Level and Quest Design Collaboration from 'Skyrim' to 'Starfield'

[Music] welcome everyone to uh level and Quest design collaboration from Skyrim to Starfield we're going to be talking about the ways that different areas of design can work together your own Studio might divide up work very differently you might have a narrative Department Quest design might do something very differently from Level design level asign might be doing more Quest design everyone does it a little bit differently but if as long as your studio has two or more designers uh this talk will have something for you so I'm will Shen I'm the lead content designer uh for Something Wicked Games we are a startup uh with uh veterans and Leadership from Bethesda game studios and Obsidian Entertainment uh prior to Something Wicked Games I was a design director at Bethesda game studios I came up primarily in the Quest

design department and by the time I was leaving Bethesda game studios I was working with the next generation of design leadership there uh with uh lead systems Quest and level reporting up to me and hey everybody I'm darl Brier World director and one of the co-founders of soft Reigns we're a small Studio of 15 right now coming from places like Riot valve Sucker Punch and Sledgehammer Games we're still early in production of our first title which is a single player first person person sci-fi Adventure game where um when I was at Bethesda I was primarily a level designer during my time there but I was also lead level designer on Fallout 4 far Harbor and also Fallout 76 so darl and I have worked together over a very long career epithe the game studios for almost 15 years uh together before we both left and this talk is going to be a retrospective of a lot of our work together both with each other

but also a lot of our former colleagues uh we're going to be talking all the way back to Skyrim uh through Fallout 4 far Harbor where we were both leads Fallout 76 and finally for Starfield uh small spoiler warning there are some spoilers for the ending of Starfield so just just know that so what do we mean when we talk about Quest and level design every shop does it a little bit differently so let's go over what we mean for our talk for our purposes so this old timey Oblivion picture here is to uh sort of remind us that we are talking about the past darl and I no longer work for Bethesda game studios they may do things very differently now but in the course of our uh career there which spaned over 15 years this is what we generally meant when we talk about Quest and level design so Quest design handled dialogue and text what a lot of places call narrative design but we were also in the

booth with the voice over actors and directors to provide them with context sometimes we would be in there for the entire session we also handled a lot of these scripting specifically to make sure that quests move forward so whenever an objective appears on the screen that is usually Quest scripting telling it to progress we also worked with all of the NPCs to have them walk around to come up to you to speak to you and finally the quest design department at Bethesda game studios was also the systems design department for a long time there was no separate systems design Department until a few years before we left and typically what system design would be would be there would be Specialists Quest designers who would do the systems design zudi when it came to level design on previous Bethesda games if there was a location in a Bethesda game typically a LEL designer owned that location there can be outliers to this with places like settlements in towns that can often be owned by Quest and supported by world

artart but typically if there's a location a level designer is owning it and doing passes through it we're also ultimately in charge of the kit pieces and the logic that puts them together since this is how we build out our locations and our layouts we work really closely with World art when coming up with the the logic and the systems that Define these the kits we work uh we share responsibilities with Quest when it comes to AI Behavior our stuff often tends to be on the side of momentto moment gameplay and kind of default behaviors how does do a creature react if he's just dropped in the world with no other linkage how can we quickly set him up for patrols or wandering behaviors we are in charge of most all the combat Encounters in the game either we we own it and are doing passes on it or we had a hand in layering the combat into a space we do a lot of environmental storytelling we're ultimately in charge of how it's done in our location but we work very closely with World art when it comes to detailing out our locations to tell these stories and we do a lot of gameplay scripting whether it's setting

up uh systems to have our gameplay triggers affect other gameplay objects we'll share those amongst the rest of the design we also do a lot of unique gameplay scripting within our own locations to either add more to the story or make it more memorable in some way and when it came to collaboration between level and Quest on on previous Bethesda games level design is in charge of the location all of the combat NPCs and any layout decisions that might come up during collaboration or as Quest design is in charge of all of the narrative all of the non-combat NPCs and any story decisions that might pop up so now let's take a typical example of this kind of collaboration in action so let's go back to Skyrim specifically we're going to take a look at a quest called the house of horrors this was uh in Skyrim in the city of Markarth I worked on this quest with a level designer named D langoa and it's a very simple concept you're being asked to investigate g a house the house ends up

being haunted there is an evil presence that comes from the basement so Quest design here handled all the dialogue there is a primary character that is leading you through the house so I did all of his dialogue I also handle all of his behavior so he comes up to you and he walks you through the house and when the quest comes up and tells you that it has started and it progresses that is all the scripting on my end uh D was handling layout but also the scripting for the haunted object so all the Poltergeist effects the camera shake that was all handled by level design and I want to mention art as well we had a few unique assets for the space including the alter in the basement that also functions as a trap there is a unique animation for it and we also did a pass on the lighting and fog effects inside the space but let's take a closer look at how it was made collaboratively so the main character in The Quest is from a faction in the game

called the vigil of sendar that faction was actually originally concepted by D they were working on other Encounters in Skyrim and uh came to me and said hey I am creating a group of witch Hunters who are going to be looking for all the evil objects in the game and since we're working on this haunted house together it would be great if we tied the two of them together and I initially said no um I already have a person that I've already come up with a whole biography for and but more the more we talked about it the more it became very clear that there was an opportunity here to ground this character in the world and also to have it uh have callbacks to other encounters that you might uh run into so we worked together and I came up with the name the vigil of sendar sendar is the name of one of the Gods in Skyrim so that grounded it in the lore and I renamed the character and his backstory so that he was part of this faction

when I originally wrote up the quest I just wrote that it was going to be a haunted house and that was going to be super fun uh but that was all I had uh when we actually sat down to work on what that meant together D was the one who pitched the idea of hey we have a full physics system I could script all of the loose objects in the house and just fly them around randomly to simulate a poltergeist and then as we were working on it I came up with the idea of why don't we have the entity speak to you like a ghost uh as you're progressing through the house art also added in some storytelling bits uh when we were doing the lighting pass on the space the effects artist who was working in it was a fan of the Poltergeist movie and she suggested that we add a little Easter egg of the randomly stacked chairs inside the space uh World art also helped with some of the storytelling I originally had the idea that you get trapped in the the basement and that was just going to be a

cage or some other very simple trap and the world artist assigned to make it came to me and said I think we have an opportunity here to tie the story together by making an altar to the evil presence actually be the thing that Springs the Trap and it's tied into the Altar and again I originally was hesitant I felt like that would be a lot more work but he assured me actually it would be easier for him to make it all part of one experience so one of the big lessons I learned here was we were writing and designing the game together when I was getting a lot of suggestions from my colleagues about what we could do for this Quest I was originally very defensive you know I thought that as you know I was a new Quest designer this was one of the first quests that I had worked on and I thought if I didn't have control of everything and have everything perfect in my head that I was a failure and so I didn't realize at the time that

my colleagues were trying to give me the advantage of their expertise and perspective and once I let down that wall and started listening to them that is what made uh the content so much richer so this is a very typical example of how we made all the content in Skyrim we had a piece that we all know we needed to work on we sort of came together everyone added their bits to it and then we would move on into another piece uh but we were also experimenting with other ways to get work done so um throughout Skyrim we had uh several level designers starting to experiment with their own quests making their own quests with the quest design tools and just telling stories in that space a little more this is this leads to kind of an atypical example that I want to talk about here where in the case of this example I'm the level designer working on the location and there's a different level designer Crea the quest that ends

up layered through the location the quest is lost to the ages created by Steve cornette for Skyrim's dongu guard DLC lost to the ages has you exploring the existing land mass of Skyrim revisiting some old dungeons as well as some new ones collecting some Quest objects for the end of the quest I had been working on a location for dard as well where all of its story was kind of tied within itself and it was all visual storytelling mainly this is the location here and you you approach it and you feel some earthquakes as you're as you're approaching the exterior you continue to feel the earthquakes as you explore the interior of this location and once you get to the main um Chasm here you see that the whole Dwarven ruin has been split in half by a massive earthquake at some point in the past you explore the rest of the location wind through these Caverns fighting Falmer and other creatures until you reach the summit where there's a massive Dwarven machine that you then turn off and the earthquakes will stop at some point Steve must have played through this location or or knew about it in some way and came to me and said hey this is a really really fun uh map to play through

what do you think about if I were to add a quest npcd or location that could kind of be a guide for the player through it as well as you know explain some of the overarching loss of the ages Quest narrative and I thought that sounded like an awesome idea so Steve started to layer katria uh a ghost throughout the location she starts out as a ghostly voice when you enter the tunnels telling you to turn back that it's dangerous here once you hit the main chamber she reveals herself you see that it's a ghost you've been talking to and she motions down to her corpse and says it's dangerous here she repeats it and says if you need help I can help you through the location if you say no she'll end up at the end near that Dwarven machine but if you say Yes she'll kind of Follow You Through the location fighting Falmer and other creatures as you're making way to the summit and this made me kind of realize that we actually do have some objects in in like unique uh elements in this location that we could use some explaining to the player which would benefit us because this is a little Dwarven mechanism that when you impact it in some way it spins to the top of its pole and then opens a door unlocks

something somewhere else and we have some visual storytelling here for it but I was like hey Steve do you think we could mention this to the player through katria and kind of explain some of this and he said that that that sounded good so he he explained some of these elements to the player they didn't have a name on them so he named them kinetic resonators which I thought worked out pretty well um Steve and I uh collaborated back and forth on different elements like this throughout the location but it all comes to a head at the end where you're in front of that big dwarf machine there's five of those kinetic resonators placed throughout it you need to hit it in a certain order to solve the puzzle and then the doors will open the earthquakes will stop and this part of the quest is completed as well if you get a bad pattern you'll get hit by a trap of some kind an enemy will spawn there's plenty of visual storytelling around the location where we're explaining the bad things that can happen um but the the coolest thing that ended up happening here was um Steve layered Kat's story Into the location Story by saying she's the one who put in the wrong pattern that caused a massive earthquake and it's also why she's dead

at the beginning of the location so that that ties in really well with the story I already had for this location he also layered in a journal on deatria corpse that not only explains some lost to the ages Quest thread you can pick up to find dungeons nearby to keep the quest going but also final moments from katria where she writes about this location in her journal and gives a player some hints to starting puzzle solutions at the end so like like I had said many level designers are learning the quest tools and get more familiar with them and this starts to De develop kind of a shared language between Quest and level design and this this language deepens our collaboration going forward onto future projects so now we're moving forward to Fallout 4 and as I had said many level designers including myself have messed with the quest design tools more and more and are developing that uh shared language with the the quest design team in this example I'm going to talk about a larger scale but back to a more

typical example of quest and and level collaboration so the quest is the last Voyage of the USS Constitution created by farret Bodwin who sadly passed away during the production of Starfield faret and I were teamed up to work on this location and he came to me with a pitch of this world war one ship the USS Constitution mored on top of a bank here rather than nearby in the water where it actually is in real life near Boston and he wanted the player to be able to explore the immediate area of the bank be able to get inside the interior of the ship and meet the crew members who are robots and then get to the top deck and meet Captain Ironsides who's a Sentry bot and Captain Ironsides will ask for help repairing some elements of the ship as well as the rocket engin so he can get this ship back in the water where it belongs Farah also pitched um some scavengers attacking at the start after you had finished meeting iron sides and then periodically throughout the quest as you're returning to repair different elements of the ship this is how the whole Quest starts this

is how it's established and everything what I want to talk about now is ferit I collaboration back and forth as we implemented the elements we kind of knew we wanted in here as well as push down those elements to to add to the gamep plan and story in different ways first off we had the the different elements around the ship that you repair including the engines I scripted up a system that would just use objects from around Fallout 4 because you pick up all kinds of random junk and Fallout to to use the workshop system and craft things so I layered that into this system and it worked pretty well fret played through it and he's like I like the the repair setup but what if we made it more personal to the player where if they have a special stat on their character that could help them bypass the the need to go get the resource or expend the resources they could just repair outright and obviously that sounds like a good idea so Fric suggested a bunch of stat checks for the different stations and I layered that into the scripting and I think that was a a big benefit to the quest we didn't have a lot of those in other places in Fallout 4 also after visiting um or talking to

irons sides for the first time and kind of getting your quest infro future visits to this uh location required you to Trek all the way through the interior of the ship to get to the top to talk to the crew members again fret and I went back and forth on maybe how we could solve this one was to move the Fast Travel marker to the top of the ship but that didn't seem as in world and if you wanted to role play arriving to the ship and leaving you still had to deal with trekking all the way up to the top of the ship so I pitched this idea and then worked with World art implement this automated Robo elevator that kind of Rises you up to the top of the ship when you arrive and and takes you back down to the street level when you're leaving farret had also played through the different uh scavenger attacks and and we were really liking how those were shaping up but farret wondered if there's a way we could add a name or a face to these this faction and if there's a place nearby we could have them hole up and kind of have a little base for them and I told them there was a good spot where I'd been spawning scavengers already kind of down in an alley so I worked with World art to kind

of turn some buildings into an old Diner that the Scavengers had made into a makeshift base and then ferret went in and layered in some named NPCs some dialogue for them so you can kind of get an idea of why they're doing what they're doing why they're attacking the ship and fret kept working on this uh this little chunk of this area with the Scavengers and this ends up turning into a possible Quest path where you can actually uh sabotage iron sides in the ship and that'll ultimately result in you getting double or double cross by the Scavengers if you take that path and also World art was adding a lot more elements to the World War I ship the USS Constitution uh kind of almost last of which is these cannons that were pretty much facing right where the Scavengers are attacking everybody who played through this sequence wanted to shoot fire the cannons in some way so fret and I started talking about maybe ways we could solve this and that's when he reminded me that the end result the end reward of this Quest is actually a weapon called the broadsider and it's a small cannon you can hold that shoots small cannonballs with that reminder

that that was in the game I was able to easily set up these cannons to fire down the whole line of them when you pull the uh the circuit breaker here and then after a short coold down you can use it again then fet went in and added in an optional objective to the attacks that would remind the player that these cannons are available for use after their cool down so you have helped iron sides repair the ship you've dealt with the Scavengers he sends you over to flip one last circuit breaker from a good Vantage point and you you you pull the circuit breaker it Powers up the engines and then the players get this reward at the end and this is just a gift there's no audio but the the rear engines kick in by themselves at first and kind of drags the ship off of the bank and just before it falls into the water the front engines kick in and keep it afloat and if you're were wondering you can actually jump across the buildings to my left and actually hit a ride on this thing and just ride it over to where it ends

up you won't survive it though we had to put a kill trigger the loader couldn't keep up with how fast the player was moving so we kind of had to end it but so um faret and I's collaboration didn't end there we now have a World War one ship in the Boston skyline at the top of one of the skyscrapers you can see it now from everywhere on the map um we were trying to figure out how to get the player up there and the least amount of you know changes to the game and I ended up uh pitching this idea and kind of going with an elevator that takes you from the bottom floor to the top and it's some empty rooms up there and you can explore a little bit of them and loot some stuff after the state change of the crash that same elevator just drops you out onto the wreck and then ferret went in added some postest dialogue to the the crew robots as well as Captain Ironsides where where Captain Ironsides doubles down on the fact that this Mission has been a success and by his calculations in a mere Century will take to the Seas so the it really really benefits you to um kind of learn or develop a

shared language with the people you're collaborating with especially if they're in other departments it'll really benefit kind of the efficiency you you hook up your content and kind of the options you have once you start pushing on it and and putting it in different directions so after Fallout 4 darl and I get tapped uh to lead a DLC and so now we are not only making content but we have to plan and review content as well well so this ended up being Fallout for far Harbor and during the planning phase of far Harbor we had a very uh concise clear set of goals one of the big ones was we knew we wanted multiple endings the base game of Fallout 4 had been out for a while when we were planning this DLC and one of the big major points of feedback we got from both the press and our audience was they wanted more choices and consequences a lot of that crunchy RPG story stuff so we knew we

wanted multiple endings we also knew we wanted to add more non-combat gameplay and solutions to Quest content we also knew that people were probably going to play this DLC after they had finished the base game so we had to have high level equipment added and then we also wanted to continue the success we had in Skyrim Fallout 4 by allowing the development team to experiment more during the DLC one of those experimentation was we had three factions and we gave ownership to the story for each faction to an individual Quest designer each so a different Quest designer owned each faction and one of the major things we did was give them more say over both the story but also all the outcomes when we were planning the DLC originally I only planned for about three outcomes uh for the ending and by the time the quest designers finished up for Harbor they had added about a dozen to 15 endings depending on

which variations you count some of them were pretty small but they were able to take Choice consequences and reactivity a lot further than I thought was possible we also added in more of those stat challenges that we had originally put in for the USS Constitution Quest so we added those everywhere we also added in an entire new puzzle gameplay system uh this was uh inspired by by uh '90s early 2000s virtual reality movies and sequences so a lot of bright colors and the basic idea here was we needed to tap into and retrieve memories from an Android uh but we were able to complete an entire tower defense system within Fallout 4 in a matter of months uh by the end of the DLC another thing that happened was we had artists who wanted to experiment with design so one of our world artists Nadia hasher wanted to write the story

for the space she was working on she pitched the story to me and darl uh we approved it a great little tale she wrote all the notes but also used her art background to do illustrations to help sell the story and another big thing we did is we wanted to continue to allow level designers to experiment with Quest design and so D langoa who I had worked with on Skyrim was tapped to do both a large Quest a big side quest but also a really large location that would encapsulate the quest and this ended up being the brain dead side quest in Far Harbor uh which takes place in a vault and is a murder mystery uh involving robots so those were our plans but we experienced a lot of setbacks as we moved uh to actually make the entire DLC so now I'm going to talk about how we went about solving some project wide problems that came up during far Harbor as well as kind of things we didn't solve and what we realized after we had finished far Harbor the mistakes we had

made in some cases uh so first off the the three different faction quest lines that will had talked about ended up requiring more elaborate battle sequences than we had originally anticipated we ended up calling in an experienced level designer Steve cornette to come in and help out and Steve had not only set up kind of scripted sequences like this before but as I had mentioned in the in the atypical example before he's worked with the quest tools he's developed a shared language with the quest designers so he's able to implement the attack as well as explained the needed elements to hook into the um quest for the various Quest designers that were running the uh faction quests kind of they could communicate quickly and get this stuff done and kind of across the Finish Line in time also the the puzzles that that represented the memory sequences in the game I worked on myself along with one of our Senior Systems designer Kurt Coleman who worked on a lot of the workshop elements in Fallout 4 he and I worked on all of the puzzle elements for these these puzzle sequences and then I

went in and made all five of the puzzle levels for for the game and the the issues we ended up having with these were that puzzles could confuse the players at time at times but ultimately overstay their welcome for most players and the biggest issue was that we had made five puzzles and the first three are mandatory and the last two are optional and I took full advantage of the optional puzzles by using all of the gameplay elements to make it kind of as fun and challenging as I could i' had been playing a lot of Portal 2 at the time and making a lot of portal Maps that's also probably why a lot of this is in here but I what the issue with a player is they would end up going into the first memory sequence and after the third memory sequence we would pop we'd pop an objective for them to come out to the main you know land out of the virtual reality area but we continue to pop objectives for the remaining memory sequences and most players assume they had to complete all five of them before they even left the memory sequence and I believe most players played all of them

also our uh large location and large Quest that was taken on by a solo level designer D langoa ended up being too much for one person to take on the way we had scoped it we had kind of done what we had done on previous projects when a level designer took on you know a location and a quest or just a lot from another Silo like Quest design is just buffer some more time and this time we did it again and just buffered more time and we realized we actually needed to properly scope these get Buy in from both Quest and level production to to actually get these completed properly this ended up requiring us to have D focus on the the quest elements to get those finished up while we brought in outside level designer help to finish up the location elements the end results of far Harbor after we had shipped were that uh will and I got to work together and lead a project which was really awesome the the criticism on Far Harbor was mostly positive which was really satisfying to hear and kind of read on the internet and places and hear from people really

the our biggest source spot were the puzzles and us not properly telegraphing to the player which were which ones were kind of mandatory which were optional and our level design quests that we took on for this project needed more production Buy in from both sides to be properly scoped but in the end our our our team morale was really high and we were excited about whatever our next project was going to be the next thing I want to talk about is what happens to our collaboration and the way we work when the work starts to get really difficult so we're moving forward in time here to Fallout 76 now and the two things to keep in mind are we're scaling up internally as well as uh working with external Studios more than we had ever done before as well as taking on a multiplayer project which we also had never done before and the first thing I want to cover is the easy stuff and what I mean by that is the stuff that was at least as fast to iterate on or as fast to kind of get the content into the game as it

was in previous projects compared to previous projects so for during um Fallout 76 we were able to do basic level design almost as fast as we had done it during previous uh Bethesda projects we able to build out our worlds up to the level you kind of can see here we could add in encounters we could do all kinds of the the the things we needed to do on previous projects kind of Flesh our locations out but the biggest hurdle comes in when we need to deal with a multiplayer element if we want to play our level with another player or play through our location with a quest enabled that would require the quest to be the server to be spun up to support all of this and our iteration time typically at Bethesda during non you know non- multiplayer projects is about a minute to kind of see your changes in game which is pretty fast and we came to rely on that we could also push changes to the game that was already running so we could do some changes almost in instantaneously once we had to deal with a server we ended up dealing with um kind of iteration times it could go

anywhere from 3 minutes up to 10 minutes on a bad day if the servers you know have having issues that day but basic level design we could still build out the world and stuff without having to rely too much on the server spinning up which really benefited Us in the long run as far as building out the world there was other caveats to that that I'll bring up later we also got to start with Fallout 4 assets from the beginning this was a luxury we'd never been able to use before other than going from you know a project into DLC you can just hit the ground running and work right away so we were able to do that on Fallout 76 and that that really helped us you know the combination of hit the ground running at the beginning as well as have the iteration time to build out the world was really beneficial to US World art was also able to iterate just about as fast as we were for the same reasons we were able to get the world blocked out the landscape our kind of Define where our locations were going to go and we had all the assets to kind of block all this stuff out right away this was really beneficial to us because the the map as we had scoped for 76 was to be four times about four times the size of

Fallout 4 so all these all these benefits to World building were were really helpful to us the next thing I want to talk about is the hard stuff and that's the opposite of the easy stuff which is the stuff that just took longer on this project for various reasons that I'll start to explain as we go through the different examples first off uh the multiplayer hurdles that came up especially for level design when it came to to scripting our our locations even if you were knowledgeable with our you know proprietary scripting language from previous projects adding in the hurdles of multiplayer could really cause issues with even somebody who's used to the scripting knowing which events in which functions fire on the server versus the client was enough to kind of make some level designers not want to script anymore because of the added confusion and the iteration time just to test if they did anything proper anyway early on in in Fallout 76 I decided to script up

or add some scripting elements to location that were elements that I had done before that way the only hurdles I'd run into were the multiplayer ones and I was also trying to figure out what a interesting multiplayer you know Fallout location would be because that was a new thing for us and I settled on this um set of haunted cabins this is Alpine river cabins and you can come here alone obviously or with a group of friends or just with random players and you'll discover that there's haunted picture frames that spin randomly there's doors that will slam closed every once in a while distant screaming I'd spawn enemies randomly and Keeny players could spot um power cables that would bounce between the the cabins as well as head up into the trees and if you follow them up into the trees you'll find a treehouse overlooking the the cabins and all of the Haunted sequences are set to be automated so it's just firing through all of them and players are experiencing them you can actually turn off the the terminal sequence and then affect them yourselves so you can mess with your friends or other random players that are passing by what this

ended up kind of proving out is that the the the difficulty and the hurdles that multiplayer added to kind of our our not only our scripting and understanding that but also our iteration time and ended up limiting the number of unique gameplay experiences we could add to our you know individual locations and that was really something we did a lot in previous projects was just add a little more scripting to bring up the story some or make our location more memorable this also ended up having a uh ripple effect of level design kind of unlearning that shared language they had developed with the quest um team because of using the quest tools and doing scripting since we were doing less of that all of the stuff that was benefited from it started to dissipate also something to keep in mind um that I mentioned at the beginning of the 76 section is our scale we were kind of starting to scale up more than we ever had before and kind of as a uh to to set the pace here is uh we ship

Skyrim with a little over a 100 developers and the process we use to make this game is the one that that will and I have been kind of describing through this talk already but as you look here moving forward into Fallout 4 we didn't scale up that much more we did bring up in some outside help from Behavior interactive on things like uh our rendering Tech and some other Graphics things but our our scale didn't change that much we still ran with the same process when we developed Fallout 4 but as we move forward into 76 you can see that our scale just kind of goes to a level where we were not obviously expecting it at this point we are double what we were on Fallout 4 even if you include our exter external help we also are adding more external help from other Studios that are kind of owned by the umbrella company that owns us and also keep in mind the blue that defines bgs devs under Fallout 76 is not actually solid it's made up of four Bethesda Studios so even internally we're trying to communicate remotely and figure out

how that works and now at this level of scale it is almost it's impossible for any one person to know everybody on the team and it's also much more difficult than ever to know what who does what who you're supposed to report to so that just affects process all across the board in the end uh fallout 76 has a a rough launch we end up taking another year adding in human NPCs dialogue for those human NPCs um instance interior locations so you can actually have Quest content that can change state of uh of of the world um we end up adding all that content to the wastelanders DLC which ends up helping revive Fallout 76 so moving forward we know we have to deal with this new issue of scale and how do we work with this incredibly large team so let's talk about Starfield so by the time we Shi Starfield we have grown even more uh probably by about the same amount from

Fallout 4 to Fallout 76 as you can see here uh well over 500 people but also we are now working with a large number of external groups well over a dozen and these groups are doing everything from effects lighting they're working on tools programming systems programming pretty much everything and they have to all be integrated into the development team for the work they are doing and another part of this is these are external groups each with their own culture each with their own leadership and they are being brought in at various times throughout production sometimes early on sometimes to finish it so life becomes a lot more focused on our task tracking software we were using jira at the time jira definitely became a way of life as we tried to reconcile and handle all of the work being done across so many people and the process we ended up settling on was one that was departmental focused so we had all of

the approvals handled by a departmental lead who also had a departmental producer so for example Quest design there was a lead Quest designer I was the lead Quest designer for Starfield I had a design uh producer specific to Quest and we would go through all of the requests and proposals approve them and then send that up to a member of senior leadership who would also approve them another another thing we had to do was also start reconciling the schedule across all areas of production so every request now needed to go through all the producers because we needed to check all of the contingent work asking for something as simple as a chair wasn't so simple what were do you need animations for it do you need sound effects for it how much does that add to the schedule can it not fit because one of those teams does not have the time that all had to be wrangled and so that became a way of life for many people and then finally because there's so much work going on we needed to make sure that

there was accountability for all of that work and so we needed to create a system where all of the departmental leads were the ones providing the feedback for all the work done and also verifying that the work was completed and saying yes it's it's finished now so this had an inadvertent consequence of favoring the department over the actual project and what the project needed and this makes a lot of sense because when you are first asked to be a department lead you're worried about the department that is the thing you're told you're responsible for suddenly you have to reconcile several people's schedules you're responsible for them you're responsible for their time and their output and when your schedule is getting squeezed you want to defend it and make sure your team has the time they need and is hitting things at the Quality that it needs but this can also lead to a siloing effect where every department is scrambling for their resources and saying no to collaboration

requests so before we had the problem of hey when we're doing these big collaborations there's a lot of found work that we aren't scheduling properly now it is scheduled properly because now we know we don't have time for it so uh so another consequence of this was as we move forward was documentation became a way of life and became absolutely critical we used to be able to just sit down found someone new to the team with someone who had already been on the project and they would just learn all of the things they needed to know as kind of almost through osmosis but now we had to rely on all of our documentation tools so for example we had a shared documentation tool with Confluence we were also using slack we had to make sure and and enforce putting communication in the major slack groups not only so that it could be asynchronous but also so that everyone could see it because we needed to make sure there were no like hallway conversations that was hiding work that

was being requested and finally we had to do a lot more tutorials so I was in OBS Studio A Lot making video tutorials so on Fallout 4 when I needed to teach someone the tool set I would just sit down with them and answer their questions but now we were working with people from around the nation sometimes around the globe I often didn't know when they would need me and I often didn't know who might need me or what they would need to know so it became absolutely critical to do a lot more tutorials and how to about every uh feature of the toolkit uh another interesting consequence uh by the way this is not a real slack profile I I grabbed this from like Slack's website uh so no secrets here uh so but this is an example of a phenomenon that happened where not everyone knows you or has a relationship with you especially if they're from one of those outside groups who's coming in to help at the end so your title became

incredibly important there are many people I met because they messaged me because they saw my title and they were assigned to work on something maybe that day and didn't know who to speak to so they came to me assuming oh you well you must know something about this because you have lead in your title and it's somewhat adjacent to what I think you're working on and so you might meet someone that day who you had never met before and you're working together simply by recognition of of title instead of uh previous relationship or or knowing each other from working on the project so what does level and Quest design collaboration look like in this environment uh you know game developers hate bureaucracy we now have a ton of bureaucracy so can you continue to make memorable moments at this sort of scale with this sort of overhead so I'm going to have one last example of content uh this is the buried Temple sequence uh at the end of Starfield so I was

implementing the main quest in addition to Leading The Quest design team and this was around the time I had done the third iteration of the main quest so I had written and implemented the main quest about three times and we were finally at a state in the project where we could play through the whole thing together and it became very clear that we were missing a large final location that was going to tie the main story together and have a satisfying action filled payoff we also didn't have a big space battle at the end and this was a space game we added a new space battle system so we had to have some more elaborate sequence added there as well now I had a problem I was both implementing the main quest and leading the quest design team so I had absolutely no time the entire Quest design team was already overbooked and so we went to a trusted uh senior level designer Steve cornette uh we've talked about Steve a number of times in this talk uh he was definitely our panic button multiple times he might still be

the panic button uh now we're not sure but we knew that Steve had worked on a lot of quest design since Skyrim and so we went to Steve and said hey Quest design is overbooked we need someone who can pitch and implement the final quest of Starfield and I was on hand to help write dialogue and review the content with him so that we could get a tight iteration Loop because we didn't have a lot of time and we also made sure we assigned him a separate level designer level designer from bgs Montreal Samuel fronto handled the layout because as we learned previously asking someone to do both an incredibly large space and an incredibly large quest in a small amount of time was not possible so we knew we needed help from the beginning so Steve pitched this idea of having a dimension hopping mechanic Starfield is ultimately a story about parallel universes and he thought we could take spaces from the main story and have you hop into them

throughout the final sequence to tie everything together we also added in a final fight with an adversary including a way to skip the fight entirely using the speech challenge system and Steve repeated the dimension hopping mechanic here so that you would be fighting your adversary throughout several iconic spaces from the main quest so this was really made possible by a collaboration a very classic collaboration between old veteran designers from Skyrim and new designers who had come in more recently and the thing that made this work was necessity and urgency so as teams get large they will Silo naturally that happens but necessity and urgency also naturally breaks silos because we knew we had to have this peace in for the main quest main quest is an important piece of content it gets priority but we also know we didn't have a lot of time

so trying to do everything formally wasn't going to work we had a lot of experience with how long that bureaucracy would take and so we know we needed to suspend things so we moved mountains and shifted things around so that a small group of people could work on this very very quickly now you don't need to wait for an emergency in order to do this the important thing is providing your team with a set of product project centered goals and project centered communication always letting people know constantly and consistently hey this is what the current problem with the project is all of the departmental leads need to know that when it comes to their individual work and their individual teams we need to all collectively sacrifice for these higher level goals so that we can hit the things that we need to hit so let's talk about uh some takeaways uh number one uh right and design your game together you might have

an area of responsibility but your colleagues and your other Specialties have a lot to add to your work they have other perspectives and you do not need to be defensive or or ultimately even in control of your own content we're all along for the ride they will make the work better also remember to develop a shared language with the people you're collaborating with especially if they're in another Department you can it'll help you kind of more efficiently put your content into the game but it ends up expanding your story um telling abilities and allows you to push the content you do end up with in the game into Fun in different directions sorry uh give your team clear goals and let them Reach For It creatives really thrive in a sandbox where things are defined but there's a lot of room to play and experiment uh this will allow them to achieve things and surprise you in ways you didn't think were possible and also remember that when the

work gets difficult people will tend to do what is most comfortable to them and often times that's retreating back into their Specialties this will end up having a negative effect on any shared language that has been developed and it will also end up focusing the content of your game on the stuff that's easier to do so make sure that if you want certain content in your game make sure it's not super difficult to implement or you won't get as much as you want of it uh and finally large teams will naturally Silo and uh emergencies will naturally break them out of it but you do not need to wait for an emergency clear consistent project level goals and communication will keep everyone focused on what they need to focus on and everyone will do what's necessary uh to make sure that the project succeeds over their department or their individual teams so we want to take a moment here to thank all of the collaborators we've named in this talk and also all of the collaborators we didn't name it takes a lot of people to make these games so

thank you for everyone who has worked on us at uh on various titles over the years and we want to give a special mention to fret bodn as darl uh said we lost farret during the production of Starfield uh we had known him for over a decade uh he was a very fiercely optimistic person probably the most optimistic person we had ever worked with he brought a lot of Joy both to the content he made and to the team so thank you so much faret um thank you all for coming uh please line up for questions remember to fill out your evaluations this is our social media handles if you want to talk to us later uh thank you so much uh hi great talk thank you for being so diligent about credit in people and your collaborators um my question is

uh you mentioned a shared language multiple times throughout the talk and I think that probably looks differently depending on you know the different teams and the different Studios that those teams are working at I was wondering if you give us any examples of like specific terms or specific uh progressions of that shared language like oh when we figured out how to have this shared term that really helped us in this way something like that okay yeah so different for specific examples of shared language so I don't know I can't even remember when we developed certain terms but like for example at Bethesda game studios we had terms like trigger box and essential NPC yeah and whatever the check checkbox got named we didn't this yeah so when we were working together there's a lot of uh like for example so trigger volumes volumes where you need to know whether something is within this space or the player has entered this space we just called them trigger boxes and so I would just be like hey darl I threw down a

trigger box it's called it's called mq101 stage 50 don't move it yeah you know and uh we talk about aliases a lot that's a big part of quest so we' come up with like what they can and can't do in an alias is just a pointer to Something in in the game world that the quest can hold on to and refer to even if it's not loaded but it's all little goofy things like that which can either come from you know your studio having proper documentation so the language keep you know stays alive or the way it happened at thda was like you'd sit down and learn how to use a tool or learn from someone else and then the the shared language would just kind of develop it's not specifically terms it's just an understanding of maybe what you're asking of them when you ask them a question you kind of know how much work you're you're asking from them that's kind of what I take out of shared language yeah definitely uh hi first I I just wanted to thank you both for these games and your work like I genuinely these games and quests and levels like raised me and like like and taught me like made me who

I am in terms of like Quest design like in game development and just as a person in general like and I really appreciate that and um and coming from that I'm I'm very curious like because youve worked at this studio for a long time like and you've left now and I'm just curious like with this large amount of scale and this huge growth in the team that seems to have created a lot of stress like and a lot of like difficulty to manage like what do you think that like a large team inherently creates this or is it like the scope of the game from the beginning I'm just curious of what could have been avoided or what could be improved I'm I'm yeah okay yeah so uh do large teams inevitably create uh bureaucracy and and stress and uh could this be ass of scaling so scale and team size are are kind of the same thing especially nowadays so the bigger your game the more people you need I don't know if scale is necessarily the enemy of agility but I do think that when teams

get larger there are certain natural things that happen one of them I mentioned is siloing but another one is so there's a neurological Theory called dunbar's number that's how many social connections that you can support in your brain and it's theorized to be about 150 so once you pass 150 it's impossible to hold everyone in your head and have a connection with everyone on on the team so you need to start kind of splitting things off and I think if you're very aggressive about chunking up the team so that it doesn't get above that size you can have individual sub teams within a larger project that can then have a collaborative sort of natural relationship or another thing you could do is just be very aggressive about titles formalization and communication structure it's it's the boring part of game development but it's like flossing or eating your vegetables especially

when you have a lot of people that's what you have to do makes a lot of sense thank you very much hello uh just to Echo what the last question asker just said thank you for these games they're awesome um great talk that was a lot of fun and a great like uh trajectory through your entire history of games my question is simple um your a big part of the talk is basically cross Department collaboration allowing people to do something that isn't their regular job but something that I personally struggle with is when you get a ton of ideas or feedback or whatever from other departments that is maybe unwelcome or just isn't going to fit or whatever how do you deal with getting ideas from a bunch of people who want to collaborate and want to contribute um but that can't make it in for whatever reason oh yeah yeah so how do you deal with there's more feedback than time I mean I I I don't know it's more like you're if you're the level designer you're ultimately responsible with Bethesda the location so you kind of would assume all of the feedback you're getting is just suggestions unless it's

from somebody who is in charge of the stuff you're doing that's what it is so it's like it can be hard you can't incorporate it all and they either have to know that or or never learn it I don't know yeah yeah that's why it's it's so important to say who is ultimately responsible uh you need to be able to say okay at the end of the day you're going to have subjective calls that you have to make and and you need to know who's going to make that call so for for level level the level designer needs to make the final call when it comes to anything that affects the layout and the Encounters in the space the quest designer needs to make the final call on the narrative story and characters that's just the way that we had divided up those responsibilities but it's important to set those boundaries early awesome thank you hey thank you guys for the talk um my question kind of prongs two ways but might end up being the same answer I was surprised from your examples at how disjointed it seemed that Quest and

level design were in terms of like the level designers were working on the level independent of what it was actually going to be in the quest and it seemed to come together later on in development so my first question is uh typically when when did your guys as teams collaborate in the design process and other than I imagine it should be at the beginning of the design when should these teams collaborate yeah when when should the teams collaborate uh you feel like it it felt like uh we were working independently and then came together and that that was you know in our experience there were different levels of that so sometimes it was hey I came up with a quest I need a space for it your space looks cool I'm going to kind of take it um and then we would start working together but sometimes you would be put together from the beginning a lot of the major story lines for example because they were so big and we knew they would take up a lot of art bandwidth and other

the resource bandwidth we would put people together from the beginning um and so those spaces tended to have a lot more uh at the start but that was part of what was great about the agility of like Skyrim and Fallout 4 was I could just come in and add stuff to a space darl was working on and it would only take us maybe a couple conversations to to hammer it all out okay thank you hey thanks for the really great games over the years um uh I you said something earlier that kind of prompted a question in my mind uh you mentioned that uh the quest designers were typically responsible for the combat NPCs and or sorry the other way around the level designers were typically responsible for the combat NPCs and the quest designers were responsible for non-combat but I'm curious how that worked in games like Skyrim or Fallout 4 where those lines got very fuzzy a lot of the time cu the player could fight anyone at any time and especially if farther back in bethesda's history you go the more of that there is and uh yeah

I'm just curious how how you decided who kind of got to make calls on things like that where you had NPCs that could at any minute in these systems kind of shift between those two Dimensions yeah so how do you deal with having particular responsibilities when it's an open world and Chaos can erupt at any time I I think it it ends up depending on like what the purpose of the NPC actually is like in a town you can fight with anything and combat will erupt but an LD is still probably not likely in charge of anything in the town so like it's more about if the location and the the NPCs that are there were set up to be a combat encounter already there can be cases where like the there can be important Quest Dialogue on a boss you're going to kill we have to converse about that decide like okay before he dies he's got to get this dialogue out like there's there's some back and forth when it comes to like like little detailed stuff but typically I think it's a shared language thing where we just understand this is a combat NPC

this is a uh a quest NPC and that's the line but there's probably like rules there that we're not explaining about how we Define those hi yes thank both of you for your beautiful games and a really insightful presentation um I was wondering if you could speak a little bit to the philosophy behind having Quest designers handle the more kind of traditional narrative elements as well as like scripting and other Quest um design elements that are seen more across lots of different Studios I'm just curious what are your thoughts on that philosophy and do you see that as the kind of the future as well of Bethesda game studios oh interesting yeah so why does Bethesda game studios have the quest Department handle both narrative and and Quest scripting a lot of that was just tradition you know all a lot of these processes started way back in like moral wind and so that was a much smaller team then and you just wore a lot of hats like Quest design also did systems design for a long time and the level design Department I believe that

was really formalized end of Oblivion skyrimish it was yeah or Fallout 3ish start of Fallout 3 was like a level design team that was making stuff in that yeah yeah and so the the specialization occurred over time and there was a level design Department then eventually there was a systems design department so it I think the thing is there's no standardization in the games industry we all do it the way we need to and the teams vary so much in size that it depending on where you at you could wear different hats so I think it just evolved very naturally thank you got to wrap up oh all right it's time for us to wrap up uh we will go to the wrapup session please uh follow folks if you want to head there but otherwise thank you so much uh fill out the evaluation forms it helps out the conference so thank you [Music]