. Combined with a healthy amount of playing around, that’s how I learned about animation too. This method is riddled with healthy frustration as you constantly try to get past the things you don’t know. I also figured if I put a bit of money down on a tutorial then I’d be more likely to do it, given that I hate wasting money. So I went and purchased a Skillshare course. This introduced me to Playmaker…

An example of a StateMachine I used in Grimsfield. This one controls animations played on the player character.

For the uninitiated, Playmaker is a visual programming interface for Unity. And for someone from an art background, it was revelatory. It allows you to program interactions through visual flowcharts, each node representing a proposition, or an action. All written in plain English. Discovering this tool was a revelation. I always knew that my main limitations would be my inability to write code. I am a terrible speller, so working with large amounts of words was going to have been a problem for me.

My first game was a little less ambitious.

If you’re an animator who fancies playing with game design I recommend you start here. If (like me) you come to it with a the kind of basic scripting knowledge you may have picked up from working in After Effects or 3D programs, it’s gonna feel like a breeze. In fact, the system feels comparable to Cinema 4D’s espresso system, which uses a simpler node-based approach to scripting. If you don’t have that kind of background the resources on the Hutton games website will help you get your head around variables, etc.

Discovering Playmaker made me feel like I could program freely in Unity, and I would actually be able to make something resembling a game. Huzzah! From there I started to hatch a plan of what to build…

The first person is my favourite type of perspective for storytelling games. For me, it gives a better sense of immersion for the storytelling. I like to feel like I am experiencing what a character is seeing. One of my favourite examples of this is Gravity’s Bone, and its series found here. I am pretty sure people will draw a comparison in art styles between this and Grimsfield.

So what drove my choice of perspective?

It was mainly driven by where my personal graphics and animation work are at the moment. To have the energy to work on a project out of hours I needed one main thing in place — the vision. Having a vision means that I can stay focused and can see the shape of what I am making (or currently think I am making). If I can see whole projects in my head and how they work it feels like something I could work on.

From my short brave new old in 2012 — one of the first outings for the cube headed people

I have been building little dioramas for a short time, I really like trying to build in a limited space, and these little renders have been a joy to work on for a couple of hours at a time. When I was experimenting with asset importing from Unity, I bought a few of my waterpark designs into the program to experience walking around them from a new perspective. While a neat experience, I thought they looked better from the isometric point of view still. It was at this point I thought it would be fun to have those limited set designs be a constraint on the game itself, and give me a very limited framework and structure to work in. I could lay out the sets as nodes on paper and control the scale of the project to something achievable.

One of my waterpark renders from the last year of my work.

With the perspective in mind, it was time to work out what this whole game was about. In recent years I have been trying to work thematically against my personal tastes a little bit in the hope it would take my work to new places, but as this was my first attempt at making a game I thought it would be nice to embrace all the things I love and try to put them in. Here are just a few.

Blue Jam Monologues

I loved Chris Morris radio show, but I particularly liked the stories and writing style he used. I probably used these stories as the basis of the plot for Grimsfield. There’s something detached and solipsistic which I like about the writing that I would love to be able to really capture one day.

Grim Fandango (2nd year)

I think it’s important to mention just how much this particular film noir adventure game has impacted everything I have ever worked on. There’s something about the 9th underworld that got under my skin. I think it’s how purely creative everything feels.

One of my favorite places in game ever. The Blue Casket.

Catterick & The League of Gentlemen

I knew I wanted to portray a side of England in games I have never seen. I and many others fetishise the likes of Twin Peaks and small town Americana for its assets and tone. But small town Britain is equally fascinating. I think everyone in the UK is either from or has been to these kinds of market towns, and I love how weird but familiar the location seems. I also totally love the absurdist comedy in these shows. Not something you come across that often anymore.

The Shipley Clock Tower actually appears in Grimsfield. Its a fascinating construction. (Image: David Spencer and licensed for reuse under this Creative Commons Licence)
Redcar Highstreet, Just the kind of place I wanted to capture (image: CC BY-SA 2.0)

My own animation

I know it seems to be a bit self-aggrandising to put myself in this kind of company, but in terms of assets I wanted to draw on what I already had worked on. The cube-headed people I used in my short Brave New Old is the character style I have been wanting to return to for some time, but never quite found the correct way, and The Circle Line is one of my favourite animations I ever made. I like the pace and energy and whilst I knew I would not be able to pull that speed across, I could bring the visual elements into the game.

One of my fav things I made. The Circle Line.

My favourite parts of most games are exploration, but I knew I would not be able to achieve all the storytelling I wanted to if I made a walking simulator or similar game. Combat was not something I felt I was technically able to achieve. So I figured the most interesting thing I could do is produce a small adventure game, similar to a twine game, but with a few more graphic elements to give it a sense of location and tone. I tend to enjoy games and films that show rather than tell, but technically I felt I was most equipped to handle this sort of work.

With this in mind, I started work on a game design document where I tried to order everything into something that resembled a brief that would allow me to work out what exactly was going to happen, what locations were available, and what tasks the player would need complete. Personally, it was a totally new process to think about a narrative that needs to be controlled. It felt like in some ways writing up some kind of complex circuit, and the set of logic gates and conditions, accept these were narrative points. I have included a few extracts from my game design document below.

Working out the quest logic

Pretty soon after that, I came up with the initial location plan. Based on what I thought would be enough for a small game. This was my first diagram.

Working out the locations

This is also my initial character list, I really wanted to try and keep everything down to the minimum, to give me maximum chance of actually getting finished.

Cast photo.

I imagine most game designers start with basic art to prototype their work and try to get systems built as quickly as possible. As an animator, I ended up working the other way round. Unity has the amazing ability to inline your 3d files directly, meaning I could get artwork and animation directly into Unity without having to learn new skills. So initially I worked on the project how I would import any one of my animated projects. I modelled all the locations and characters and, once I had everything ready, I could then ‘animate’ — except this time I would be programming.

Playmaker allows you to add the level of functionally to the geometry (or any other jobs), in a similar way to tagging in 3D programs. So the next phase was to build the ‘State Machines’ (or circuit as I like to call it) on top of the artwork and animation. This basically worked, though I’m not sure it’s the correct way of doing it.

The town layout was one of the first bits of artwork i made. The final layout was quite different.

As I mentioned I have never made a game before, and the idea of the player been in control seemed quite strange. For this reason, my initial prototypes used this really weird system of node-based movement. You would click on one of a number of points, then your character would be allowed to move there. It kind of looked cool, but from a user point of view it was awful because it meant you have to keep pressing down to move around, for no reason (a bit like hitman go but without the puzzle element).

A Terrible Idea

This was a real insight to me about game design itself. And how it’s about building machines and systems that people can use, to do the heavy lifting of your storytelling. I know this is a basic thing for most game designers, but as an animator, it was a huge learning and takeaway about the differences between the mediums. I’d never had to think ergonomically about animation.

Building the point and click character controller was a real revelation. It felt like I had saved myself huge amounts of time by not having to script out every position a player could stand in.

With these lessons in hand, I set about building my vertical slice (I know enough of game design to know about this). I won’t dwell too much on the finer details of the production itself, but here are a few broad strokes of important elements.

A big important part of the game was the dialogue, With enough perseverance I might have also been able to build this in Playmaker; but I instead opted for another custom solution, called Dialogue System. Again it had a bit of a learning curve, but it allowed me to edit and build dialogue a bit like in twine when you’re writing. It’s another node based editor and lets you see your conversation trees in a nice overview. It also works well with Playmaker and allows you to pass variables between the two.

A screen capture of the Mayor Robbinghams dialogue tree. These things are tricky to keep track off.

The dialogue itself was written in the evenings over 2 weeks or so. I personally have real trouble with words which is why I got into graphics when I was younger, so I anticipated a lot of trouble here. But I found as long as I knew what the outcome of the conversation was supposed to be I could pretty much improvise my way through it (after a few bottles of wine). Although I don’t think the writing in the game is its strong point, it’s definitely better than I anticipated, and I put that down to rhythm and consistency. If you ever need to do a similar thing I suggest you write as much as possible at once in order that the tone is consistent, as long as you know what you’re writing towards. Having the big picture in my head again helped the story flow.

Dialouge System also uses a simple logic system, to manage the variable effects on gameplay. It even writes the logic scripts for you.

Once major systems like locations, movement and dialogue were in place I built the ‘quest machine’ as I called it. In my head, there was going to be one empty object in Unity that would be responsible for managing the whole quest business, and be a giant central Playmaker circuit. This (I don’t think anyway), does not appear to be how game design is done. Instead, I discovered that my main quests would be distributed across all the relevant components in the game, that would activate and deactivate as needed, and when relevant. Its distribution of logic means you can quickly locate any bit of programming as needed, and means you don’t need to build everything at once, again quite insightful for an animator.

This null object in unity controlled all the systems, such as menus and sounds.

Finally, one of the biggest oversights I made from the start was the inclusion of ‘game stuff’ — menus, pointers, save games systems. With stuff like that, the reason for my oversight was it just sounded really boring. And to be honest this stuff was the plumbing and I kept putting it off. Building a save system is not a fun task, and if you’re an animator who’s planning on making a game I suggest you try to work these tasks into your project gradually, rather and letting all the boring jobs stack up will the end.

For some reason I thought that you could just use some kind of default save system. Thats only half true. I had to cobble alot of it together myself, am still not sure if I did it the right way.

I also need to mention that throughout the project (and my life) I have constantly relied on the troubleshooting ability of the internet. The nice thing about Unity is there is so much information out there on the internet. But both the people behind Playmaker and Dialogue System work extra hard to support their products, personally answering my questions and helping me develop when I ran into problems. Making Grimsfield would not have been possible without this support, for which I’m grateful

Music is probably one of the most important elements of Grimsfield. The reason for not writing about it more here is simply because it is also the only component I did not make myself.

This is the inspiration playlist I sent along with the music brief originally.

A few months into the project, when I had working prototype I approached Chris Reed, who I had previously worked with on my film Vandals. He was the composer for a stage show I watched a number of times called The Paper Cinema (which was one of the most wonderful experiences I have ever been lucky enough to watch). I really loved his work on Vandals, so I’d asked if he fancied contributing a few tracks to Grimsfield. The briefing was not too extensive and was limited to me offering a few references points, and the kind of mood I was going for. In return, I got this….

This is my fav Grimsfield track.

Which blew me away so much.

After this process that took about 4 months I was left holding something vaguely playable. It had some content holes here and there. Also, the ending was kind of broken. But for the most part, it resembled a game. It was at this stage I started sharing some builds with my friends and I started to think about the biggest and most difficult choices I would have to make.

I do animation most of the time, here are a few of my favs found in Grimsfield.

I am a believer in free from a personal point of view. My personal animation, in addition to being a hobby, has frankly also functioned as a kind of showcase for my skill set, in the hope that with will grant me the opportunity to get paid to make things I like working on. I also want as many people as possible to watch and enjoy what I have made. Not that anyone would ever buy said animation work as there’s just not that kind of value on video work. In a lot of ways I wanted to extend my thinking to the game I had made. I want as many people as possible to play Grimsfield, and by deciding to charge I have made the choice to deny the chance to build and audience.

Jim sells his stuff for like 3.99 thats the same as the game!

So why did I decide to charge? Here are a few reasons. Games are different, and longer. I think they offer people more for their money. Grimsfield has a playtime of around 45 minutes to an hour, most animations run at minutes. I also think there is an element of ownership that is not present with animation, you don’t ‘keep’ an animation, you watch it in a browser.

As much as I am a believer in free, I am also a believer in paying for things you like. I’d like to try and get into a place where I can be paid make things without people telling me how they should be made. I don’t think there is a single version of how this will be done. And in future, I will likely release things as free interactive projects, but for Grimsfield, I wanted to create a paid-for offering as it’s a more rounded product and story.

And of course, I’ve spent many weekends and evenings building the game. The decision to build it, and the decision to charge for it, are not ones I take lightly.

Having made the decision to publish as opposed to giving away, I thought I would work towards putting Grimsfield on Steam. A Greenlight campaign would also serve as an announcement of what I was working on. By comparison, I feel that most games announce themselves fairly early in production. However, I was not comfortable announcing until I felt that Grimsfield was accomplishable and of releasable quality. For that reason, I did not announce until I have full version I could play all the way through.

The Greenlight campaign itself was fairly painless. Chris provided some trailer music, that I cut together on one quiet afternoon. The whole Greenlighting process took about 3 weeks before we got the thumbs up. There are a lot of articles covering Greenlight out there, but I found this one particularly helpful.

Final greenlight score. Not many, but enough.

By far the hardest part of this entire process was uploading the game to the Steam store once it had been Greenlit. Hours and hours of trying to figure out, as a non-developer, how to use their SDK was very painful. But I got there in the end.

I think it’s also good to give a bit of space here to what I consider to be Grimsfield’s failings, and the things I wish I could have included.

Not enough surprise
The primary thing I would have liked to get in is more surprise. I love surprise in games (and life) more than anything. It seems like such an underused tool in the grand scheme of things. I had great plans for the ending of Grimsfield. And while what is there breaks the established rules of the game, I wanted to do so much more. Unfortunately, I came up against my own technical ability on some of these points.

The whole end section of Grimsfield was originally first person. It worked quite well, but I felt it came across as underdeveloped compared to the rest of the game. I really wanted a perspective shift for the last section and so I found a compromise with the poem-telling moment.

Too much dialogue
The largest trappings of the adventure game are branching dialogue trees. And Grimsfield relies on them heavily. I have worked hard to try and make sure that all the interactions with none player characters are worthwhile, and don’t fall into the trap of padding. But I would have liked to create some more unique gameplay elements (though not really object-based puzzles). However as this is my first significant effort, I will accept I had to work within my means for this project.


Lack of support.
I worked independently on the whole project. I have never made a game before, or studied game design. So I felt like for a lot of the project I was making it up as I went along. I’m sure there are a lot of time-saving tools I could have used, and better ways of approaching programming problems.

It would have been sensible for me to team up with a programmer and writer from the start and stick to what I know and I ‘m sure it would have produced a better product. On the flip side of this point, it feels good to have made something all myself.

Over the last few years, I have worked on many of my own projects. Making a game has certainly been the largest learning curve of any of them. But with the resources available online, it was possible. Coming from a visual background gave me an excellent jumping-off point for all aspects of the project. And having the vision from the very beginning meant I could hold the whole project in my head from start to finish.

Second Grimsfield Trailer, made a few weeks ago.

I would not consider myself a game developer, more someone who created a game based on what they know. But by carefully working within my own technical means I have been able to craft a playable story, which was what I set out to do.

Would I do it again? I am not sure I would set out to make another game of similar scale destined to be sold without more technical support. But I am sure I will make some of my 3d work interactive again in the future. And I am quite excited to do so, given how much I already know.

Go Buy

Footnote; Grimsfield is out on the 12th of May, a few weeks after it’s out, I may write about how it did, and the marketing, publishing part of the game, if people are interested.