Building interesting emergence in Dying Light 2

At last year's E3, Techland introduced Dying Light 2 to the world with the help of a new lead writer: Chris Avellone. The former Obsidian writer was on hand to explain that not only was he working on Dying Light 2, but the game would also be adding an incredible amount of choice-based divergence.

Instead of following a linear adventure like the first game, players will be constantly making choices that will reshape the zombie-filled city they're parkouring around in. At E3 this year, Techland showed off a demo behind closed doors that established how far these decisions will go. Small decisions can open or close quest paths in a way that will be familiar to many RPG developers, but Techland also showed off a prominent divergence that literally impacted the physical level design of Dying Light 2.

At the end of the demo, Dying Light 2 protagonist Aiden Caldwell must choose whether to open the pumps to give water to the whole city, or practice restraint and listen a the Colonel Kurtz-like NPC who has another plan for the city. It's a (complicated) moral choice, but more importantly, it's a decision that will impact the actual game geography. 

Lower the pumps, and a new section of the city becomes available, filled with a new type of zombie. Keep the pumps raised, and the player will never see that section of the city, or those enemies, in the rest of their playthrough. 

After the demo, lead game designer Tymone Smektala was on hand to chat about the game design process of Dying Light 2, and how Techland is taking inspiration from Avellone's choice-driven storytelling to make a more emergent game experience. 

What have you learned in a year as a game designer since you got to show Dying Light 2 to the world?

What I learned is how complex it is to work with narratives that are nonlinear, and with sandbox space that can have a lot of variation and variance in it. You need to support every possible combination of these too. This is really stressful, and we had to come up with some internal methodologies for how we wanted to handle this.

This is really a headache, especially if you're doing this for the first time. But thankfully we have support from Chris Avellone, who's basically a master of nonlinear narratives. He has been supporting us very much, first of us he shares his knowledge with us, he shares his experience with us, he helped us build the world that Dying Light 2 takes place in.

The city, the factions, the whole lore is either made by Chris himself or with support from Chris. But I think the most important thing we benefit from right now is the fact that he also as I said shared with his knowledge with his internal team of writers.

So he did a lot of nonlinear games where you have choices and those choices have some kind of consequences, of course these games never really offer the opportunity to change the actual game space, but still this helps out. He knows how to do it, he knows how to approach it, so it's manageable. 

You said a great word in there, methodologies. We're seeing a lot of procedural games this year at E3, would you be able to share anything about the procedural methodologies developed for Dying Light 2?

If you are pushing this in that direction, the biggest tool for us in that regard is something we are calling CityBuilder. It's not the fanciest of names...

It's a good name!

This is our internal tool that lets us create a city like you have in Dying Light 2 with a lot less effort than we would need to have if it was all done manually. There's a guy outside who is the architect of this system...our engineering programmer, he could answer more questions about it. This is our tool that allows us to build whole cities from prefabs, small pieces. 

When I say small pieces, I don't mean like building stuff, basically each part of the city, of the geometry you see, every ledge, every window, every door opening, everything is one brick that's used by CityBuilder to create various city parts which doesn't require that much input from our level designers. It understands our game, it understands what works in parkour, so it's a great tool that lets us create something quite quickly, and iterate quickly. 

Emergent gameplay is a promise that everything you do has a consequence. What have you found is a way to let your team make very emergent gameplay without it explode out of control and break your game?

So I think there are two questions in one. The first one is actually about emergence itself. We have to introduce a lot of different solutions for that. One of the biggest - and it always works - is basically physics. We have so many cool emergent moments because of the physics and how it plays with things like our AI behaviors, with our destruction models. 

For example, we had one presentation today early where in the first scene of the demo, we want you to see that emergence so we try to cut off the head of the first guy. The player did that, and the head flew off. But the head hit another guy in the face, and it broke his nose! So suddenly he started bleeding from his nose because he got hit by that head cut off by another guy. 

So that's emergence. When we talk about the number of variance not blowing out of proportions, you have to put a limit somewhere. Because the limit is either your capabilities as a developer or even the limits of the Blu-Ray disc you have to ship the game on.

We decided what's important, what kind of changes are important to us. The goal is to create a game where each player can feel that they were able to shape the city by their own decisions, and we look at things that really support that. Things that let you activate or deactivate. Reveal or destroy, heat or unheat interactive elements of the sandbox space. 

This is how we kept our head cool, let's say, and how we were able to wrap our heads around it. 

You're working with a custom engine, right?

Yes it's our in-house engine, it's called C-engine. It's something we've built purposefully for the games we're making. Techland said internally it wants to make open-world first-person games, and this is an engine that supports that with a lot of different things, how it streams data, how it streams spaces with the high fidelity of graphics you see in the scene. All of those things, we have built in the engine on purpose because we know what kind of games we want to make. 

What are some other things you found that emergent gameplay system rewards you in terms of making interesting moments for the player? How do you create a sense of triumph, or partnership with the NPCs?

When we talk about partnership with NPCs, this is an important part of our game. This might not be that visible here in the demo, but there's a theme to Dying Light 2 which asks you to look at the selfish needs of the player character and the collective needs of the NPCs around you. When you first start thinking about it, you may think "screw the NPCs, they're just computer people on my monitor, I don't really care about them, I want all of the good stuff for me as a player character." 

But there's a theme in our game where you will have to choose between the needs of your player character and the needs of those computer people. But when you finish the game, you start thinking "maybe I shouldn't be so insensitive. Maybe I should think about the NPCs more. Maybe there's something those guys are trying to tell me. Maybe it's not always about thinking about yourself. Maybe we should sometimes think about other people as well."

Is there anything in the gameplay about letting players see those consequences that you think has emerged?

This is something we knew by gut instinct, but it's clear in playtesting it's really empowering for players that they feel their decision made some kind of a change. The bigger the change, the bigger the scale of it, the bigger the feeling of empowerment.

We had a lot of people leaving our playtests saying "wow, this was amazing when I climbed to the highest building, the shape of the city, what's in it, is because of my decisions." maybe this is not a new finding but it really solidified in our heads and experience and knowledge knowing it makes a difference and knowing that it excites people and they know they can impact what's happening in the game they're playing. 

A lot of devs talk about their willingness to let the player miss content a lot. You mentioned that the player will miss 50% of the content in the game, because they're going to make binary choices and content will be locked off. As a person who makes games and loves seeing people play your stuff, how does it feel to put a lot of work of something and only a few people see it?

No! But I still believe people will see it. We had a very high number of players who completed Dying Light. So we had 50 percent of people completing the first game---

A 50 percent completion rate?

That's crazy if you compare it to other games, if you think about it. 

That's incredible. 

We had also about 35 percent of people playing co-op in a game which wasn't really about co-op, it was just a mode we added. 35 percent of people were playing co-op. We had guys really going crazy with it, the guys who played 4,500 hours. I believe they will go back to the game even when they finish it, this is something they will really experience, and this will really get them interested. 

I think also we did some artistic choices that we hope will really engage people, like the modern dark ages thing. You see things from the modern times but we'll say they're medieval-ized. This is something that really interests players, because this is kind of familiar, but also alien, so you want to explore that. So I think people would want to stay in our game for a long, long time.