An interview with underVillage’s Matthew Wills and Jason Sampson

underVillage is a New Zealand made indie title currently in development – a survival game driven by town building and a cacophony of morbidly cute and quirky characters. Although still in the early stages of production, the game is already taking shape as an intriguing work with unique approaches to both its aesthetic and design. I had the pleasure of discussing the game with two of its developers: Matthew Wills and Jason Sampson. The interview took place over a google doc which turned out to be a great format given the comprehensive responses that both of them gave about their work and the processes that have shaped their ideas so far.

Could you briefly describe the premise of underVillage?

MW. underVillage (which is still a working title by the way, there is a band by that name out there somewhere) is about a band of four small monsters, called underlings, coming together to form a civilisation away from the servitude of their former masters, the Underlords. Your goals are essentially to build a town, grow that town and fortify that town against lurking monsters who will mess you up.

A big part of your game revolves around procedural character generation, what kinds of combinations are possible in the current system and do any of these elements have effects that extend beyond aesthetics?

MW. Currently there are five sub-species of underling (Humanoid, Puppet, Bug, Ghost and Monster) each of which have 100 possible pairs of eyes, 40 different mouths and 40 different ears/horns/protrusions each. An underling isn’t limited to the parts within its species though, so right now there are about 20 million possibilities, if my maths is up to scratch (it’s not). That doesn’t even include all the skin tones and textures – they can rock some pretty cool scars on their faces too. We’ve toyed with the idea of every body part having some hidden stats – being a fishman makes you better at fishing, having bat ears makes you more alert… That sort of thing. But currently it’s not a feature.

JS. To add to Matt’s maths above, there are also 80 facial features, 37 different pieces of headwear, plus 10 types of neckwear that are not species specific. I think that brings the total number of possibilities into the billions, excluding colour variations. I should point out we’re not trying to be the Borderlands of character generation, it’s simply a function of how much art Matt has put together and the desire to have characters that are truly unique. Some of the assets may be removed, but for now we’re going ahead with most of what we have. The underlying code for all of this is more modular than when we first started out. Each body part selects and manages it’s own asset from an external bundle. The ability to include other bundles is planned. This could lend itself to some community customisation in the future, themed asset packs, etc. I do want the character generation to be more than just aesthetic, the details of which we still need to work out. If we tie that to asset packs, then I’ll need to sort out a way to include those attributes as part of the bundle, and prevent them from being manipulated to bring the game out of balance.

A sample of some of the procedural body parts used to generate characters. These are all mouth pieces. 

When creating artwork for the game you have taken a scanned, hand drawn approach as opposed to creating assets entirely digitally or with a tablet. Can you talk us through your workflow and why pen and paper is your method of choice?

MW. Pen and paper is how I learned to draw and it’s always been the way I do things. There is no concept art stage – all the art in the game is freestyle, drawn straight in ink, with no pencil sketch first. I don’t think I could have drawn 900 pairs of eyes if I had to do it in front of a computer – I was making these assets on lunch breaks, in bed, on the bus… Just whenever I had the time. I do, however, use a computer and tablet to add some greyscale shading, fix up the occasional wonky eyebrow, create blinking animations (its mostly erasing lines I’ve already drawn) and line everything up on sprite sheets. I may switch over entirely to digital if I invest in a Surface Pro 3 or some other art-friendly tablet-pc, but it seems unlikely at this point.

How has taking an ‘art first’ approach influenced the overall design of the project?

MW. Early on the team were struggling just to get the mechanics of randomising all these body parts down. There have been lots of fights about if I should now have to draw 500 eyes from multiple angles (the answer is “No, stop asking!”) and complaints about odd-shaped eyes obscuring mouth parts. I guess if I was going to start from scratch I would have probably made 20 eyes and done them perfectly – we’re way past that now.

JS. Early on the eyes were giving me fits because I was trying to come up with a way to make these 2D characters appear to be looking in the right direction when walking sideways. I wouldn’t say there were fights, but until I could come up with a way to make these assets look reasonably correct, I did ask Matt a couple of times to his annoyance to consider it. *smile* I’m pretty happy with the look we have now, some simple transformations and the new head type were all that it took.


Would you say the game is working to a defined, overarching design or are things more iterative and spontaneous as you implement the core mechanics?

MW. Right back at the beginning of the project, I was just making art and throwing it up on twitter for people to look at. I had a 10 page design bible for people interested in helping after that – it was pretty vague, but I think most of that stuff is off the table now. Originally the underling leader that you play as was going to be invincible (or able to respawn), but now he just expires like the rest of them (if you suck).

JS. I’d like to say we have a well defined overarching design right now, but honestly that’s a work in progress. We had a fairly aligned undocumented vision early on and I’ve been working mostly with that vision in the back of my mind. I place a lot of weight on our original vision as being the “correct” one because it was untainted by the realities of how hard game development can be. Lately the development has been more iterative and spontaneous, the switch from 2D to 3D world was based on a mockup I did and showed it to Matt. He almost immediately said “Yep, let’s do it.” As I dig more into implementing the core mechanics like movement, inventory, crafting, etc, it’s forcing us to revisit that vision and finally get the design and mechanics down on paper. I’m sure there will be more iterative and spontaneous components in the future, especially as we start getting more feedback from the community.

What are the touchstones of underVillage’s design in terms of other games? I imagine there are some obvious comparisons but it is often interesting to hear about less expected sources developers look to for inspiration.

MW. Well as long as we’re addressing the elephants in the room, I’m a big fan of games like The Binding of Isaac and Don’t Starve which are major influences. I actually wanted to use Isaac’s character customization for another game (Lucinina – Vampire Astronaut Cowgirl Princess) where the enemies are all ominously disfigured cows living on the moon (it makes sense, I promise); but it didn’t really take off. Everything I was doing there set me up for underVillage, which wouldn’t have started otherwise.
I mostly envision this game as a sort of spooky version of Animal Crossing though – the village already exists when you start up the game, the characters are unique enough that you create a bond with them as a player… but in my game a big monster can come along and smash it all to pieces. It’s a survival-horrible kind of game, but I still want the player to feel relaxed while they build their town and explore; and then terrified in the face of monsters much more powerful than they are.
Early on I wanted to include a Patapon-esk language in the game (underTongue), consisting of four letters that form words, so the whole game is presented in this weird foreign language. There may yet be some Pikmin elements involving using teamwork to move heavy objects. Late in the game breeding might become an option, with inherited personalities and body parts, like in Creatures. A wandering Smellchemist might teach you how to see the world in colourful clouds of smell, like Detective Vision from the Batman games – There’s ideas drawn in from lots of games that I’ve played and enjoyed in the past.
Most importantly is that once these ideas are on board, I have to work to make them comfortably part of underVillage’s DNA and not some tacked on feature.

JS. Matt has a much much broader exposure to games than I do, so I’m relying on his lead to define the core mechanics. Roguelike features of random world gen are key for replay value, and I’ve put forth some ideas in our design document that we have to sort out. I think Matt mentioned to me that Dungeon Keeper is another influence, especially with our use of the village fire as a key health system component. I agree with Matt that everything we do needs to feel integrated and really has to make sense with the story and the world.


How has your experience been learning an engine like Unity as you develop the game?

MW. Mostly I’ve been learning that I need a better computer to run it on. Support my Patreon.

JS. I experimented with UDK a few years back, so Unity wasn’t terribly difficult to get into. Overall I think Unity has been pretty easy to pick up, partially because there is such a vibrant community of devs out there willing to share their tips and techniques. On the flip side, there is also a lot of content out there that is really old and not applicable to Unity 5, so it’s a matter of sorting the wheat from the chaff. Interface and architecture wise, it’s a piece of cake. Other than that, simultaneously learning Unity, game development in general, C# (easiest thing I’ve done thus far) and probably too much personal pressure to keep showing progress, has made for some stressful times for me.

From reading your devlog and seeing the amount of detail contained within its entries, it is clear that a lot of work has gone into the thinking and process behind the game. What is the main draw for developing in such a forward facing way?

MW. Being open and receptive to ideas is exactly how the game first started forming – if I’d had a strict set of guidelines for what the game should be, I don’t think anybody would help get involved. It helps to keep revisiting things as we hit technical limitations.

JS. Matt and I talked about what to put in the blog early on and we decided to be pretty open and honest about the development. Initially, I was a bit worried about my inexperience in game development, and that it may scare off people who really are looking for more proven devs. He rightfully pointed out it’s better to be up front about our experience and allow the blog to show that despite our inexperience we can still make a game. Apart from the that, the main draw is that even if we’re successful or not, we’ll have a decent journal to look back at and help us see where things went right or wrong and learn from it. Personally that is what I hope to get out of it. After that , I suppose in being open about my development challenges that it may also help other new devs out there to learn or gain some inspiration from us. It’s still early days to see how it all goes, but I generally feel good about the content and we have been picking up new followers almost every week.


What kinds of things can we expect to see from your team in the coming months?

MW. I would love to have something playable for everyone to explore by the end of the year. The original goal I had in mind was a demo back in March with a mid-year Kickstarter; but I’ve had to learn to be patient since then.
Most importantly we need a musician and a business mentor both willing to invest a little time in the project; and more fans to keep us motivated.

JS. “MVP Plus” is my first goal, hopefully within the next two months or so. This will be based on our new design document and have most of the main features included. I say “MVP Plus” because we probably had a MVP back in the 2D world version: character gen, autonomous NPCs, building, crafting, and a tiny bit of combat was all there, the only thing we didn’t have is world gen. Much of that is still actually there, I just need to hook it back into the new architecture I’ve setup. Beyond that, would be a playable demo by end of year as well.

You can keep up with the development of underVillage over on their devlog or follow Matthew and Jason on twitter.

Jordan Browne

Jordan Browne

Jordan is a maker of games, music and various artistic oddities. Currently a game studies lecturer, he is has a passion for video games of all kinds, especially indie titles and works challenging the boundaries of the medium.

More Posts - Website

About the author /

Jordan Browne

Jordan is a maker of games, music and various artistic oddities. Currently a game studies lecturer, he is has a passion for video games of all kinds, especially indie titles and works challenging the boundaries of the medium.

Post your comments

Your email address will not be published. Required fields are marked *