Level Design Workshop: Balancing Action and RPG in Horizon Zero Dawn Quests

[Music] hello everyone all right we're gonna go ahead and get started quick warning some things in this talk might be considered spoilers also please do not record the talk and please fill out your talk evaluation that you'll be sent over email a copy of this talk will be available on Gorillaz website after the conference so you'll be able to download this presentation if you want to see it rather than having to take pictures of every slide or something like that the title of this talk is balancing action and RPG and horizon zero dawn squares so how much is horizon an action game and how much is it an open-world RPG truthfully I can't answer that question we never drew a line in the sand and said yep this is it this is the perfect balance when I talk about finding the balance I'm gonna talk about my own experiences searching for it the open-world RPG concepts I applied during

development that worked and the action again concepts I had to learn sometimes painfully here's an outline of the talk first I'm gonna describe my background a bit and I'll that'll give you some context for what I brought to gorilla into horizon second I'm gonna walk you through an early vision of horizons quests third I'm gonna talk about open-world quests versus what we call bunker quests and do a deep dive on two examples from Horizon to illustrate the differences in technique and approach we used for each fourth I'm going to talk a bit about my own shortcomings and the things I learned from guerrillas institutional knowledge of action game level design and fifth I'm gonna try to introduce some rules for thinking about balancing action and RPG lastly I'm gonna leave a little bit of time for questions but if you want to please come out come up after the talk and ask me any questions you might have that we don't have time for hi this is me my name is blake ruby Shea

I'm a senior quest designer guerrilla games I was hired in 2015 to join a new team that the studio leadership identified would be necessary to supplement the design team that gorilla already had met new team was the quest team this team this is the quest team at guerilla games hi guys so prior to working at guerilla I was a quest designer on the Elder Scrolls online at ZeniMax online studios ESO embraces the Bethesda model of open-world RPGs like Skyrim and fallout quest designers it sauce right design and script their quests but level design is a separate discipline entirely before that I was a Laurel builder on Star Wars The Old Republic at Bioware Austin Old Republic is a take on the Bioware model of RPGs again we had separate level designers or architects at Bioware the common theme here as far as what I brought to the table is that my experience was scripting narrative design and quest design not level design I arrived this production was swinging into full gear horizon had a very strong

very compelling vision but one that didn't fully encompass the game we ended up making majestic wilderness all inspiring machines and exotic tribes these were the aspects of the game that drove our early design thinking and let's not forget a LOF course well what about quests that's what my team had been hired to make what a quest looked like in the world of horizon let me paint you a picture evening it's raining a lie wakes from unconsciousness lying in a smoking crater she stands dazed and starts moving towards a burning village in the distance elo enters the village everywhere she looks chaos reigns a villager approaches a Louie and begs her not to pursue the machine that attacked the village it's too dangerous a lie being a lie she follows the path of destruction out of the village passing the bodies of Nora warriors slain by the Beast a herd of machines blocks her path hey Lloyd

picks the Machine off from the herd in the plane and overrides it making it her mount now with a ride she makes haste to pursue the rampaging machine Ellie catches up to the hunting party sent after the Beast in the nearby mountains all but one are dead the last hunter tells a load to beware the Beast is near from nearby our quarry the mighty Thunder jaws shrieks and he engages with and defeats the machine in epic fashion pretty cool right this is an early illustrated version of the e3 trailer that we would eventually use to reveal horizon to the world this kind of quest I can understand and wrap my head around given my background at open-world RPGs a quest like this is very similar to a quest like the ice giant in The Witcher 3 big monster destroys town protagonist tracks monster protagonist kills monster I'm being reductive but you get my point this was my wheelhouse but this was not

exactly the kind of quests we ended up making this is the story I just told you there are many quests and horizons similar to this but almost half of the main story of horizon differs that difference led to a lot of lessons for me on action game design you'll notice that I've bolded and underlined all the indications of place there are no interiors in this story and indeed there were no interiors at all in the vertical slice of horizon these are the main quest chapters of horizon the quests that I bowled a request to take place at least partially in interiors 7 out of 16 more than 40% of the main story of horizon takes place at least partially in spaces we call bunkers but what is a bunker and how is the design of one inform us about balancing action an RPG as I mentioned the early vision of horizon focused a lot on exteriors machines and tribes but we needed spaces in which to tell the stories of the old world bunkers are those spaces where as

open-world quests in a horizon feature big outdoor spaces and combat against large monstrous opponents bunkers are relatively tight linear interiors featuring cover shooter combat against primarily human opponents if open world quests resemble the quests and other open world RPGs what a bunker quests resemble I'll give you a hint but more on that later too returned to the FEMA talk each of these types of quests requires a different balance of action and RPG design thinking each required a specific set of tools and experience that didn't always completely overlap to illustrate all that I'm gonna talk about two quests specifically free heap and the grave horde free was one of the first side quests I pitched and Dilton it was also a bit of an experiment as I mentioned I'd come from designing MMORPGs and so did our quest design lead at the time we wanted to see if we could build a quest

using the same kind of structure we were accustomed to that would also work for horizon it was also a bit of an experiment on how open worlds we wanted our quest to be this is what free heap looks like today but it was not always so initially we tried a very open quest concept three quests locations each essentially a pocket level connected by the open world a benefit of this type of open design is it gives players a lot of options for how they approach your quest and the challenges within it also within the context of a multiplayer game it's spread your players out and it prevents bunching unfortunately a major downside of this design is that players can bumble into the goings-on of your quest without realizing what is actually going on mini open-world RPG suffer from this problem to some degree one way we solve this problem an Elder Scrolls online was to get the quests post over to come to you whenever the player approached the vicinity of the quest some Yahoo would come running out of the distance hey I have something for you to do this just

didn't work for horizon as we developed the game it became more and more cinematic for any significant character we wanted to stage their conversations and we needed a definitive location for that case in point that was store for free he Petra I've calluses older than you I have calluses enough if you want to test me she sparks I believe it flame hair petra Forge woman uh a boy machine hunter mm machine hunter eh huh I could have a use for you straight of it is we're being muscled out of our own claim that's why I've been working on this grab spitting beast it's a dispute settler that conversation happens here in free heap and it can really only happen in free heat because of the staging the things that she references the propping that needs to be there we really wanted you to have a

context of the situation in free heap as regards the Bannock camp or the scrap or den the other locations and the quest before you went there so we employed a tried-and-true strategy we added big dam walls this is an MMO technique used to onboard open-world questing quest locations basically you limit the number of approach angles and make sure all those vectors approach lead to a quest bus tour but now with the add of the walls we start to lose the open feeling of the quest it became more and more like a linear level to combat this we focused on the level design of the transitional spaces this is an early gray box of the transitional space leading up to the bandit area because we lost so many vectors of approach to this quest location I wanted to maximize the variety of traversal options available to the player in the space we called this the jungle gym concept these early paint overs emphasize the verticality the route variety and the range of

movement options through the space if you saw Davis talk earlier he was talking about this as affordances so we really loaded this area up this is what it looks like today you can see if you have a look that there's a real variety of options presented to the player here we wanted to let them come to this bandit camp and get into it any way they wanted to you'll see plenty of zip lines you'll see tight ropes you'll see vertical options across the top you'll see low options in through the gate and tons and tons of affordances because we lost the vectors of approach we still wanted to give players a sense of flexibility this new more closed designs still offer the player a chose about which quest objective they wanted to accomplish first but that too eventually fell to the wayside again this was due to the fidelity the story we wanted to tell

this is the quest today the quest has three acts each of which correspond to a movement within the story an area and an objective the quest has bestowed here by Petra Petra needs a Leah to recover some components she will use to restore a weapon she's working on act one takes the player to the Bandit bin where the human bandits have stashed stolen Power Cells a way Briggs in takes the cells and we foreshadow retribution by the bandits in act 2 AoE enters the lair of the scrappers and grabs the parts she needs from the scrapper piles in act 3 Ellie returns the quest of a store petra to give her the collected Power Cells in parts petra uses the returned materials to restore the weapon just as the bandits arrive and attack the town so here was the rub and this is why we ended up going with a linear quest if you did the scrapper lair section first and then the bandit hideout it didn't make any sense that the bandits showed up immediately after this guy shook his fist at you and said we're gonna come

and get you it wasn't as strong a story so we eventually decided that we wanted you to go to the Banat layer first and that's just the way that we directed you to do it this narrative structure naturally suggests a looping flow so we built a special one-way shortcut from the scrapper layer back to free heap to facilitate that why because backtracking is the devil the problem was players never used it they didn't know it was there what we needed and we didn't have the time was a bread crumbing system so we built one very importantly there was a talk earlier today that talked about bread crumbing and I think this is a very good time to point out the differences between the action RPG because in action games bread crumbing literally means placing like little objects along the path to lead you somewhere when we talk about it in terms of open-world RPGs we usually mean a system like this so in world and on the compass we add these markers and we literally tell you where to go it routes you along

it routes you to doors and I'm gonna explain a little bit about how that works because it's very important in the context of an open-world game let me share with you some images from the design dock I wrote for this system three years ago as you can see I am an artist if a quest objective takes place in a space with a discrete number of entrances and the player is outside of that space the bread crumbing system will point you to a valid entrance into that space the eggs so we do that with a marker like this important point a door in this context is a crack in a wall a ladder whatever it doesn't have to literally be at door it's just an entrance into the space a level in the traditional sense doesn't have to worry about stuff like this because it is a bounded concept a level starts and a level ends and everything in the middle is the level in an open world quest you can go in go half way through decide you want to go somewhere else get something else powerup and then you have to find your way back that's why your bread crumbing system is really vital in an open world context so this is what I did

I define the scrapper layer as a discrete space with two doors like so when the player collects the parts necessary in the area I simply invalidate the door they came in through and I open the other door then the bread crumbing system takes them out simple as an experiment in open world quest design what if we heat teach us so multiplayer online RPG's the open world require you to structure requests in a way that single-player open-world RPG is don't story informs our quest structure a lot openness will often take a backseat to a great story in horizon lastly invest in player guidance systems this is extremely important will save you a lot of heartache and open-world games as a contrast I'm going to talk about the gray Ford gray Ford is a linear bunk request taking place around the middle of the main story of horizon as I mentioned before there are a lot of main story quests that take place in Bunker's

here's an illustration of what I'm talking about if you add cauldrons to that you begin to get a sense of how much linear interior space there really is and horizon this is the earliest wireframe gray box of the grave work done in Maya if free heaps flow is like a loop this sucker is a straight line compared the gray box wireframe above against the shipped version below I've added these colored areas to indicate the spaces that correlate to one another as you can see a lot of the spaces stayed they hung out between the first design in the last the macro-level design of this quest stayed pretty consistent throughout development because that's what I was familiar with and I got it pretty much right that is not true of the micro-level design quickly let me define what I'm talking about here macro level design each space is a significant moment aka a beat that's what we call it of the level as a whole the image on the left is an

in-game map of a starship level from Star Wars The Old Republic that's what I mostly worked on when I was there on the right is an image of an in-game map from the grave Ford in both cases room-service story beats or gameplay beats with the cores in between size to help the pacing of the overall level this is the type of level design familiar to RPG designers like me given my background building dungeons and Elder Scrolls online and starships and Star Wars The Old Republic so I thought that this was like directly applicable micro level design each space within the level is itself a story and must be designed with its own beats this is the type of level design familiar to action game designers more so than role-playing game designers this is also the type of level design I was least familiar with given my background this is the earliest playable grey box of the armory combat area of the gravelord the Armory's right smack dab in the middle of the level and

it's one of three combat spaces in the bunker the beat before the armory is a puzzle and the beat after the armory is kind of a story moment so it made sense structurally to do big combat beat here in the middle here's another angle for you the lines indicate the intended player path for clarity's sake the red line is kind of what I expected most players to do the purple lines were I don't know other things you could do if you were gonna be sneaky and the yellow box is where I expected the bad guys to go so feedback on this early gray box was a negative to say the least and I think you'll know why if you saw David's talk earlier this led me to one question I got bad readability overly simple uninteresting like a bad kill zone level but what is a good kill zone level I had no idea our creative director had 10 years of experience making kill zone whereas the quest team had comparatively little experience in

fact remember the team photo from earlier only one guy in the quest team had worked on kill zone the rest of us arrived to gorilla with horizon as our first project of the company I went to my boss at the time distraught and I said him I don't know if I'm the best guy in the building to do this my knowledge of action game level design is limited and there are a bunch of people who guerrilla way better at it than I am his response to me was you're right let's get you some help let's make use of some of that institutional knowledge this led to me being sent the kill zone guide for good level design TM I want to share some wisdom from a small part of the document with you and show you how I applied it so readability a very important page for me entrances exits alternative paths flanking options need to be clear to the player the player also needs to recognize where the enemies are it needs to be clear what can be used as cover and this needs to be done in conjunction

with his environment art and lighting this slide helped me realize something very important about this room and about the way I was thinking about this I was thinking too much about the content above the content of the room from a top-down perspective and I wasn't thinking enough about the way the player saw the room this is a view from the entrance to that gray box room this is the exit and the goal because this is a small cross Basin because genius I put a whole bunch of boxes in front of it it's not very clear speaking of the boxes this is where all the bad guys are supposed to be because there's so many boxes you have no idea where they could be standing these tanks are supposed to be your cover but you can't tell because they're irregularly shaped and they're much too close to the insurance to be perceived in their entirety anyway you get my point red red red bad bad bad my readability really sucked this slide was also gold for me but not in the way you might

first expect so encounters must be multi-staged routes leap layer efficiently through the environment provide opportunities for flanking keep multiple paths within the same area create variety and Heights I thought to myself I think my level has these things but then it occurred to me this this made sense to me I could do this so what I decided to do was I tried to take elements from this and put it into my level to see if they worked if this was a good kill zone level I think I could do this are you guys familiar with the term a thing we say aping when we mean that we copy something right that we that we emulate what we see other people do the funny thing is apes actually don't learn this way humans learn this way and adding elements of this to my level design taught me a whole bunch of things so I reworked my gray box to match the

example as closely as I could I kept the armory feeling with the tanks and the ammunition boxes as well as adding small sub interior rooms as offices basically I learned a lot by trying to copy the example as close as I could by aping it the new design offered multiple routes that led the player efficiently through the space the routes offered flanking opportunities and a variety of heights this is the new view from the entrance to the room much more readable and much more optimized as we moved from proper arted building blocks into proper arted building blocks we ditched the straight ramp that was into the room and replaced it with this switchback staircase for a very important reason jumping down right as you neared the room is an option but this banister is kind of a a mild barrier right in your way helps to encourage you to have a look around the space before you just go straight in guns blazing and that's what I really wanted you to do look around here's what that looks like to the player now I'm going to show you what this looks like in the finished game so

as we round the corner and we actually get our first look into the room you'll notice that the goal is clearly lit that's where you need to go as you proceed in the total area reveals itself to you and the bad guys reveal themselves to you and your options so they really owe themselves to you if you look around the room the enemy placement will help to inform you what your options are in the space also a very important tip I got from when the other designers use your pickup placement to actually show people where they can go lighting is really important but pickups are actually a great way to tell players where they can and can't go here are some still images of the finished product kudos the environment artists and the lighting team that worked on this it's an incredible result as you can clearly see the goal of this space is clearly Layton highlighted enemies

aren't hidden their placement and packing makes player options clear this space is multi-layered it has multiple paths and contains sub interiors with the collectables and goodies to find I don't know if it's a great level but I do know it's not like a bad kill zone level anymore so putting it all together what lessons can we abstract from these two level examples from horizon about finding the balance between RPG and action first these two concepts are not opposing forces I'd encourage you to consider them as lenses in the way that Jesse Schell uses that term in his book the art of game design each represents a viewpoint from which you can consider your design priorities so concept 1 macro versus micro as I've demonstrated open-world RPGs focus more on macro design whereas

action games focus more on micro design when you think about a quest macro design think about structuring your quest and service of the story when you think about a space as micro design ask yourself if the area is designed to demonstrate the challenges and the options at the players disposal concept to approach free heap shows you should try to design your quest space to funnel the player to the quest of a store and thus the beginning of the quest it is equally important as I demonstrated with grave or armory that you need to think about how players approach individual spaces not just entire areas a player's first view of room has to tell them a lot about it concept 3 movement RPGs are principally interested in players movement from space to pace and from beat to beat I explained a bit about

bread crumbing to demonstrate how manipulating the player's guidance system can ensure optimal flow through a quest space the jungle-gym concept in free heap was an attempt to capture how moving itself can be engaging and can bribe provide a feeling of openness there are many things you can do as a player to make spaces themselves more engaging more fun and more joyful case in point the big ride from the end of the gravelord so we knew through our development that we were gonna have this quest in a high place and we needed a way to get off of a mountain so I added this gigantic zipline I put this in and I expected it to get cut every single time it went to review because it's crazy but people got it when they played it they were like wow this is kind of fun it's kind of joyful and I kind of get it I'm amazed that it's in the finished product and I'm so happy that it is concepts

or final point though we may fancy what we create is art it is also entertainment to make a film reference sometimes where Wes Anderson and sometimes we're Michael Bay strive always for engagement as well as emotion for feeling as well as fun and meaning as well as explosions thank you I'd like to thank gorilla games and our publisher Sony for supporting me giving this talk I wouldn't be here today without the faith patience and gratitude that they show me I'd also like to thank my team and the rest of the designers our gorillas from whom I learned so much and now I've got some time for questions

microphones bring him on also if you'd rather you're welcome to come and find me after the talk or email me or add me on Twitter at bigger Evo whatever makes you comfortable hello there so you have a lot of experience working on mmo's and quest design there did you face any resistance from the team of gorilla when you came in and we're talking about like let's take some of these MMO learnings and put them into a single-player game where did that process look like did it feel organic to kind of take that and transition into single-player self well like I pointed out the quest team was really new and I was really lucky that I came on board along with a guy named David Ford who was the quest lead on horizon and I had worked with him on elder scrolls online I followed him to

Amsterdam so things that we wanted to try like I said the experiment with free heap was really hey David can we do a quest objective like we did on the Elder Scrolls in horizon let's see if it works and he was always open for that kind of experimentation granted not everything we tried worked out we had to break a few eggs but gorilla was really supportive and they knew that they needed the kinds of experience that we brought to the table they knew when they were making horizon that they needed people from experience and from types of games that they didn't have a lot of experience with well I think it worked thank you hi hello do you have any advice as far as making an action-adventure map and communicating with your art team to make sure that the design is still communicated when they go through with 3d modeling and make another woman yeah you know I could probably give a whole other talk on that but with her horizon we knew in the open

world that we wanted to emulate the natural world the real world so when I was building something like free heap I actually knew kind of the part of the United States where it was supposed to go and we looked to real locations in the real world for inspiration when you're building an underground space a bunker you don't have those natural touch points you don't have a natural frame of reference it's a lot like a kill zone level another thing is when you're working with really big areas if something changes between your gray box and the kind of arted product you naturally get small changes but with big areas those small changes don't matter as much they don't cause a lot of turbulence whereas if you're dealing with very small spaces small changes can change the feel a lot so you have to work very very closely with artists especially if you're working in tight areas and small areas they need to know how changing a boulder from you know moving a boulder 5 meters they need to know how important that is sure thank

you sure yes on the very macro level as far as players can do quests in any order especially the side quest you have any lessons learned for making sure things ramp up properly or whatever over the course of the game well this is a slide I wanted to add but it's interesting I think if you think about most open-world games like I think that Skyrim is a great example the game has shaped a lot like sort of like you the world and horizon is actually shaped like a funnel that gets bigger and bigger so the first couple quests of the game you're on a linear path and then the world starts to open to you and you need to be keenly aware when you're making an open-world game of when that world opens if you know when those points of broadening are you can get a much better sense of how you can design your interconnectivity does that help yeah sure I just was curious if there's any lessons that you guys learned in that sounds interesting

yeah yeah kind of going off of the environment art question what was your experience with like when working with art I mean like I'm free he being like okay we need to add some extra affordances we need to potentially limit player movement we need to like put this massive wall like he described that process and kind of like how you guys solved some of those problems whether it's like quick concept or you know stuff like that it's all about goals if you make it clear to everyone which your goals are and what you're trying to accomplish and I often say the design is like sort of being that like the Advocate in chief you know you have to be the chief advocate of a piece of content you have to make sure that people understand not only what it is but why you're doing the things that you're doing we know that we needed to get the player to the quest post or something that we did on elder scrolls is we actually had the ability to have a quest be bestowed anywhere but and on

horizon we never developed that technology it's something that we made a deliberate decision about that we wanted our conversations to be staged in very specific places so they could be really high quality when we made that decision it was easy to go the artists and say hey look we made this decision for these reasons and this is why we need giant walls that wasn't always a popular decision by no means but when you can understand it and explain it it's the natural decision I think we have time for maybe one more i can- if you'd like to you can come and ask me any more questions you might have thank you so much for coming to the talk [Applause]