Kenshi – Rezzed 2018 Interview

Editor’s note: The yearly Rezzed event run by Eurogamer.net is a great way for indie developers to showcase their games. One title we got to look at was Kenshi, an open-world RPG with a map that reaches over 870km² and a new danger seemingly lurking around every corner. After initally getting absolutely wiped out within minutes of starting, we discovered that this isn’t like most titles around. Kenshi requires strategy, wits, and leadership skills.

How did you become a game developer and end up working on Kenshi?

I started about ten years ago, self taught, and just decided to start making it from scratch. The first version of the game was made completely single-handedly. It looked a lot rougher due to this – it was a lot smaller and less complete. I got it alpha-funded. So, before Steam Greenlight I started releasing and selling it on my own website and Desura, and then eventually got it up on Steam myself. Before then, I was working part-time as a security guard in order to get by, but now I can work on this.

You mentioned Steam Greenlight, and one of the questions I had was about Greenlight and Early Access, and how helpful they’ve been. Which one has been the best overall?

Aren’t they the same thing? Oh yeah, Greenlight was community voted.

Yeah, it was a good judge of if player demand was there.

Kenshi was actually in the first ever batch of Greenlight games ever, the first ten, along with Zomboid and a few others. Early Access is the entire source of funding for the game, so I wouldn’t be here without that.

You developed the game from scratch, how difficult is it from that perspective to create such a huge world, but then to create so many different play styles and characters that have to have their own stats accounted for?

It’s a lot of man hours, basically. We’re a small team, and a lot of it I augment with programming that helps generate things, randomly generates characters, for example. We also have a dynamic dialogue system that branches and alters dialogue semi-dynamically.

How does that work?

It starts with a standard dialogue tree that can go different ways, but there are conditions on each branch that apply to the state of the game world. For example, the target character is a woman, so you can have a different response based on the gender of the person you’re talking to. If there are enemies around, if you’re injured, any factor can affect how the dialogue works.

They act and react to the surroundings and what’s going on so they can change direction. So, bandits could be following you and assaulting you, then suddenly they can say “oh no! Police! Let’s get out of here!” and completely change the path.

It’s a massive world, it’s much bigger than say, The Witcher 3, which is most people’s mainstream experience with massive game worlds recently.

How difficult is it to create such a large world where anything can happen, while also making sure there’s enough there that it doesn’t feel barren?

It’s really difficult. Plus, as a small team we have a limited amount of assets. We can’t just churn out thousands of different creatures and models, so I have to ration them; not to be stingy, but to segment things into cultures and regions.

So, one of the areas you explored with the Hive race, that area has many different characteristics; red rain, it’s where the hivers live and you won’t find anything else there. Then elsewhere on the map you have the Holy Nation, who are more advanced than the Hivers, but they’re still kind of primitive, they’re big on religion, they’re racist against robots…

And that will affect the dialogue trees too, right?

Yeah, it affects the gameplay dynamics completely. You wander in there with a party and you have to make sure you don’t have a robot in your party, or they’ll attack you on sight. They’re also kind of sexist, so if you have a female-only party they’ll treat you differently to if you’re a male-only party.

Some zones are civilised and populated, others are wild and empty and good for exploring ruins, others have natural disasters like acid rain. I try to vary the characteristics of every zone.

It also means you have to make sure you optimise your party for each area, so there’s always that element of strategy and thinking.

Not just thinking, but learning from experience for the player, and figuring out how it works for the player in each area.

One example where that’s demonstrated is that the characters learn from failure as well as success.

You learn from both success and failure, because you want both to be fun, don’t you? You want success to be fun and you want failure to be fun. Success is more fun if you failed a lot before.

It’s more of an achievement if you fail so many times, then succeed. Going from everyone dying to everyone succeeding was a massive feel good moment for me, even in the short time we were playing.

So, you have Early Access on Steam but you also show the game here every year. How important is it to get feedback one-on-one to help build the game later on?

One-on-one honestly isn’t too much of a difference because we’re publicly released. We get thousands of people playing the game and reporting everything that’s wrong with it with feedback and bug reports. That is very valuable as it’s our main source of feedback.

We’re a small company that can’t have its own testing department, and especially with a game so big you can’t feasibly go everywhere and test everything – we need those thousands of playing it to give that feedback. It’s not just free testing, it’s on a much larger scale.

Something I’ve noticed is that there’s quite a large Wikia page for Kenshi, with a lot of contributors. Was that something that you and the team came up with, or was that the community.

That was the fans, they made the Kenshi Wikia themselves. I like that the game inspires that attitude in them. They get excited about the history of the game world and the background lore, as we’re kind of stingy with how we let people find out about it, it’s all through hints and the layout of the world.

It must be really motivating to know you have such a large community behind you already! Something else I came up against in Kenshi was prosthetic limbs, which is a really great idea for a game of this scale. Where did the idea come about, and how easy was it to implement?

It’s actually a more recent feature, we’ve just recently added it. It was originally an idea we ditched because we thought we didn’t have time and that we’d just have to live without it, but it’s been an idea since the early days of Kenshi.

You saw the complicated medical system for the body parts, so it adds a lot more to the gameplay dynamics, because every little thing you add to the design interacts with other elements, so you can create a whole new player story from your guy losing a leg and having to carry him to the Hive Village and find a leg you can afford, like what happened with you.

Rather than, because you don’t have a specific plot or main quest, these gameplay dynamics are what pull the players around on their own stories. Losing a leg itself is a story which means you need to do something without a pre-scripted “Go to John. Ask about legs”.

Do you find creating a game where those stories are present would be easier than creating something with a very strict plot where you have to go from A to B to C to D?

I don’t know, really. I think they’re completely different designs that require different skills from the designer. I wouldn’t be very good at an ABCD game myself – that’s a similar skillset to a writer or someone who writes stories, as it’s essentially a pre-written story, where my skill is designing chaos that interacts with itself.

One thing I noticed from the descriptions of the game on Steam is that the world doesn’t necessarily scale with the player, and while it was a fun for me to experience failure in the game and I learned my lessons, there may be players who do it once or twice, get frustrated, and stop playing. How do you balance Kenshi to make sure people learn organically without bouncing off because it might be too difficult in the long run?

I don’t, screw them! There’s no way to do it really; if someone really doesn’t like that balance of game then you can’t really make them like it. You let them play and see what happens. Everyone does that the first time they play; they run out of the gate and die quick, and that’s their first experience, and those players name that as what first really got them into the game. The fans of the game say “yeah, that was the moment I liked Kenshi, this is something different”. Others might say “ah, yeah, I didn’t like that”.

That was the moment I definitely fell for the game, because I can’t just do everything off the bat. It makes you want to understand the world around it, and one of the biggest triumphs of Kenshi is how it does such a great job of organically teaching how it’s played. How difficult or easy was it for you to create all the systems in such a way that it is so easy to learn organically, without having to be explicitly told that you have to do X-Y-Z in order to progress?

I don’t know, it just sort of happened! I basically made the game for me, I made it what I wanted it to be like, and it just sort of worked out well. I guess I’m instinctively good at making it work naturally, I’ve spent my life playing games so I guess I have a feel for it.

That, or I got lucky. It’s one of the two.