Building New York in Marvel's Spider-Man: It's Still Just Level Design

[Music] so today i'm going to talk about my level design process this is a process that i developed while i was working on sunset overdrive and then i later applied while i was doing the initial layout for manhattan on spiderman and so i'll go through each step in the process and describe how the lessons i learned on sunset overdrive helped me while i was building manhattan so there's a little background on myself my name's josue benavidez i'm a senior designer at insomniac games and i've worked there for over 12 years now and one of the great things about working at insomniac has just been all the different games that i've got to work on so on the resistance games i worked on multiplayer and co-op and then on sunset overdrive i got to design most of the night defense bases and traps and then on spider-man i got to do the initial layout for manhattan as well as work on hero progression and traversal and something that i actually struggled with quite a bit early on was level design this became a real

challenge when i was working on sense of overdrive because i was tasked with building lots of different bases back to back without a lot of time per iteration in between and so i try to figure out how i could be a more effective level designer how i could build bases faster better and more efficiently and really i was looking for a step-by-step process particularly to help me overcome one of my biggest challenges which was this figuring out where to even begin and i looked up a lot of guides and best practices and tips and tricks but nothing that i felt was totally comprehensive nothing that could take me from this to the beginnings of a layout to bringing it into 3d and then to final and so because i couldn't find anything that could help me guide me through the whole process i started building my own process and i started developing my own step-by-step process so on the first base after i first built the first base i wrote down what worked and what didn't work and then tried to carry that forward onto the next base and then i kept iterating and kept updating this process after each base

and so by the end of the project i had a pretty solid process for how to build bases in sunset overdrive and then i moved on to spiderman and on spiderman i was tasked with doing the initial layout for manhattan and this felt like a huge task and it was almost a little overwhelming because building something like manhattan feels like it's so different from building something like a sunset overdrive base but i learned that when i felt overwhelmed the best thing to do was just to try to rely on my process and so i actually went about laying out manhattan using the same steps that i had laid out for building bases in sunset overdrive and so that's what i'm gonna go over today i'm going to go over each step in the process and talk about how how each step applied in both sunset overdrive and spiderman and so this is an overview of the whole process and it's define research experiment confirm execute and then reflect and it probably seems pretty simple and that's because it is in my opinion good design is really about just applying simple ideas in the

right order i don't really expect anything that i say today to really blow anyone's mind but at the same time along with this sort of simple step-by-step process there are also 27 sub steps so this allows us to get into a little bit more detail so step one is defined and this step is really about clearly stating your intentions for a level or for a space and in my opinion this is both the most important and yet also the most difficult step to get right because it's really hard to know what you need to do before you do it but if you can clearly define what you're trying to do up front it allows you to avoid a lot of rework and so i have a lot of techniques that i use that tried to help me define my intentions early on and so i try to break down my intentions into high level goals and i divide these into thematic goals and gameplay goals so really thinking about what are the thematic elements like the audio visual experience the emotional experience and the

narrative experience and then what are the gameplay elements what are the mechanics that are in the level what's the interactivity of the level um what are the different functions in the level and i form these goals around key questions so i ask myself what sort of story do i want to tell what emotions do i want players to feel do i want players to feel scared while they're playing through the level or do i want them to feel joyful what challenges do i want players to overcome and most importantly what makes this level special what are the lasting impressions that players are going to have what are they going to remember most and then if i'm having trouble deciding between goals i always try to pick out ideas or goals that either scare or excite me the most because those are ideas that are interesting at least to me and i think self-reflection is really important when you're building a level because the level should be fun for you if a level that's fun for you is at least one better than a level that's fun for no one and i know that i appreciate risk

and also these are your high level goals they should push you they should be trying to get you outside your comfort zone and it was actually kind of scary when we started working on manhattan and thinking about trying to build a city that was much bigger than anything we had ever built before but in the end i think i think the payoff was worth it and in a lot of ways i think we went beyond even our own expectations because we set these goals that pushed ourselves and so now that you're trying to form your goals it's also important to make sure that you're aligning your goals particularly when it comes to thinking about how your thematic goals and your game play goals interact because a lot of times you may find that you have conflicts between these goals and this is some in any of these conflicts are things that you want to try and resolve early on because later when you get further into development they can be major points of friction and cause rework this is a lesson i learned the hard way on sunset overdrive when we were building the first night defense base so i was struggling a lot with trying to make the gameplay work so night defense uh is like this tower

defense mode where you have enemies streaming in from different angles and lanes and i was struggling with making that gameplay work with my thematic elements and so i ended up just throwing any consideration for the theme away and focused solely on the gameplay and i actually did that by just sort of carving out the lanes that i wanted the enemies to take to get to the base and i got something that from a gameplay perspective i thought worked pretty well and was pretty well balanced and then when it came to trying to figure out the layout i just sort of filled in the in-betweens with geometry but didn't really think about what the actual thematic element of the base was going to be and so this put a lot of pressure on the environment team to try to realize the space and make sense of the space and in the end it didn't really make very much sense because i didn't bother i didn't have the forethought to think out what the theme was going to be and so what we ended up with was a weird mishmash of things where it's sort of like this diner at the bottom of a corporate office in the back of a neighborhood next to a shopping mall it didn't really

make sense even a game like sunset overdrive so when i moved on to the next base i wanted to make sure that my thematic and my gameplay goals were better aligned so i knew for gameplay i needed things like rails and surfaces to grind on but i also needed ways to create lots of small substructures to form lanes and so i came with the idea of brewery which would allow me to have a lot of things like pipes and bats and things to bounce on for traversal but also allow me to create lane structures better and the transition from the rough art to the final art was a much smoother process because i had worked out the theme ahead of time and so the artist and the environment team can really just focus on applying sunset overdrive style and not trying to figure out what the space actually was so on spider-man we knew we had potential conflict with some of our goals one of our goals being that we wanted the world to feel like a real place like the actual city and the other goal being we wanted the traversal to be smooth and fluid so when you're building things like a

real city you come up with lots of elements that are potential hazards for traversal and i didn't bring a laser pointer because i thought it would be a lot closer to the screen so sorry so you have important details uh for real world elements like things like billboards little uh air conditioners you've got like water towers right those things are necessary to build a real world but they become potential hazards for traversal so we made sure that we could support those things but that also whenever we included an element like that we had some traversal move or some way that the player could overcome or flow over or through them smoothly so while you're building your goals it's also really important to rank your goals you want to try to form them in a hierarchy if you can and most importantly if you can try to figure out what the most important element of your level or your space is this will help you while you're iterating on the space this will help you while you're resolving feedback but also this will help you when you're

considering spaces between each other because you might have similar goals between different spaces but you might find that the priority shifts a little bit between each space so this is the boat base on sunset overdrive and in this base the the boat theme was was really the theme that took priority so we had to pull back a little bit on some of the gameplay elements but that was okay because we knew we really wanted to have a of a base on a boat this was an idea that came from pre-production this was an idea that we thought would be really important to include in the game and so and i'm glad we did because i think it adds a good amount of variety to the total number of bases that we had so when we were thinking about the goals and defining our goals for manhattan again one of the main goals we had is we wanted it to feel like a real place and we wanted to feel like the actual city and more importantly we wanted it to be familiar to people we wanted it to be familiar to players even to players who have never actually been to manhattan who maybe are only familiar with the

city through tv and movies and what this meant was that we had to focus on things that um people had seen before in tv and movies and that meant we focused a lot on the landmarks and so and i'll just i'll go over to this side this time for fairness six and so what we have here is we want to make sure the landmarks were called out even in the rough form so you can see that there's this big tall structure thing here and even though you don't know what it is exactly you can tell that it's probably pretty important and we wanted it to stand out and also we've got another kind of little structure here too that's called out and also seems like it might be a pretty important thing and this is something that we wanted to be evident even in the rough form because when we took it to final those things ended up becoming uh important landmarks like the empire state building the chrysler building which are probably two of the most iconic buildings in the city so we want to make sure that players could see those from different areas of the city and the landmarks ended up being the actual foundation of the initial map

so we went about creating the first map by just plopping down all the different landmarks that we wanted and then including some of the roads that we knew that were going to be important things that were really iconic like park avenue and then from there we filled it in with the more generic roads and locations and this became the layout and the first map for our city so step two is research and i think for a lot of people this step seems pretty obvious of course you're going to do research on a location particularly if you're building a location like new york which is based off a real space but i include this step and i think it's an important thing to consider because i found a lot of times i was overconfident on my own knowledge of a topic i would think that i already knew a thing so i didn't need to do that much research and then when i eventually went back and actually did research i found details and things that i had missed also when you're starting on a new project or you're starting on a new level it can be really tempting to try to just jump in and start building right away

but it's important to take a step back and try to do a little bit of research because a lot of things you can learn there's a lot of actual time you can save yourself by doing the proper research first so as an example this is the first layout for the brewery base actually that i showed earlier and i was super excited to start on a new base so i just started laying things out in the uh in 2d because i figured i know what a brewery is i mean i've been to bj's they have things like pipes and bats and things i get it um and so i just started laying things out but then i realized you know this stuff's not really working out uh the layout's not vibing right things don't seem to make sense i couldn't i was struggling with why was i having such a hard time with the layout and i realized it was because i didn't do my proper research and so i ended up stepping back doing more research on breweries and then i found structures and things and more useful tools that i could use when i was laying out the space so when you're doing research it's also really important if you can try to go beyond just reference images you know places in the world

are meant for people they're meant for people to use them and a lot of times there's a lot of a lot of times the getting into the mindset of the people that are meant to inhabit the space will help inform you on the decisions on why those spaces were constructed and so if you can get if you can get an understanding for the actual people that i have in your space it helps you understand the reasoning behind that space and for spider-man i actually ended up reading a lot of travel guides for the different neighborhoods around the city and i even joined some of the online communities that were really just meant for people who lived in the city and one thing that i learned from this research was that people like to to talk about the the subway system a lot uh people who live in new york are always posting about either there's a delay or there's a shutdown or the subway's flooded it's really a major element of their everyday life and because we were trying to bring the city to life for people for our players we knew this was something that we wanted to include and even though our city doesn't have an actual functional subway network

we still try to deliver on that theme by having it as part of our our fast travel so whenever the player flash travels you get to see spider-man taking the subway and it's just a good way we can connect with people on and make the space feel more real so speaking of talking speaking of allowing research to help you do some of the work for you is you can really learn from the learn from what other games have done a lot of times something you're trying to do has been tried before and my favorite version of a prototype or first pass at a level is something that's been done in another game because you can see what sort of issues come up when you try to take something to final implementation and just because something doesn't work in another game doesn't mean it won't work in your game there might be techniques or circumstances that you can apply that will make it work but at the same time there's no reason to repeat the same mistakes exactly as another game did so one issue that i knew that we might run into from playing other spider-man

games was rooftop spaces so rooftops actually are a significant part of a city space and yet they're a big challenge for a character like spider-man because spider-man needs tall things to swing from and generally when you're on a rooftop you don't have anything to swing from so we went about trying to handle this by first making sure that whenever we have a low space right here we try to have a tall building next to it so there's something for you to swing from and then on our rooftops we also try to include lots of elements for the player to either parkour or vault over to help you cross the roof quicker and in the end we actually ended up creating a brand new traversal move called the point launch where you can actually just zip to an object on the roof and then bounce and leap off to allow you to quickly and fluidly overcome rooftops and so we ended up taking the rooftops from a problem space to something that's more more in the player space and included in the gameplay experience and then speaking of playing the other spider-man games during research i think

it's really important that you actually play the games particularly the games that relate to the content or the project that you're on i hear a lot of discussion about as developers how much should we should we be playing games and i think as a designer it's really your responsibility to play the game and then also give the game the time it needs to really understand the experience a trap that i see people fall into is they'll pick up a game and play it for maybe like an hour or two and then drop it and think that's all they need to do to get the experience but i found a lot of times mechanics or the real insights in other games don't really unlock until kind of the later half of the game and then for spider-man i actually went through and played every spider-man released on console since the first spider-man on playstation 2. and i played through them most of them all the way through and there were a lot of things that i learned in those games that i would would have missed had i dropped the games early on so again i think it's important to if you can play the games and give them the time they need

so step three is experiment and this is where you get to dig in and start trying trying ideas out but i call this step experiment because it's important that while you're at this stage you're trying to answer questions so again i think that when you're starting on a new project or a new level it's very tempting to want to just dive in and start building things that you know if you know what a thing's supposed to be you just want to make it but at this stage it's not really helpful to work on what you know it's more helpful to work on things that you don't know so you want to focus on things that you need to prove big questions that you have or major risks you want to try those things out sooner and then get those things out of the way so they don't cause conflicts later on and while you're building these experiments try to build them in a controlled way where you can control for variables and you can get results that are quantifiable if you can this is where you start building metrics and this is where you start to define things like how big should spaces be how many

characters are going to fit in there how long is it going to take the player to get through here and when you're doing these these layout these experimental layouts it's also important to not really think of them as a rough i think of them more as sketching because you're trying to get to your answers quickly and so you want to focus on just the things you need to do to answer your questions and not get too hung up on details that don't really relate to the questions you're trying to answer so as an example this is what i would consider a sketch for the first layout of manhattan and i call it a sketch because none of the building details are really accurate in fact a lot of buildings aren't even called out there's a lot of blocks that are just sort of one square per block and there are a few landmarks in there but again it's not very accurate the only thing that this really does answer is how big is the city going to be and was it going to feel like when you're inside a city this big and that was what i was trying to answer with

this i wasn't really trying to get into the details of building heights and swing metrics i was just trying to figure out what's it going to be like when we're in a city the size of the one we want to build compare this to something like this which is a little bit more accurate to what we we built in the final game so you've got buildings that are called out correctly things are at the right height this is definitely a much more detailed rough this is what i would consider a rough and not just a sketch this layout however took us weeks to get to and in fact i needed to work with a houdini expert to generate a lot of this procedurally it's not something we could get to quickly whereas the the previous the previous example this one i built in i think maybe a day two days max and so i was able to quickly get to get to answering the question of does this feel like the right size space for what we're trying to build and during experimenting it's important to try to be agile again try to work quickly try to get to your answers quickly and don't be afraid to throw things away

this is the safe space to try things that aren't going to work and then just toss them when they don't work and if you can work in physical i think it helps to bring things into the real space and bring it out of the computer so as an example hopefully you guys can see this um when we were talking about traversal right i got this little toy on a string that i was just like okay how does swinging work because our swing model is a physics-based model so i was like okay well can he just do this what does it work by doing this you know and so we would work in the physical space a little bit we were talking about traversal and then how that applies to level design um can come back to that sketching idea so i talked before about how i sketched the lanes out for that first sunset overdrive base and while not considering the theme was a bad idea um sketching things out on paper actually did was really helpful in defining the gameplay and so when i built the other bases i actually started by just sketching out all the different lanes on paper and so i could quickly get through lots of different lanes and lots of different layouts by just drawing lines and sticks and circles and then i just had to make sure that

whatever layout it came out came up with would have some kind of recognizable theme in the end and while you're also working quickly in this sort of experiment experimentation phase it's important to make sure that you're keeping the player's perspective in mind so i think one trap that's easy to fall into is when you're developing when you're developers you can kind of locked into editor space where you have top-down knowledge of the entire level but players don't have top-down knowledge players have a limited fov generally speaking and so they can only fit so much into one view and so while you're building and while you're sketching it's important to consider that what is the player going to actually be able to capture in one view and you can do this even if the game's not playable and in fact again talking about sketching sometimes i would start a layout before i even did a top-down 2d layout i would just try to draw a little perspective sketch of what a player might see standing at one point in the level just what are the different things that are called out and again it just helps you get into the mindset of the player

so our experimentation in spider-man was focused a lot on defining building metrics sizes and scales and also traversal standards and one of the first metric we worked on was the city block size we wanted to define how big where our city block's going to be were there going to be a real scale or are there going to be something scaled down and so i started by building a bunch of different city block layouts at different sizes and again trying to create experiments to control for variables i tried to use the same geometry in each experiment and making sure that the only real difference between them were the size of the city blocks and so at the top we have something that's a little bit more like what you would the size of cities that you would see city blocks you would see in the real world and in the middle is something a little bit more in between and that's what we ended up on and then another big metric that we worked on early on was trying to figure out the size of street widths and so again our swing model is based on a physics model and with that kind of a model the

smaller gap you have between swing points the more reliable swinging feels but at the same time the tighter the space you put the player in the more confined they feel so we had to try to figure out what's that sweet spot between street widths and i did that by building one almost infinitely long street and it started with very narrow streets at one end and it progressively got wider and wider and wider as you went down the street and then we would just swing up and down the street and try to figure out where was that sweet zone where did it feel right on both ends so for the city for the for the building heights we wanted to make sure that we could get a good diversity in building heights and i also wanted to make sure that we could standardize we could say that hey this area is a low building height area this area is a mid-range building height area and this area is a high range building height area but we wanted to make sure that those different height ranges would read against each other so that if we have an area like hell's kitchen which is supposed to be low height area you could tell where that is supposed to be versus an area like midtown which is a medium to high rise area so i built a

little test level where each area was called out by a different color and each area was contrasted and set next to a different height range so we can make sure that those two different heights would read well against each other and then we experimented a lot with central park too central park is a major element of manhattan and it's a major consideration when you're building manhattan because it is central if you change the shape and size of central park you actually change the shape and size of the entire city so we wanted to make sure that we got the size right we wanted to feel big and impressive but at the same time we wanted it to be manageable and so at the far right we have something that's a little bit closer to the actual size of the city and again we ended up on something that's a little bit more in the middle so at this point you've been playing with a lot of different ideas and you've been doing a lot of different experiments and you probably feel like you're ready to really start building stuff but before you go into more final implementation it's important to take a

step back and to verify that you're on the right track first thing to verify is just your goals the thing the fundamental things have been driving you so far sometimes you set a goal for yourself and you achieve it but in the end the results you get really aren't the thing that you're looking for and so this is a point in the process where you can take a step back and redefine your goals if you need to you also want to make sure that all the experiments and everything you've been working with are at the proper scale and have believable logic and believable flow you've been doing experiments on the wrong scale the results you get aren't going to be accurate when you start doing final implementation if you only have if you've only been working with a single scale clue like maybe like the player height you may not realize that things feel out of scale when you start implementing things like doorways or smaller props and you also need to make sure that it'll eventually actually run when you deploy it on your platform so again at this point you've been kind of blue skying it just trying out ideas and you may not have been taking a lot

of consideration to the technical aspect of what you're trying to build and the tricky thing about this step is that you don't always know the exact parameters and the exact technical constraints you're going to have when you do finally deploy on the platform so you need to consider whether or not you need to build in safety nets and have ways that you can scale back if you need to and you also before you go forward need to make sure that you're sharing your vision again at this point in the process you might have been working very quickly this whole up until now it could have only been a few hours and before you go through to final implementation or start doing the more expensive work you need to make sure that you're sharing what you're trying to do with everyone else and also explain your goals and try to explain the intent behind what you're trying to do if you show people what you were trying to if you tell people what you were trying to do and they see what you actually did that really helps that helps inform them on the shortcomings and if you can as much as possible only move forward with complete confidence at this point you've been working fast and you've been

working loose hopefully everything you've built so far can be thrown away and you can start over if you need to and this is the point to start over after this point things get more costly and becomes more much more expensive to do rework and so you want to make sure that you're very confident when you start doing final implementation and i also think this is the step that scales well with the project so for spider-man we had been built we had been defining a lot of metrics and a lot of standards but i wanted to be very confident that we were building the right thing and that we were going to build the right thing because we didn't have enough we didn't have a chance to try to build the entire city and then start over if we didn't build it right and the city is very different and very varied there are lots of different areas and architecture and layouts throughout the city so i wanted to make sure that metrics that we were defining for one area of the city were going to work throughout the whole city so i took our map layout that we had made i took slices from it and i composed those slices into sort of a scaled down version of the entire city

and this became our first real practical test of all of our metrics and this became our main testbed for all of our metrics and traversal standards that we have been building so after you have confidence and you feel like you're ready to go forward now is when you start to apply all the metrics and the standards you've been building now is when you start to really define the final rough you want to snap things to metric you want to clean up your edges ideally this is where the rough actually starts to look like the final game or as close to it as you can with your level of resolution and you also need to make sure that you're integrating the rest of the gameplay that you've been building so hopefully before this point you've been playing with gameplan you've been doing experiments but you need to make sure that it's all connected and that that the game plays the way you expect it to play and inevitably you're going to run into some conflicts when you start doing final when you start doing implementation of the final gameplay and so you may need to adjust and hopefully you don't have to do major rework but this is a case where you may have to

step back to the process and redefine some of your goals so one issue that we had in spider-man was the road network and the bottom half of the city is very complex and that's because this fall a lot of these roads follow most of the roads in actual manhattan and so the traffic system we had been building wasn't really built to handle this and the traffic system we were building was mostly built for four-way intersections but some of these intersections have five or even six ways but we knew this was still important to have these really complicated intersections because again we wanted it to feel like the real city and the real city has these types of things so what we ended up doing was for some of these complicated intersections we ended up kind of just blocking off a few of the roads so one road might have be under construction so there was no actual traffic going through it or one road might just be have parked cars in it and so we converted some of our six or five-way intersections back into four-way intersections so that we could achieve both the goals of building something that felt like the real city but that

would actually work with the traffic system and then at this point you get to a point in your level where you basically just continuously test and iterate on the level so you've been testing and iterating on it before this point but this becomes the entire process and this is pretty much where you take the game to final and during the testing and iterating phase you're going to come into contact with a lot of feedback and so it's important to have a plan for how you're going to resolve your feedback and this is where i like having those goals and that hierarchy of goals because this allows me to better evaluate feedback and it's important to also not be dismissive you can have a very personal attachment with your level and so sometimes you can get an emotional response to feedback but that's the advantage of comparing everything to your goals because it allows you to detach a little bit emotionally and so when you get feedback you can go okay does this feedback align with what i'm trying to do maybe this feedback is pointing to a goal that's a lower tier than a than the goal i was trying to service and so it's okay

to ignore it or maybe this feedback really does tell me i need to make some adjustments to what i was building so when we were starting to execute the final rough across the city we started to actually build the rough into more detail so once we were confident with all of our metrics and we were we knew that what we had been experimenting with was the way we wanted to go with the city we could start building the rough and start matching the buildings more to their final form so the rough started to look more like the final level and we could build those details with a little bit more confidence that the details would carry through and wouldn't be lost in the final level so the last step is reflect and even though this is the last step this was actually the first step in the process for me because this whole process started with me just thinking about how i could improve my process and how i could do better the next time an unfortunate part of this step is it

usually comes after you run out of time i find that i always have the most insight after i've run out of time to uh to act on it and so i try to learn from that and i try to carry that insight forward and do better the next time now i also have a little bit of what i call kind of a dubious personal philosophy and that even though i've run out of time i always think to myself is it really too late so because i in that reflection process i try to think to myself well is there anything that i can still act on is there anything that i can still do um to improve this and so i i include this as an important part of like trying to think about what can you do to actually improve the level in its final stages so one example of that on spider-man was something that we didn't wasn't part of the initial layout of the city was the faction bases so we have these different bases around the city that have enemies and thugs in them that

you can go through and battle in sort of a challenge type mode and we ended up implementing these kind of late in the process um we and we actually had to rework a lot of final art to get them in but we knew it was worth it because we wanted to have those spaces that would challenge players and we wanted to have spaces that would really allow players to to exercise their mastery of the game and reflecting on the whole project and thinking about what i could do better uh i do think that initially i was so focused on all the metrics that i didn't build enough space in for the environment team in some of the is to add details and to some of the street level stuff so and we actually did end up having to rework some of the street metrics so after we built prob about half of one of the first districts we realized we didn't have enough space on the streets to include the props that we needed to really make the world feel detailed enough to make it feel real and that was one of our major goals we wanted it to feel like a real space we

wanted it to feel real and so we made the decision to change the steep the street metrics uh which ended up having to do a lot of rework in that first area luckily we did this soon enough to where it wasn't detrimental to the whole project but it would have been better had we considered or had i considered that that but had it considered a better buffer zone for the propping spaces early on and we also did end up shaving off some space from the city so we actually ended up taking off some spaces from the sides but this ended up being okay because we found that if we took space off the sides it still felt right it still felt like manhattan and it didn't really disrupt the layout or didn't disrupt our landmarks very much we also considered if we could take space out of the middle of it and we found when we tried to take space out of the middle it disrupted a lot of things and made the shape not feel like manhattan and it disrupted the position of all of our landmarks

and so considering our goals of trying to make this feel like a real space we knew that we would be okay if we took we took space out of the sides but we needed to leave the space in the middle alone and i think that all the experimentation all the confirming that we did and all the planning that we did and all the research that we did really shows off when you compare the initial sketch to the final product because there are a lot of similarities between the two and so reflecting on the actual process itself during the during the applying the process on spider-man is when i really thought about experimenting with intent and so thinking about this as a scientific process and not just as free-form uh free-form experimentation but trying to actually actually answer questions and then also thinking about using evidence to support what i was trying to do so again one of the scary things about building manhattan was it's a huge level

and we didn't have the time to try it once and then restart over and try it again so we had to be sure that and confident with everything we were trying to do and we basically needed to actually predict what the outcome was going to be and i find that when you're trying to make accurate predictions it's good to have evidence so i started thinking about the experiments i did and the research i was doing as evidence when i was trying to predict how well the final level was going to come out so in summary you want to define so define your thematic and your gameplay goals form goals around key questions and resolve conflicts you may have between gameplay and thematic goals you want to do research you want to make sure you do research even if you think you're familiar with the content and you want to try and go beyond reference images and be sure you play the games and make sure that you give the games the time they need to really understand the full experience and then when you experiment try to answer key questions try to build controlled experiments think of this

phase like sketching so don't get too hung up on details yet and be okay with throwing things away being okay with starting over if you need to and then confirming you need to make sure you step back and verify your goals make sure check your scale your logic your flow and if you can only move past this step with confidence and then execute this is where you define the final rough this is where you lock down metrics you clean up your edges and everything should start to finally play like you expect it to and then finally reflect consider what you could have done better consider what about how you can improve the process and consider whether it's really too late to try to make trying to improve your level and so the main point i have is a base in sunset overdrive is a very different type of space from the city of manhattan yet the same fundamental process was used to build both and that's really the power of a good process and that's really

the power of good old-fashioned level design thank you so i think they're doing questions now if anyone has any questions there's a microphones if you have questions yeah you can go first you go first i'll go back and forth hello hi there can you hear me okay um so i'm wondering each steps how long for example like for text spiderman as example how long did you take for each steps how long for each step uh well it would have been good if i had uh recorded that um so to for example to get to one thing

i do know is to get to the initial map basically that first map where we had everything filled in for me that was about three months of work i know that because i started on the project in january and i had to get the map done before my daughter was born and she was born in march okay so it was about a three month process and then from the rest of that the process kind of blends into most of the game and sometimes you do step back and go back and forth so it's not always done in discrete chunks okay thank you so i just want to ask about the dlc when when you guys had to work when you when you finish the game and you had to start on the dlc how did you approach the level design for the dlc as well knowing that you're going to reuse the same map but change some things in it well i wasn't as involved with the dlc so i can't speak to that as specifically i do think that there are a lot of constraints that go into that though when you're

trying to uh approach spaces that are already built and that was actually some of the challenges that we had actually with the just the initial release game because the city was so big we had to basically start building it from the beginning and we ended up having to do a lot of final art before we had a chance to build all the gameplay in every area because we needed to basically just start building final art almost right away and so there were a lot of challenges that a lot of designers had because they ended up having to build in a lot of gameplay spaces that were already final and so it's always just looking for opportunities where you can build something new and looking for opportunities where you can reuse spaces that are already used okay thank you how you doing my name is josh i work on a relatively large map too it's 8k by 8k and on real units oh nice mine is 3k by 8k yeah i saw so um early on in development once the levels really started taking shape we obviously ran into performance issues it being such a big map obviously it probably came up when you

guys were building the city of manhattan like what kind of challenges did you have with performance and how did you kind of solve those issues so with i mean performance was the big consideration for us um actually part of the ak limit was beyond 8k there would be dragons i guess uh there's some sort of math issue with going beyond that and so they're like don't go bigger than that so that was the initial boundary but we did have some previous history with streaming on sunset overdrive so we had some metrics that we could build from and we knew and again talking about scaling we knew there were certain things that we could scale back and so we had idealized you know metrics that we wanted to hit and we knew there was ability to scale there and then ultimately you know one of the things we do at insomnic is just sort of rely on our awesome engine and technical teams right i don't know what those guys do i think they're wizards um but i just say hey i wanna do this thing it's gonna be insane can you do it and somehow they figure it out okay hi uh what part of the process do you

start worrying about where the story missions take place and how many collectibles there are everywhere all those sort of gameplay specifics like that i mean we were trying to map that out from the beginning we tried to make sure that each of the location i think one thing that helped that we knew that a lot of those things were probably going to pivot around landmarks too because that's makes for interesting gameplay if you have a mission or something that takes place at a well-known location right particularly in something like manhattan so we try to make sure that there was a good distribution of landmarks to help people build missions around there but you know we would map things out early so we would map things out early we would do sort of speculative mission flows throughout the throughout the city and then just try to accommodate whenever missions did come up thanks um okay uh you mentioned playing games and studying them uh to learn things from them and and play through the whole thing to learn things do you have a specific process or things that you look for or or how you

go about doing that to learn the most and and get the most out of it as a designer yeah i mean obviously it's a little bit easier when you're making something like spider-man right because you just go play all the spider-man games um but i also look for i try to find common elements or common threads or things that i uh want to achieve that either have been done well or maybe were tried in kind of a crazy way in a different game so um obviously played through all the spider-man games but then there are other games that take place in new york and so it's important to try to play those games um there are also games that are trying to go for just the feeling of new york and so i work on i work with looking at games that are achieve critical claim right that uh or are related to what i'm trying to do um or to just do something unique a novel and in terms of finding those games i guess that can be a difficult challenge right there's lots of diamonds in the rough out there and so i feel like that's why it's important just as much as you can try to do research and try to play games because what you'll find is

um by researching for one game you may see an interesting or a unique element that you can maybe remember and then you can come back and revisit that when you're doing research for another thing so it just helps us try to like as much as you can try to build that repertoire of interesting mechanics and interesting games to reference okay thank you hi um i love all the transference from sunset overdrive to spider-man that seems to be like a key theme was there anything you thought players overlooked in sunset overdrive that you were really proud to bring back in spider-man that maybe they didn't catch the first time around um i mean you know we were really we were really proud of and i guess felt like sunset overdrive was overlooked in general uh but again a lot of it just comes again with just sort of the basic principles of the design and trying to build the diverse and different layouts for the world right there are a lot of other principles of just trying to invoke one one big thing that

i was really big on with sunset overdrive was including natural elements i was really big on just adding like trees and waterfalls and anything i could anywhere um we wanted and for for manhattan i was super big on central park like i wanted central park to be this really beautiful natural space because i felt like it was important that we gave players a break from the city and so a lot of the effort that went behind making sure that central park is this map actually compared to the rest of the map a pretty massive space even though it's totally different from everything else a lot of that comes from all the desires to create natural spaces in sunset overdrive and so i think that was a big inspiration that was brought over so i thought it was really fascinating the uh what you were showing about like the the burden on on you specifically building the the way out for this space um this massive space um what uh and specifically with regards to level design um what kind of division of

labor uh came on later in the project uh to what extent did were you able to then start like breaking chunks up and when were you able to start doing that uh from the uh from your base layout state uh if you did at all you mean um the vision labor between the design side and the environment side specifically actually within within design like did you then once you got your uh core layout built did you then start dividing up districts for other designers or or was that primarily your responsibility the whole way through and then working with environment art or well that was actually the original intent was to basically try to do like a first pass layout for the city first pass like designer layout and then the other designers would come in and build the gameplay into the level unfortunately what happened was because we found we needed to start building the level so building the actual final art so quickly and a lot of the initial gameplay ended up being things more like interior spaces

a lot of designers ended up having to actually build spaces into final art instead of being able to come back and work in the look work in the layout and one thing we started doing later on um was a lot closer to the phase where we started considering bases that um i think would have been good to to include sooner if we were to rebuild the whole city again which i don't think we will uh is to try to get people to basically just build speculative space you know we did that in a few areas where we were just like we don't know but this is maybe a space for something cool it's like an arena put something there if you want right like um and so we built some of those but i feel like we could have built more of them because we ended up we ended up doing things quicker on the art side than we could keep up with on the design side thank you all right all right everyone thanks for coming