Mooncrash: Resetting the Immersive Simulation

My name is Rich Wilson. Uh I'm the level design lead at Arkane Austin. I'm here to talk about uh Mooncrash. Uh more specifically um the talk is about questioning dogma. Uh it's about sort of getting the opportunity to take something um that you love and break it into pieces uh and then build something new with those pieces. Um And without further ado, here we go. All right. Um so the project um and this was a DLC for our our core game Prey uh that we launched in 2017. Um so the project was defined uh by the dynamic tension between three major forces uh that we were kind of dealing with. Uh the motivation to do something different um and a bit unorthodox to to kind of evolve our genre. Um the preservation of the studio and design team's core values um as well as testing ourselves to see which of those were worth keeping. Uh and then adaptation to the new problems that we were going to get to solve um as we incorporated all of the

new ideas um given the direction change from Prey to Mooncrash. Uh so to speak to that um a bit of a sort of digression uh a quick summary of Prey and Mooncrash just to make sure everyone's on the same page and and kind of you know build a foundation. So Prey was an immersive simulation. Uh it was kind of a hybrid first-person shooter RPG um where you had basically alien abilities and and science gadgets and tools and things like that to solve problems that came your way. So it wasn't just about shooting your way from point A to point B. Uh there were guns, but they weren't the only way to solve problems. It was set in this confined but fully explor- explorable space station um orbiting the moon. So you could travel anywhere within the station as well as without um through zero-G travel and things like that. Um this design uh was inspired in part by games like System Shock and the studio's own history with with games like Arx Fatalis, which was also inspired by Ultima Underworld. Uh we use the term space dungeon colloquially within the studio uh to kind of express this idea.

Um the objective is to escape the station. Uh you know, it's kind of a a science fiction trope um complete with this uh alien infestation uh creatures called the Typhon. Um but we added wrinkles to that so it wasn't as tropy, I guess. Um so there's a psychological element uh to the game that makes the player just kind of question their reality. Um you know, you start the game in this fake apartment. Um and upon breaking your window to try to escape the you know, the apartment, you discover you're in this simulation, this sort of uh observational situation. Um and for the rest of the game, you're kind of you know, un- unable to trust your eyes. There's this sort of a pervasive sense of unease basically that that permeates the game. There's also this looming moral framework to the game uh where you have to weigh the fates of everyone on the station against the people on Earth against your own life. Uh so it's this sort of immersive trolley problem uh with the player's hand on the lever, you know, the whole game basically. Uh so next up is Mooncrash. This was you

know, a DLC again uh expansion for the main game. Um it's a mechanical extension of Prey so it uses the same player mechanics as the core game. Uh you know, weapons, abilities, uh things like that. First person, you know, we didn't we didn't diverge too far in that regard. Uh the story was adjacent to the main game. It wasn't really a prequel or a sequel. It happened kind of alongside uh and it was set primarily within a simulation um of a of a moon base owned by the same company that owned the the space station from the original. Um but the simulation gave us a lot of wiggle room uh to play with the sort of meta structure of the game. It used a framing story for the narrative uh the main narrative delivery rather than moment-to-moment drama. Um you're a hacker diving into the simulation so your story is separated from the moment to moment game basically. Uh one of the player's primary goals while in the simulation is to escape the moon base. Uh and this is part of a list of player-directed goals uh that took the place of Prey's series of story-based objectives, and we'll talk more about this uh later as well.

Uh the game uses a roguelike framework, um which means death is incorporated as part of the core loop, um and there's a lot of variability between sessions, among other things. Uh I'm going to use roguelike for simplicity's sake. I know it's a very contested term, so don't @ me on Twitter or anything. Uh and a quick rundown on immersive sims as well. I know most of you are probably uh well-versed, but again, just to it's it's a wide-ranging subgenre, so kind of make sure everyone's on the same page. Uh so, immersive sims are they're a particular flavor of first-person shooter RPG hybrid, where we inject an extra level of simulation fidelity into the mix. Uh so, it's somewhere between, you know, the the depth of Dwarf Fortress, right? Um and the immediacy and then spatial inhabitation of Wolfenstein. So, instead of a god's-eye view and turn-based uh progression or whatever, you you're actually inside the simulation touching it with your, you know, hands. Uh so, this means there might be simulated fire propagation instead of fire just being a damage volume with a particle effect attached. Um AI may have

an organic awareness of the player and their surroundings instead of magic knowledge of where the player is at all times. Um and so, that's sort of what separates these from from other first-person shooters. All right. So, first up is the uh motivation section. Uh this is our first step uh in the path to developing Mooncrash, uh and it's kind of set the tone for everything that followed. We took some creative risks given the smaller scope of the project. We could have done something that was a standard story DLC extension. Uh we could have, you know, followed the the main antagonist from the core game, uh Commander Doll, and did like we did with Knife of Dunwall in Dishonored. Um we could have even taken the moon base that we made, the the DLC out of, and and did more story-based um A to B kind of uh progression adventure there. Uh so this section is about why we did not do that. Uh so one reason was to grow outside of our domain. So when you go from from

project to project in AAA, um a lot of the times you're solving the exact same problems every time uh with very little in the way of innovation. So you know, are we going to use lean this time? What are we mapping jump to? Are we using face buttons for the use? Stuff like that, right? Um this doesn't offer much in the way of creative growth. Um you're solving problems in the same domain, you're not getting to stretch your legs any. Uh sometimes this can even cause people to jump ship and go to other studios um between projects um and you're not really learning anything new. So with Mooncrash, this meant a kind of a cross-genre exploration. Um we explored some ideas for random level elements in Prey uh that ended up getting cut for scope. So we took that and leaned more heavily into that um and brought some of that stuff back and more roguelike elements on board um because it seemed like a natural extension of those ideas. So once we made that decision, then we could look into the roguelike space and look at their problems and their solutions and things like that. Um and uh and and learn from them basically. You know, issues like respawning and persistence, uh difficulty escalation, character variety. So games like Spelunky, uh

Rogue Legacy, and Risk of Rain, they were basically master classes in these other spheres of thought that we hadn't really explored yet. Uh and so why did we pick, you know, the roguelike genre specifically? Uh part of it was design team bias. A lot of us were playing some of these games at the time, you know, Risk of Rain, Darkest Dungeon, things like that. Um a big part of it also is the kinship between the genres. Uh so roguelikes generally rely heavily on a player's internalization of systems. Um they don't rely on rote memorization and they generally allow for expressive players in emergent situations. And a lot of those core values really speak to what we do in immersive sims as well. So there was a lot of overlap there. So another thing we wanted to do was sort of deconstruct our genre. So, you know, searching for these new problems to solve, um it afforded us the opportunity to to deconstruct our games. Um and basically take the constituent parts and spread them all out and kind of evaluate them individually, right? Um you know, we take for granted that the player can hack keypads or or crawl

through vents or things like that. Um so, isolating these parts and then recontextualizing them and and um looking at them in a new light means we could ask ourselves why we use them and why we adopted them in the first place. It's like moving between houses and then you have this, you know, this dog clock that your grandmother gave you and you can finally ask yourself why are you holding onto this thing? It's a win in either, you know, either case because you either get rid of baggage or you remind yourself why this thing was important to you and you keep it and take it along. Uh you know, some examples from from Mooncrash, character skill trees for instance. This was something we we had on the on the block and we're looking at. And we decided to keep it um and and brought it over to to Mooncrash. Um a linear sort of beat-by-beat narrative. Uh that was something we examined and decided did not belong in in the DLC, so we set it aside. Another motivating factor um was was DLC being this sort of stable launchpad. Um one of the cool things about a DLC project is that usually um

you're starting with this really extensive and refined tool set from the previous project. Uh and you've got studio expertise, um assets, and the ability to prototype things really rapidly and extensively. Um your metrics and mechanics are also in place. This is something, you know, you can't really take for granted um because it's it's very valuable. So, even if you're planning to diverge um a bit, um this starting point can really be immense for prototyping and getting, you know, getting started. Uh you know, these sort of pre-existing um assets and and engine and things like that act like a force multiplier for small teams and DLC are usually small teams. Um so, as a motivating factor again, this this kind of leveraging encouraged us to kind of flex a bit uh and try something unusual. A good example of this as well is uh we had a game jam in between, you know, the the main uh game and the DLC. Um, one purpose of this was to help the team blow off steam as people ramped off of the project and onto other things or on you know, onto uh the next project. But, um another thing this served was as an

incubator for ideas for the you know, the DLC. So, we had things like, you know, VFX-laden basketball games complete with scoring mechanisms and jump pads. Um we had this dance party involving the Typhon organisms. Um, we even had some things that made it into the to the DLC, the uh like the tentacle grenade here and the mimic companions. Uh both of those things actually made it fully into the the DLC. Uh a lot of the content wasn't ultimately used for anything that shipped, but the exercise was really valuable. Uh a lot of the team members, they got to play with development tools they wouldn't normally use. They got to kind of step outside of their their box for a bit. Um, and it also helped motivate the team to try new things. It was sort of engendering this this, you know, this spirit of of exploration. All right, so uh preserving what we do best and what we love about immersive sims was a big part of this project. It was a big deal because on other projects these values generally just tag along as a matter of course. They're just

something we take for granted. But, in this project we were explicitly calling these things into question and sort of explicitly bringing these values along um the ones we thought worked and and such. Um, we also didn't want to lose sight of what worked for us in pursuit of something new. But, we didn't want to cling too tightly to dogma out of blind habit. Uh so, there wasn't like a clean deconstruction phase that preceded actual development. Um, but as we pared things down in meetings, we gradually got a more focused picture of the values that we wanted to to as we moved forward. Uh so one worry early on was, you know, especially from the design department, um was that moving to a roguelike model would mean we were completely tossing Arkane's knack for world building and environmental storytelling out the window. Um but this was one of the dogmas that we wanted to hang on to and and because we explicitly sort of analyzed it and decided we wanted it, not just because it was something we always do. So this survived the transition remarkably intact, actually. A large part of this is because Arkane's emphasis on is usually on player story.

Uh so the bulk of our storytelling work also happens in the environment. Like a lot of our prescriptive stories, environmental stuff. And this worked really well with Mooncrash um because the game takes place in a defined fictional location. Um and the main character's job was to find out what happened there and so that that pairs very well with environmental storytelling. Uh so speaking of values, um quite a while back Chris Crawford he coined the term process intensity um describing what computers are better suited for versus what they aren't. And you can see him here at GDC '92 brandishing a sword as you do. Um and this is what he said on the subject of of process intensity. I'll drink water while people read. So the short version um is that processors are more useful with game systems than they are with game content. They're more useful with verbs than they are with nouns. Um and this is what games can flaunt that other media can't, basically.

So immersive sims embody this concept pretty pretty well. Um pretty heavily. Uh they pull content towards the process end of the spectrum, basically. Um and we leverage this more extensively in Mooncrash than we even did in Prey. Part of this was a question of economics. Um so if you look at more process intensive games, they've got a much higher consumption time to development time ratio. If you look at something like Counter-Strike or Quake, for instance, the floor, you know, the footprint of one of those levels, uh you get a lot more hours of gameplay out of one of those than a more linear narrative kinds of game. Um this isn't uh it can consumption time is not a be-all end-all of games. I'm not preaching that by any means. Um but I'm mainly talking about like development economics and and how you know, we were able to leverage that. Uh there's a whole spectrum of experiences on the the sort of process to content continuum. And that's fine. Uh process intensive games also respond more dynamically to player choice, uh which is a boon for for agency, you

know, the player's sense of having an impact on the game world. Um there are multiple tiers of agency. Um and again, this is kind of a recap of design 101, but low-level agency is stuff like footsteps sounds, bullet impact effects, you know, the enemies reacting to your presence, uh sort of base-level acknowledgement of your existence at all. Uh then you've got high-level agency, which is more narrative impact kind of stuff, um dialogue options based off of the fates of characters or, you know, different endings to the game. Uh the lower-level agency stuff is much more process intensive, uh while the higher-level stuff gets simpler and simpler on the process end, and more intense on the content end. So, the fate of the world uh can just be like a boolean flag or a an integer index that's checked by one script node uh at the end. But, on the content end, you're going to have to author, you know, four different cutscenes with voice acting and and things like that, right? Uh so, in Mooncrash, we didn't retain Prey's sense of higher-level player agency. Um that was something we we set aside. Uh you can't really affect the ending of the game. There aren't any conditional blocks of dialogue based on your actions or anything like that. Um

because monolithic moment-to-moment story doesn't play as big a role in that game. Uh it was all, you know, intentional. Uh what we did double down on, however, was the the lower-level agency and a sort of sense of intentionality uh in the player. Uh so, you know, again, speaking of of sort of core core right? Um as I mentioned earlier, one of the handy overlap between roguelikes and immersive sims, uh, and one of our core value, uh, you know, our core values are are a bias towards systems. Um, so the drama and gameplay with both generally relies on novel situations emerging from the interactions of systems, uh, rather than hand-scripted cinematic scenes. You know, the the player's story is is paramount in these games. Uh, so a quick example of of systemic design, uh, and again, this is kind of 101. Um, so let's say there's an ice wall in the game blocking the player's path, and the player has to get past it to progress through the game, right? Uh, the There we go. The non-systemic approach, uh, might be to have a variable that's tied to the player picking up a specific item in the game, uh, and then there's a

trigger in front of the wall, and when the player enters that trigger, uh, it checks that value, says yes or no or whatever, um, play some VFX, steam, plays, plays an animation as the ice melts away, things like that, right? Uh, this is really rigid, uh, it's a rigid relationship, there's a very lock and key kind of thing going on here, they're not very flexible. Um, so in immersive sims, um, you start with the concept of heat, right? Um, you know, you you this abstract concept, heat is a signal that can be passed between objects. Um, and then in that case, anything in the world can emit and respond to heat, and so we say, "Okay, the torch is a thing that emits heat." Uh, the wall is something that can respond to heat. Um, and so now, when you get close to the ice wall with the torch, um, you know, you can say, "Okay, the ice wall responds to that by melting." You can do the same animation and VFX as in the you know, the non-systemic version, um, but now you've got these systems that can more dynamically play off of one another, um, and potentially produce novel and player-driven results, uh, and

provide plenty of novel scenarios for the QA department in the process. Uh, so when I say player-driven, this means uh, the player's conceiving of ways to solve the problem rather than reading the designer's mind, uh, and that's kind of that's kind of key to our, again, our our values, you know, the things we want to preserve between projects. Uh this also means that in most immersive sims the torch it won't be the only way to bypass the blockage. Uh there's no real point in having a novel simulation-based system if you're only going to have one way to solve a problem. Uh but now that you've got heat, now that anything in the world can, you know, can generate heat, um you can have any kind of heat sources. You can have a barrel explode next to the wall um and melt it. Uh you can have, you know, a dragon that breathes fire at you and you dodge out of the way. Um you could have, you know, Channing Tatum dance next to the wall and melt it. Um so you any any kind of thing you want uh that generates heat. Um also this means the torch can also go and live a life of its own uh as a generic heat source, so it can cook food items,

it can spread through grass, it can burn enemies, etc. Uh so systemic design ties back into agency because a player with an expressive set of tools like this um in a consistently responsive world can genuinely have an impact on that world. Um the consistent rules give the player a sense of intentionality. Uh and so that means they can make informed decisions. Uh and the emergent nature of the game can provide surprises and novel results. Um so things don't get too stale or predictable and it's kind of a balancing act. Uh right? Like making sure things are interesting enough but also predictable enough. Uh and you have kind have to kind of find that the sweet spot there. Uh so one example uh of of the systemic design in Mooncrash um is that you know, the player learns there are these devices called Typhon gates and they block your passage if there are enemies nearby. Um so you can see the gate there and it tracks how many enemies are close and all that stuff.

Uh there's a an enemy called the Moon shark that patrols in levels underground. It's blind so it chases whatever vibrations it can sense in the soil. And it's also massive and really hard to kill, so the player is disinclined uh from engaging with the thing. So they want to avoid confrontations um especially in the beginning when the player is weak. Uh so you have a situation where you know the Moonshark is near the Typhon Gate and how does the player get past the gate, right? Uh so the player can put all of these kind of elements together and decide, "Okay, I'll throw a projectile really far away. The Moonshark will sense that. It will go chase it. The Typhon Gate will drop. I get to pass through." And that was all, you know, the player's sort of sh- generation of of the idea, not us saying, "Well, this is exactly what you have to do. We'll you know tutorialize it and we'll step you through all the processes, right?" This was sort of a bottom-up kind of solution on the player's part. Uh one metric we use for game objects in Arkane is how many systems interact with them. Uh so the Typhon Gates for instance interact with the AI ecosystem like in the example here. Uh but they

also interact with player progression, so if you get too many alien upgrades uh the things will actually lock down when you're nearby. Um they interact with the EMP system, so much like the heat stim you know stimulus that I mentioned earlier, EMP is also a stimulus in our game. So you can use any kind of like stun guns or exploding batteries or things like that to take them down. Uh player powers also interact with these things, so the player can burrow beneath or use an EMP uh power as well. Um and this is similar to the thought processes behind things like the Walls of Light in Dishonored and and so forth. Um but what you know it may look at first like there are a thousand different ways to get around these things, uh but the truth is in the beginning the player only has a handful of these tools like one maybe two tops in the beginning. Um and then as they gradually progress over the game, they get more and more tools to get past the things. So there's a sort of organic process where by the time the player would be getting frustrated and tired of dealing with Typhon Gates, they have all these expressive tools to get past Typhon Gates. And so it's a problem that kind of actually takes care of itself over time and it's uh this interesting

relationship. So why do we make these things? Um again, you know, looking at our values, looking at moving forward, what do we want to keep, you know, this stuff around? Um, you know, immersive sims can be a sea of chaos, systems interact in unpredictable ways. Um, organic AI awareness means you can put a basket over someone's head and then rob them blind. Um, you know, if you get too close to a quest giver and you have a torch and they have a broom, maybe their broom catches on fire and then they, you know, freak out and attack you or something like that, right? Uh, you go to pick up a key card and you kick it into a lake, you know, tons of of things, right? Um, but when you have this terrarium of systems that interact in these internally consistent ways, uh, these things begin to take on a life of their own instead of, you know, depicting something that happened to a player, you're actually building a microcosmic reality. Um, and the players that these games attract end up falling in love with them because of that sort of internally, I don't know, the sort of small world you're building.

So, a large part of Mooncrash uh, was taking these new mechanics, these sort of roguelike ideas and things like that, um, and adapting everything to this hybrid that we were creating. A lot of things we took for granted, we now had to, you know, figure out how to work with with new ideas and uh, yeah. Uh, so one of the the big things were playable characters. Um, and this is one of the systems we debated most on the project. Uh, some versions we had, you know, one playable character we were going to have that could be retooled in a lot of ways. Uh, some versions on on the drawing board, uh, we had dozens of variants on the character that were, you know, there was no real narrative veneer to them, they were just kind of buckets of mechanics. Um, and we settled on what I think ended up working best for our game and our studio. The characters had names and faces, they had personalities and backstories. Um, and this worked really well with our traditional storytelling methods. We could put a lot into the backstory and then let the abilities kind of speak for the characters, right? And let the players speak through that. Uh, so we took the all-encompassing sort

of Morgan uh, and broke it into pieces Um, and then we look for themes and play styles in those pieces. There were some that kind of naturally clumped together in this process and there were many different configurations that we ended up you know settling on the five that you see there. Um So yeah, we look for themes and play styles in those in those clumps. Um So you might have more of a tank character or someone's more of a a space wizard kind of guy. Someone's more tech inclined. And so from those templates character traits emerged then we wrote personalities to go with those. So the tank becomes VJ, you know, the engineer becomes Joan etc. And this approach yielded some really cool results that ended up solving some problems that we deal with in immersive sims without us without that being the intention going into it, right? So first we lost the the sort of choice paralysis you can get when you get a massive skill tree presented to the player like right off the bat. Each character was focused so the player can make informed decisions. You know, this character is about dealing damage so here are my options to

do that, right? And all the smaller trees they had enough wiggle room within them for the player to choose their their style but they were all heavily themed so the player is making informed decisions. And this is an example of the the Morgan over overwhelming suite of powers by the way. So there was very little overlap in these skills as well so they could complement one another. So maybe one you know, only Claire for instance had hacking so she could go open doors and then other characters could come in subsequent runs and go through those doors. VJ was combat centered so he could go clear an area of enemies and you come back as another character and and you know, have a a bit of a reprieve. So second, since all these characters are mandatory at some point or another in the campaign, we have the potential for pushing the player outside of their comfort zones. You know, through the the sort of focused skill sets of each one. One thing about RPGs and immersive sims is players tend to gravitate towards a style you know, after game after game. Like I'm the sneaky person, I'm the you know, gun haver, I'm the whatever, right? Um

and uh but by pushing other templates in front of the player, we can make them see the game from different angles. And I think that's very valuable. It can be a double-edged sword cuz some players don't want to step outside their comfort zone. It's comfortable. Um so uh but I think everything was under a small enough umbrella in our game that it worked out okay. Um yeah, we weren't making anyone do match three games or cart race or something as part of their characters' skill sets. Third, uh so by playing characters with set affinities, you get out of what I call guilt-driven gameplay. Uh this is something I really liked about Heat Signature from Tom Francis. Um you get these characters that are already leaning in a direction. Uh so they have a temperament and the associated tools um to to go in a certain direction, right? So with Dishonored or Deus Ex, you can end up playing the whole game in a very narrow fashion because like in in my head, Corvo doesn't kill people. In my head, JC didn't is never seen by anybody, right? So you kind of self-impose these these limitations and that can be that's fine, right? But you

can end up playing um in a sort of self-flagellating way where you're like, "Okay, I was seen by somebody. I'm going to reload my save. I you know, I accidentally killed somebody because I slid down a hill into the river. Um I'm going to reload my save, right? Uh even if the game isn't technically saying you failed, um you're you're again you're imposing that on yourself and you end up hating the game in the process sometimes. Um so when you're given a variety of characters, you can break out of that prison for a bit because okay, this this person is already kind of chaotic. This person is already very violent and I can I can do that through them without, you know, without sullying Corvo's good name or something like that, right? Some of the best experiences in Dishonored uh came when players would fail their sneaky approach uh and then have to sort of improvise in the messy aftermath playing like a rat-infested jazz in the process. Um those experiences were completely negated again when the player would just stop and reload. Um this also combats uh sort of ambient anxiety built up from fear of missing out. Um, if you're wondering, "Oh, well, that whole other skill tree over there, I'll never get to do. I

wonder what it's like." Um, or some players are like, "Okay, I'll play that and, you know, I'll do my my second play-through with that tree or something like that." And then they never get to it because they've got a Steam backlog of 500 games or something. Uh, so one of the first things we had to decide after choosing to go roguelike uh, with Moonlighter was what model of persistence to use in the in the characters specifically. Um, traditionally in our games death means you have to go back to the last checkpoint and try again. Death is outside of the core loop. It's, you know, it's outside basically. Uh, in roguelikes however death is an expected and integrated part of the experience instead of an interruption to that experience. Uh, they're loops uh, basically. So to adapt to this new model uh, we had to decide if we wanted the player to retain some semblance of progress between runs or if we wanted to completely wipe the slate clean. And then what kind of meta structure we wanted to use to hold the whole thing together in the process. Uh, so we looked at Spelunky and Rogue Legacy as our sort of go-to examples of of two different models, two different schools of thought.

So Spelunky is this kind of pure expression of replay. Uh, you start on the same footing every single time with the exception of shortcuts which don't really count. Um, you get this sort of stateless timeless feeling to the game. It's like checkers or chess or something like that, right? Uh, the thing that changes over time is the player's skill level and the game itself is this immutable thing. Um, and also due to the random generation of the levels you're learning how the systems interlock instead of memorizing a layout. Um, you're learning a language rather than memorizing a piece of music. Yeah. Uh, Rogue Legacy on the other hand has a progression system that persists between your trips to the castle. Uh, so you're not mechanically expected to be able to beat the game until many runs in. Some people do it but, you know, they're outliers. Um, but every death, you know, you get to bring experience points and and gold and things back uh, to give you the sort of incremental advancement uh to push you over slumps in progress. It's a really effective negative feedback system that also keeps you playing in light of defeat. And so ultimately we settled on the latter model, uh the Rogue Legacy model,

um mainly because like uh for one the the character progression in Rogue Legacy mapped really well to what we were already doing with character skill trees in Prey. Um and character skill trees are something that generally play best over multiple sessions. We we had this one model uh that we entertained where you would have the skill tree, but it would reset every single time. Uh but that ended up being right way too tedious in our heads of having to spend those points every single time you you get killed or something. Um and we also eventually decided we wanted to have an overarching story to narratively tie this game to the franchise. Um and that worked best with a sense progression doled out over multiple sessions um rather than one sort of stateless experience like Spelunky. Oop. There we go. Um so the next question is what do we want to persist in the world? So you've got character progression and world progression and and state basically you have to kind of keep track of. Uh so we wanted to keep a sense of agency in the game, um you know, the player having some kind of lasting impact on the world, but we also kind of wanted to have our cake and eat it, too. We wanted, you know, the the sort of

fresh start of a roguelike uh where you can try new things with the at the back of the without the baggage of previous mistakes. Um and we knew we'd have to reset the world at some point to make the escapes challenging and repopulate monsters and things like that. Um so we introduced the concept of resetting the simulation uh again to sort of you know, have have both. Um the conceit is that everything in the world persists until all five of your or however many characters you have, until they're all dead or have escaped. Um and then things are are reset. Uh so each character gets one life basically. So for instance, um Joan might make her way into the shuttle and escapes. Uh now she's no longer playable for the rest of that run of the simulation. Um then the player picks up VJ and goes and dies in the middle of the crater to the moon shark Um and he's no longer available and now there's like a phantom with his name on it carrying all of his stuff. I'll talk more about that later. So then the player comes back in as Claire and you know, she can either go fight the VJ phantom or she can go try to make an escape or whatever and so on until everyone's dead or has escaped.

So doing this meant we got the rogue-like consequences of permadeath sort of as well as the ability to have interplay between the characters in a meaningful player impact on the world that that one will talk about this in a second actually maybe the next slide. Yes, so this adaptation ended up as a happy compromise between the new and the old. So yeah, speaking to that you get this concept of an afterlife when you do this sort of persistence in the world. It's a legacy you're leaving for yourself. We really like the idea of the player sort of playing a co-op game with themselves in a way. And this adds a layer to the emergent nature of these kinds of games as well. So if you get the interesting combinations based on the abilities and equipment of one player, you can only imagine what you can string together if that player can leave stuff for another player another you know, character. You know, so one character might be on the way to an escape route, but they've got a lot of cool weapons and and gadgets and things like that. So they you know, if they escape with it then it's all gone,

but instead they can leave it in a chest for the other you know, character to come through and and grab or if you skirt around an enemy encounter as one character those enemies will still be there for the next character. So you have to think a few moves ahead basically. And lastly as I mentioned earlier, we added this wrinkle where characters that die come back as phantoms for other characters to deal with right. So if you die with a lot of cool stuff, you'll have to fight a monster with that cool stuff to get it back. It's much more interesting or it was more interesting to us than just picking your crap up off the ground basically. It adds this sort of dynamic piece to the board uh that you now have to deal with because of your ineptitude in a previous run. All right, so let's look at replay and game length. So, this is a diagram of a narrative-driven linear AAA game. I guess it doesn't have to be AAA, but uh narrative-driven linear game. Uh the dollar bills here represent budget

allocation in an abstract sense. Um so, this can be cutscenes, character interactions, weapon pickups, entire levels, whatever. It's It's abstract, right? Um And then this is the the player's path through that game, right? You can see most of the the dollars over the 15 to 20 hours have been hoovered up by the player um because it's all kind of laid out in a nice row in front of them. Uh then if you look at the budgets on immersive sims, uh this is more like what you get with those games, right? Uh the same those are the same, you know, the same number of dollars, um but it's all over the place. And this represents things like branches in dialogue or mutually exclusive, you quests or large levels that that are non-linear and the player can just, you know, beeline through and miss everything. Um stuff like that, right? Uh and so, this is a single playthrough, a hypothetical single playthrough. There are a lot of diehards that will eat all the content in these games, um uh but they're they're outliers. This is more

like the average kind of playthrough in one of these spaces, right? Uh so, you can see a lot of dollars are missed here. This is not a value judgment. Again, I I wholeheartedly believe in this sort of model. Um I think there's an intangible quality you get from doing this kind of thing that you can't, you know, that you can't really quantify, um that justifies the spread like that. Uh and so, here is an example of multiple paths through our hypothetical immersive sim. Uh so, you can see different player players, um they have vastly different experiences here. don't even finish it. Um but if they only play through one time, then maybe they don't know there were choices. They don't see the breadth of the whole thing um through their single play-through. Uh so we looked at this phenomenon going into Mooncrash um and the effect it has on the player's sense of agency and and choice and outcomes. So if you've got a game that's 15 hours long um and there are choices to be made in the beginning or the middle or wherever, actually. Um by the time the player finishes the game and comes back to those choices, it's possible they won't

even remember what they did the first time around or what the outcome was. Um and this is assuming they even knew there was a choice happening at all. Assumes that the player finished in the first place um uh and didn't just move on to another game. If you've got a shorter game, players are more likely to remember those choices in the short time it takes for them to get back to the choice. Um and they're also more likely to play the game multiple times to experiment with different outcomes. Uh additionally, there's a phenomenon we discovered with with player choice in games uh when we were play-testing Dishonored with people um is that uh if the choice is too subtle and organic, players don't know they even made one. Um and this can be really great for immersion. Again, this is not I'm not I'm not saying you shouldn't do this, uh but it's a phenomenon that we noticed. Um you know, we we did a a play-test with like the first hour of the game with some people um and everyone at the end were like, "Yeah, we we played this and we thought you know, it's pretty linear experience. I went and killed this person or infiltrated this place or whatever." And then the person next to them would turn around and be like, "You meant who? You went where?" And then when they compared

notes, they could see the sort of breadth of the space, but only then because everything was so organic and you know, um it was all based on on the their sort of moving through the space instead of a like a hard choice, like a big left or right or A or B or something like that. Uh so to address this uh this all these kind of ideas in Mooncrash um you know, we we adapted the more roguelike approach. Um and so we wanted to shorten the replay loops for starters. You know, a single pass through the campaign uh would actually involve several passes through the content, several loops through the content. And by a pass through the content, I mean like traveling through a few different levels to do an escape or completing a story mission or maybe just dying as the result of a plan gone wrong. This results in the player seeing a large percentage of the game and developing a developing a familiarity with the world. One problem this model can cause however and you can probably guess from the diagram here is is boredom from repetition. You see everything is kind of branching off of that one route. So

that one section of the game is going to get very stale very fast. It's the alarm clock from Groundhog Day basically over and over again. So to address that, this is the the same diagram more accurate with loops and dotted lines and stuff, but this is a lot to swallow. So um Yeah. So to counter this, we had multiple starting locations in the game. So when a new character started, they would start in a different location than the previous character. But they would be in the same level that they had died in as that previous character or escaped in. Part of this was was purely logistics. We wanted to negate load times as much as possible, especially when you're doing loopy stuff. Load times can really kill the game. But also had the side effect of keeping the player moving forward because they could basically start really close to the kind of scattered remains of their last run if they had failed. So they you know, they would pick up and knew exactly where they where they needed to go. It's worth noting that this concept can be taken too far as well. Some of our earlier models we had the player start in any level in any location, all these random spots

basically. But it got really disorienting really fast. And the players wouldn't you know, they it would take them a while to sort of get back into the context of what they needed to be doing and where to go and stuff. So even with these varying spawns though, you're going to have the potential for for boredom due to overexposure to the same content. And so Let's Let's about that. Uh so earlier I mentioned how systemic design and focusing on process intensity helps lengthen the experience for the player. You get more output for your input as a developer. The dark side of this can be a game that keeps the player playing even when the novel experiences have run dry. We dipped our toes into random content shuffling and generation so this topic became very relevant to us. The concept of infinite content it can be a facade over an endless grind basically. Um you know by the time your game stops having something new to show the player it should make a graceful exit instead of trying to compel them to stick around. This isn't an always an obvious choice as a developer either. There can be a

lot of times where you're like oh yeah I can easily generate another 3 hours of game here for the player just by doing this one little thing right? The players want more game right? Have some more pancakes player like you know. Um So there's not always malevolence behind this you know. Yeah. So instead of focusing on the amount of variability we were offering we started filtering for meaningful variability. So what could we offer that was different from session to session in a tangibly meaningful way? So in this straw man example above you know I could say this TV static offers the viewer hundreds of unique images every minute. Um it's ridiculously exaggerated but it it sort of speaks to this one exaggerated extension or this one approach. It's it's very scalable basically. But the thing about TV static is not only are the images individually meaningless to the viewer they're also individual or they're sorry they're they're meaninglessly different from one another from from frame to frame basically. A more gamey example might be a dungeon that's randomized but you're only changing the color of the tiles on the

floor or something like that. The opposite end of the of the spectrum is to hand craft everything individually and that doesn't scale very well at all. So in Mooncrash one of the ways we tackled this was through what we called hazard channels. So we took each each the four environments the four levels and we layered on top of them five different hazard states. So, you've got fire, radiation, flooding, the power can be out, and you can be oxygen deprived, right? Some of these had to be offered. You can see like the fire, for instance, requires the effects and mesh alterations and things like that. Radiation and flooding also need custom geo, but we authored these all in chunks, so then they could be mixed and matched. So, any any one particular level that was on fire could have five different versions of being on fire with different rooms being affected and things like that. Um Yeah. So, these states can then be layered on top of each other in in different combinations as well. So, you could have, you know, fire and radiation, you could have the power be out and it could be on fire, things like that. Um some of these

were mutually exclusive, so if there was no oxygen in a level, then it couldn't be on fire. And then some modified each other, so normally if there's flooding, then you have electrical hazards, but if the power was out, then you had flooding without the electrical hazards and things like that. So, you had these interesting sort of combinations. What this gave us was really strong. So, the the combination of channels gave us a lot of outputs for over a few inputs. And that was especially given the fact that it was a DLC team, small, you know, small scrappy team. The spaces could be recontextualized meaningfully for the player. So, you know, the staircase that you see here is normally perfectly fine. You can walk up it and and there's no there's no problem. But when it's on fire, now you have to decide if you want to find another way around or use your player tools or things like that, right? The spaces also look dramatically different in subsequent visits while remaining recognizable. So, this means you get visual interest over time, but since we're not re like we're not randomizing the layout of the level, you can actually learn these spaces. So, the staircase will always be the staircase, but now it's it's different

tactically and visually. Also, power outages cause there's a tram that connects levels, and so power outages can cause it to malfunction. So, it overhauls the topography of the the game every time, basically. So, you have to decide which levels or you have to learn which levels are connected to each other. Uh so, your approach from level A to B might be completely different in subsequent visits. And so, this is the same staircase um without the fire, for instance. Uh you can see here, whoop, did I skip? Yeah, okay. Here are some just brief examples of, you know, flooded and electrical um normal radiation etc. Oh, and this is, yeah, without fire, with fire. And so, we worked a lot with the granularity and frequency of the random elements within the levels.

Um we we found that if every level had a little bit of everything in it, um it was what I called the chop salad effect. Basically, you've got a lot of ingredients in the salad, but every bite tastes the same as the last one. Uh the same thing with the levels, um your your net sort of impression going from level one to level two here is going to be the same. These two levels are almost 100% different from one another. Um but like the TV static, they're not meaningfully different from one another. Um so, um what we ended up doing was sort of consolidating and clumping these random elements. Um so, now you've got, you know, one that's primarily fire and one that's primarily radiation. And so, your net impression uh from going from one to the other is is is very palpable, basically. Um and then uh let's see. Yeah. Uh this also works temporally as well as spatially. So, when you revisit these levels subsequent times and subsequent like resets of the simulation, um you you get again a a very different impression each time you you visit it instead of the sort of samey kind of kind of chop salad effect.

And this also was echoed in our enemy encounter design and several other random elements that we did over the course of the game. Uh here's an actual level layout with the same concept applied. Um you know, chopped up version and the more consolidated version. Another way to add variety uh to the game was the corruption meter. Uh so this, you know, kept the player from just kind of hanging out in the game and pushed them to accomplish their goals. Uh we ratcheted up the difficulty uh during the time they spend in the simulation, uh and we culminate in a forced reset after the the sort of corruption meter hits level five. Uh this was inspired partly by Risk of Rain's difficulty system, as well as the ghost in Spelunky. Um we liked how it makes the player choose between kind of taking their time to gather resources uh and sort of pushing them forward to accomplish their goals. It also adds a layer of variability to the experience because further down the corruption

meter you get different, you know, encounters with enemies and things like that. Uh this was one of the more contentious adaptations we made, honestly. Um a lot of people thought it kind of undercut the exploratory nature of our games. Um we mitigated it some with the with these items that you can get that reset the clock or or wind it back some. Um but at the end of the day, it was something some people, you know, some people didn't like. It was a It was an intentional design decision and uh yeah. Uh one of the larger departures from Prey was the concept of narrative progression in Mooncrash. Um so the game primarily consists of a series of orders uh from the main character's employer. Uh the larger the percent of items checked off the list, the farther along the framing story progresses, basically. Uh but the player can take these things in any order. There are some that are, you know, that have a sort of a dependency chain, but otherwise you can take them in whatever order you want. Uh the actual items are more than just fetch-questy stuff, as well. Some of them are entire story missions or or elaborate escape methods or things like that. Um the point of it is to allow the player to proceed at their own pace and in the

order that they choose. Um this opens the door for different strategies as well and and caters to different play styles. So, some players might want to save big stuff for last or focus on one character at a time or go on a collecting spree or things like that. This is also a win for us because we could decouple the player's actions from the prescribed narrative thrust of the game. Uh, we didn't have to deal with problems like, "Well, what if the player goes over here and gets gets this thing before they find this other thing?" Or, "Once they've done X, what is their motivation uh to then proceed to Y?" Uh, that was a big deal on Prey because the entire, you know, game was a sort of continuous string of of cause and effect. Um, but this this sort of unasked those questions. And here you can see a brief example of the elements that change over time as the player finishes items from the list. Uh, so you get more hazards introduced to the simulation. Uh, the corruption meter gets faster um over time. Uh, the player's mastery of the space is basically tested more and more the further along they get to the end of the the campaign.

And finally, the escapes. Um, so yeah, one of the first drafts of Mooncrash was basically using Talos from the core game and just adding, like, opening the whole thing up and adding random escape methods to it. Partly inspired by this game called Monstrum where you're on this freight ship escaping an alien or a sorry, a monster that's uh that's escaped and is with you. Um, so the Talos version was a much lighter weight version of what we ended up doing uh within Mooncrash, but the overarching concept of of the sort of escape uh the randomized escape remained interesting to us. Uh, so the base in Mooncrash has five different ways to escape. Um, each one is mechanically and thematically different from the other. Uh, so there are a few examples. Um, one is a shuttle that requires a character to install a piloting skill. Um, but you have to find the random location of the last person to install that skill using the personnel locating system, basically. Um, and then that skill persists for whatever character installs it. So, that person can now use that escape method um once they've got that

skill. One of them is a cargo container you can stowaway in um and get blasted by rail gun to Earth basically. Uh, but first you've got to scavenge uh scavenge for a randomized list of of supplies basically so you can survive the trip to Earth. Uh, one of them is an experimental portal but you need a hacking and repair which are separated into two different characters. So, those two characters have to work together to enable this escape method to even function. Uh, so these differences kept the game interesting by varying up the types of task between the escapes. Um, in addition to all the other world variables that you have to deal with in a given session. Um, the differences also add very deliberate choices for the player. So, maybe Joan has the piloting skill installed from a previous escape um but you need to use her for half of the portal before sending her off in the shuttle or something. Um, maybe you found some supplies for the crate as Riley but the only way you can use the brain upload escape is with her. Um, so you have to find somewhere to stash those supplies. You know, there there are lots of choices basically the player

has to kind of think up or deal with. And not only do you have to decide which escape route is the most opportune for the character you're currently playing but you have to consider characters down the line. Um, and again none of this is prescriptive. This is all up to the player to determine the best strategy. We don't say you have to do this in this order or whatever. Uh, so the most elaborate list item uh is the five-person escape. Uh, you have to escape in all five different escape routes with all five characters in one run without the simulation resetting. Um, we kind of we wanted to incorporate this um because it turns the game into this puzzle box basically. Uh, where you have to consider things like the strengths of the characters, the complexity of the escapes, and the order that you want to perform them in, uh and the corruption meter because again the difficulty difficulty is ratcheting up over time. Uh, so then the player has to weigh things like should I get an easy escape like the escape pods out of the way earlier or save it for later cuz I'll need an easy win um or I could escape with this character right now, but they could do things in the world to help other characters with other escapes down the line. Now, the speed running community has

illuminated the optimal path by this point, but um for regular players, for regular human beings, um there are there are a lot of interesting variables to sort out, um and you know, in combination with actually playing the moment-to-moment game. So, you've got the planning phase and the you know, the execution strategy and strategy and tactics kind of playing with each other there. All right. So, in conclusion, um so, in retooling Prey to make Mooncrash, we were looking for what might be next for immersive sims. I originally wanted to have like immersive simulations are dead or something dramatic here, but I don't think that's the case and it's kind of hyperbolic anyway. Um immersive sims are not dead. Um I think the monolithic vent crawler we're all familiar with is no longer the only expression of immersive simulation. Um you know, I think the regimented model is dead, uh and in its wake are all these really amazing possibilities. Um you got things like a lot of indie projects experimenting with the form and adding their own systemic wrinkles to the mix. A lot of mainstream games

are also adapting more and more of the simulation-driven elements into their mechanics suites. A big takeaway from this project uh was that you should be willing to break from tradition if you're given the opportunity. DLC is a really great, you know, fertile ground for that. And even if you're not happy with the results, and your video is still open, you can always bring some of those lessons back to your core projects, you know, the next time around. Also, don't be timid looking at other genres for inspiration. Uh we didn't quite go as wide as, you know, cart racing or collectible card games or something, um but there were enough different elements and problems in roguelikes uh that you know, resulted in this really unorthodox kind of hybrid uh at the end of the day. Um it was one of the most engaging projects I think I've worked on. And you know, I don't think at the end of the day that Mooncrash is the future of immersive sims, but I think it is a future of immersive sims. Uh special thanks uh to everyone on this big list. Uh I didn't come up with all these cool design ideas. There was a you know, design team that we collaborated with uh at Arkane that came up with a

lot of cool ideas. Uh special thanks to Clint for mentoring me and uh everyone that sat through my talk. And uh yeah, any questions? I think there are two microphones. Hi there. Um I think I noticed you said one run takes about 1 to 2 hours. Does that mean completing five runs in that kind of similar that that's for each character to each escape is about 5 to 10 hours? And All total I I think the game ends up it I think it ended up longer than the core like than Prey, honestly. Um any given run can take you know, an hour or two a a couple of hours. And also there was like the the the branchy thing was kind of a simplified example. Um so I think there are probably like 15 to 20 1 to 2 hour runs in in the game, something like that. And um uh the corruption meter, does it reset on on a failure and or and is it persistent across runs and it kind of

kind of keeps on ratcheting it up? So when the simulation resets, the corruption meter also resets. So it's not like a positive feedback thing. So you do get a you know, a reprieve basically. It starts over. Okay. Thank you. Mhm. Uh Uh yeah, yes. Um first of all, just I want to thank you for the talk and just working on these amazing some of my most favorite games of all time. Um A few questions about the both the time you took for the game jam itself and the time you spent on the prototyping and just how big was the team working on the game? The team I don't know if I can actually technically speak to team size. Um but it was about half of what we had on Prey basically. Uh and the game jam was about a was about a work week of time, maybe two, something like that. Um and we they were managed to turn out some stuff really quickly. Like again, having that leverage of the existing platform and the existing engine was really yeah. A lot of a lot of uh levers there. Thanks. Mhm.

Yes. So, um when a player discovers some emergent gameplay that lets them solve a problem in an immersive sim and makes it very easy for them to complete something, how do you make the decision between this is a glitch, this is an exploit, let's get rid of it in our design and uh this is working as intended, they've discovered a wonderful novel solution to something. Do you ever get that kind of decision to make and how do you make it? Lots. Um in general we err on the side of saying yes to the player. Um usually if if the bug crops up and it it prohibits a a well-meaning player from finishing the game, then we're you know, we'll we'll spend our time and focus on that. But if some clever player comes up with a way to subvert things, then let them have it. You know, as long as it doesn't completely like break everything, um we generally let them let them get away with whatever. Yeah, thanks. Yes. Uh when doing the design for Mooncrash, was there anything you expected to be able to carry over from the single player that wound up not quite working besides the obvious? Um

like anything you really wanted to get in but couldn't quite make work? I think if we had a bigger budget, um some of the like each of the characters have their own story missions. And so if we had had the budget, I think it would have been cool to do a complete like re-redecoration of the moon base for each player's like individual sort of vignette um and have characters walking around and stuff like that. That was the one thing that I would have liked to do if we had a bigger budget, I think. Thank you. Mhm. Yes. Uh so, you were talking about how like low-level agency or low-level agency like, you know, kill or not kill, seen or be seen, uh it can vary with these, you know, new and varying characters and the their play styles they present. Is it possible that if you were to expand that to more like high-level agency, like changing up like narrative or dialogue choices based on these individual personalities, like I'm the crazy guy, I'll pick the crazy dialogue option more than just, you know, the more natural neutral response? Mhm. That's

definitely something you can do. It's again, it's much more content intense and you need much bigger sort of team to support that kind of thing, I think. Um but yeah, it's definitely something you can you could uh entertain. Okay. Thank you. Mhm. Yes. Hey, two questions. Uh why did you um what was the idea behind making the corruption meter happen at the very start of the game rather than like 25% of the way through um with like progression? And then also, did you guys have any mechanisms for helping struggling players who are like failing over and over again? Uh so, the first question was about the corruption meter? Yes, and like why it was at the very start of the game when you first playing your first play through, it existed rather than when you've done a few of the escapes. It uh it actually only kicks in, I think, after the first escape, but it's it's very early on because we wanted to get a lot of play with that system, basically, and indoctrinate the players like right off the bat, like this is what the game is about, um dealing with this sort of you know, among other things, dealing with this corruption meter and and things like that. And what was the other question again, sorry? the other question was about um if a player was failing over and over again,

did they have any were there any mechanisms to help them? Yes, so we had the we had this currency called simulation points, uh sort of like the experience points in in Rogue Legacy or something. You accumulate those over time and you can spend them on on gadgets and upgrades and and blueprints and things like that. So, you can give yourself more gear going in the next time to sort of get over the hump. Got you. Cool. Thank you. Mhm. Hi. Uh another question about the corruption meter. Like when I was playing Mooncrash, I feel like the corruption meter give me too much pressure and I have no time to read the notes and books in the game. Like and a lot of friend of me is say the same thing. So, I'm just wondering why there's no an easier mode in the Mooncrash to let player like enjoy to explore the environment and to read the books and like that. Primarily about scope, honestly. Like if we had more time, there would would have might there might have been a a mode where you could like explore more. Um one thing we were kind of banking on was the player will be through this area like half a dozen to a dozen times, so then they can read that book the fifth time they come

through. Um but primarily a scope issue, honestly. Uh if you look at like if if if you look at the the book on Spelunky and stuff he talks about like having a staunch sort of like difficulty model and don't diverge from that and and and stuff, but I think we we would have probably Yeah, if we had more time development time, we would have. So, if there's more time, it would be better to add such easier mode in Mooncrash, is that right? Yeah, I think I think so. I think if I recall correctly, like when the Mooncrash is updated, there's also a new update for the uh core game that is a relaxed mode or something like that. Mhm. There was an easier mode for the core game, for sure. Yeah, so so that's why I'm wondering why there's no either mode for the Mooncrash. Yeah, primarily just development economics. Okay. Sure. Thank you. Mhm. Yes. Yeah, I guess I had a Oh, first of all, I want to say the Easter egg in the static was really really cool. Um the other thing is it kind of um plays on the previous question. I'm wondering what you think the impact was of losing a lot of times um in immersive sims, I think particularly for perhaps less skilled players, that kind of die

back to the quick save, try again loop helps people kind of figure out the possibility space and makes encounters quite a lot easier. Did you find that you were cutting out a a chunk of potential player base? Uh cutting them out because it was too difficult, you think? Yeah, well, I'm not saying necessarily, but I think there might have been some people who might say play Dishonored or something and they uh by exploiting that loop, that did you feel you were potentially cutting out people? I think it was definitely a much more focused set of people for for that particular project. Um Uh yeah, I mean it was it was kind of a a smaller subset for sure. Uh a more focused kind of experience. Um Yeah, cool. Yeah. Thanks. Last two questions. All right. Thanks. Hey, uh thanks for the talk. Thanks for the games. They're awesome. Um Thanks. I was wondering uh over there at Arkane when you guys are working on these immersive sim mechanics themselves, what the sort of ratio is to like when you're creating the individual sort of building blocks, like

intentional bespoke interactions versus ones that you sort of discover? Like in the example of the torch and the ice ball, it's like we're going to make heat cuz we want a means of melting this ice ball, right? Then you accidentally find out like, oh, but when you heat ice it makes water and then water does this to electricity. Like what's the sort of ratio to like sort of pre-planning versus like and what are design philosophies to create components that sort of play nice with each other? Um so in in general, the the mechanics uh often like the game objects that I mentioned, we we look to see how many different systems that they can interact with themselves. Um and then a lot of the times things will just kind of bubble up over the the course of production and iteration. Uh so we'll be, you know, we'll we'll have these systems out in the wild and some level designer comes up with a like, oh, hey, wouldn't be cool if the electricity conducted from this thing powered a computer or something like that, right? And so then we have to evaluate that and say, okay, do we want this to be a one-off thing or do we want to now sort of bring this up and and elevate it to a system itself, right? Um so a lot of this stuff happens organically over

time. And then there are there are some rare cases where it's like, okay, well, this these two things interact in a way that we don't we can't support because there are expectations that then cascade out of this. Um and then we have to cut something off, but in general it's like the more the merrier. Um and again, QA, you know, loves loves that stuff. Um yeah. Thanks very much. Mhm. Yes. Hello. Um so, Mooncrash is a pretty big departure from Prey. Um moving from the the Prey style of game to uh the Mooncrash style of game, how did that affect your team? Um assuming they were the same people who worked on the base game. It was the same people, yeah. Um but it was a smaller sort of subset that weren't going on to other things and stuff. Um so, there was definitely a like a bit of a learning curve and people were getting used to the sort of thing, but overall I think it was uh it was kind of a good exercise to like rejuvenate people between big projects and get them to let them try something different and and sort of stretch their

legs a bit. Cool. So, it it didn't like maybe take a big morale hit to the team. It was more reinvigorating instead. There were definitely some people that were a little like they were questioning it. Um but it was kind of a mixed bag, um honestly. Cool. Thanks. Yeah. Thank you. All right. Thanks,