Mining Your Own Design: Crafting the Crafting System in Astroneer

[Music] okay thanks for tuning in this is minding your own design crafting the crafting system in astroneer let's get right into it so first just a couple of quick introductions i'm aaron bittelcom i'm a technical designer with system era and i've been in games for about eight years and been working on astroneer for the last three of those before that i worked on galaxy this 2d roguelike shooter and on envelope vr which was a desktop vr platform most of my career has been focused on gameplay and interaction design especially features that are highly physical and tactile but i've also worked on everything from system design to engine architecture i'm sure any indie devs listening are familiar with wearing lots of different hats and i'm elijah o'reir a gameplay

programmer at system era i started in games around the same time as aaron but i worked in mobile casino games before making the jump to pc and console in addition to programming i also have a design background that's been part of my work i've often worked in ui and ux and large gameplay systems i made the absurd seagull simulator sploot to try my hand at vr and mess around with its tactile interactions and i did independent work for a while after that also wore many hats and joined system air about three years ago all right so what is this talk actually about well we wanted to give you a look at some of the lessons that we learned while working on our crafting update which came out near the end of our early access period we'd identified that the crafting loop was kind of struggling to scale up with the additions that we'd made to most of the other systems in the game but to improve it we had to make big changes to widely integrated parts of the game it was kind of like trying to perform

open heart surgery but on a conscious patient once we got into working on it we quickly started second guessing ourselves and it was kind of a red flag that we didn't understand something about astroneer mining your own design is really just a way to describe the process that we developed to overcome that paralysis we treated astroneer kind of like an archaeology dig site digging into it and deepening our understanding of how it worked it allowed us to make changes that could be almost seamless as a new player you'd be hard-pressed to spot the new from the old this meant reconciling some conflicting ideas about the core aspects of our game which made our goals seem mutually exclusive at first that conflict only really became apparent when elijah and i started pushing forward what at least felt like competing visions for the game we didn't really set out to use this

back and forth process it's just what emerged over the course of putting this update together mining your own design happens whenever we were in conflict with each other i was coming at the crafting system with more of an outsider's perspective well aaron was really familiar with the existing gameplay and how players were engaging with it soon after we started working on the update we both accepted that uh we both had valid points and diving into ashnea's mechanics to break them down and understand them wasn't so much to prove who was right but to find some sort of missing or misunderstood piece of the design that could reconcile our perspectives we didn't realize it at the time but this is going to change our understanding of astroneer and how we build it working together also revealed holes in the design process we were using and advanced it further it established the immense value of bringing the engineer into the design process early on so if you're not familiar with it astroneer is an exploration crafting and survival game

set in space it focuses on ground-based exploration rather than spaceship travel and it's limited to a small set of planets in a fixed solar system by this point in development we had a lot of things that already worked great and made our game really unique our planets were procedural but each of them had a distinct identity and this made exploration both beautiful and unexpected our terrain was fully deformable and you could sculpt it like plato item interaction was extremely tactile and diegetic which made it not only charming and toy like but it also was intuitive for new players and we had this chill low stakes vibe that emphasized relaxed problem solving with only the occasional moments of danger to kind of spice things up all of this was presented in a vibrant aerospace fantasy package so what was the problem well player engagement dropped off after just a couple hours we were in early access and the opinion

of the player base hovered around hey great concept can't wait to see the full game this had not shifted despite months of updates it was clear that there was something in astroneer that was sapping player motivation or just not really keeping their interest there were a lot of fun and interesting things especially what we had added recently to the game that players just never even got to see i believe that it was the crafting system that was kind of the main offender for all this it was flat compared to the new drift of experiences you got from spelunking unlocking items and visiting different worlds baseballing and exploration both generated resources that followed into crafting it was kind of the end point of almost all steps of your game play nothing changed how crafting worked though past the first hour of the game it lacked a big moment where mechanic got introduced that changed some part of the crafting system you're familiar with and gave you this totally new experience with new opportunities

i like the cadence of crafting games how they managed to fold in new resources new steps new ways of building the longer you play these are big important beats in the experience that give you something new and interesting to play with and they make me excited to see what's going to come next it felt like astronomy was really missing this opportunity i thought that we should have more resources in the game but just not make them all available from the outset i wanted to add a few new modules that combined or transform the existing resources into entirely new ones this would give us a couple new resource sets unlocked by these modules to divide the crafted items up in between each module would feed into the next giving us a nice natural progression so that we could divvy up the recipes between early mid and late game pretty effectively i didn't actually disagree with elijah at least in principle but i was having trouble squaring it

with that same laid back low stakes 5 that felt like a core part of astroneer's identity the team had often talked about simplicity and minimalism as watchwords for the game and i was worried that the various additions we'd made to the crafting system already had moved us away from those concepts despite the fact that the overall crafting loop was shallow and repetitive it also had a lot of moving parts that completed for the player's attention on top of that after the smelter each of our crafting modules only got more complex in their interactions involving things like reusable containers which were needed to hold resources even though players use these modules less often than the smelter their complexity took a lot more focus and it wasn't clear how much the game had actually benefited from it the more elijah and i battled over this the more we started to realize that we were really arguing over the issue of cognitive load in our game

normally when you hear that term it's in a negative context as in you have too much of it but you can just as easily end up with not enough which results in losing player engagement this was the problem that elijah was pointing to with our crafting system its core loop centered around just three simple steps that the player repeated ad nauseam you mine resources out of the ground toss them into the smelter and use the output to print new items on top of that all but one of these resources was available on your starting planet and there were no choices in how to use them outside of what you wanted to print but at the same time we had added several additional crafting stations and instead of creating a stable long-term loop they'd started chipping away at the simple elegance that players had found appealing to begin with from my perspective i felt like we were already placing too much cognitive load on players to the point that we were losing them to

frustrating complexity you run into an apparent contradiction of opinions like this you know too much cognitive load versus too little don't be too quick to just discard one argument and take the other it's easy to use a common language that is actually obfuscating what you're really talking about here and i both had an opinion about the crafting system but it was only when we laid out the specifics of our positions that we realized we were talking about different subsets of the crafting system aaron was talking about the visceral moment-to-moment interactions that the player had with crafting you did have a really good point about how these modules were complicated to players for players to use which i hadn't concerned myself with very much instead i was focusing on the systemic interactions within the crafting system and how to beef those up we were both right but my relative distance from working on the crafting system day in and day out gave me a pretty clear view of the core loop of astroneer's crafting and its deficiencies

and speaking of the choreograph crafting loop this diagram illustrates it and kind of how it works to start with players use the deform tool to mine resource nuggets from deposits embedded in the soil two of those mined resources could be fed through the smelter which would turn them into metals and the final crafting step was using uh the resources that you would mined and that you would converted with the smelter in order to make printed items a printer would consume a couple nuggets and output some kind of item like a solar panel or an oxygenator you know something that's sort of useful in space the smelter was the only module that created new resources in the game and it was also the first crafting module that players would unlock after about 30 minutes of play it only created two new resources that

had to use oars didn't have any of their use other than to be smelted they weren't used in any crafting recipes directly so the player wasn't facing many trade-offs when they were engaging with the crafting system using it was very simple you can see in this video over here that all you have to do is take an ore place it onto the smelter and then it turns it into a nugget of metal it's a one-to-one conversion um and again you can't use the or for anything else so the smelter being so simple in its form and having such a simple ux it really felt more like a beginner crafting module in order to kind of ease players into the concept of using them but you could make literally every item in the game using just a sponsor and printers so let's take another look at that crafting diagram as we said in addition to the core loop

there were these three other modules that were involved in crafting but were on their own separate track we'll call this track the spur because it kind of poked out the side from the core loop the first of these was the soil centrifuge and it was practically the module equivalent of your terrain tool is it produced all the same resources as it branching out from that were the chem lab and the condenser both of which made the same single resource hydrazine which served as astroneers rocket fuel but wasn't actually used in crafting the soil centrifuge worked by taking a sample resource that you would have mined out of the ground and effectively cloning it in exchange for the soil that was collected when you deform terrain there were a few parts to it that can make it confusing to use which is probably a little easier to understand if we see it in action so the orange buttons that you're seeing would allow you to select whatever output you wanted but it also made it clear which resources could be cloned

but you could also just select your output by placing any valid sample resource which would automatically set it as the selection the amount of resources that you got back depends on how much soil you provide and different resources have different exchange rates all of this presented a bit of a learning curve without adding anything fundamentally new to the crafting loop in the case of the condenser in the chem lab like we said both of them only produced that single resource hydrazine originally hydrazine could be either mined out of the ground or produced from nothing but the atmosphere and electricity through the condenser mining something as complex as rocket fuel felt kind of bizarre but the condenser was way too overpowered and unbounded to serve as the only way to produce hydrazine so to address both of these issues we previously previously added the chem lab which crafted hydrazine from a new

resource that we used to replace the hydrazine deposits at the same time we changed hydrazine to depend on the same reusable canisters that were also used to store soil this introduced a constraint on the previously unbounded production in the condenser as players now needed to manually supply it with fresh canisters but the downside was a finicky and kind of tedious ux these players had to juggle their canister use between soil and hydrazine and while we had intended for the chem lab to eventually produce an entire set of complex resources in the meantime we were presenting players with two ways of doing exactly the same thing without a clear trade-off between them so let's see how those both worked with the condenser using it was actually pretty simple you just attached an empty canister and started it running as long as it remained powered it would slowly fill up the canister until it was complete

you could also put empty canisters in storage next to it and they'd be pulled in automatically when the previous one was full for the catalyzer that thing that looks like a pizza oven hold the mind resource ammonium for every nugget of ammonium you also needed an empty canister if you had a mismatch between those the machine wouldn't even start otherwise you'd set it going and a few seconds later you'd have some hydrazine the issues with the sperm modules was more subtle than just an overtly broken design it wasn't that players didn't enjoy using them at all outside of that initial learning curve it was actually full of the the kind of tactile and dynamic interactions that astroneer was so special for the real problems only manifested in the interactions between these modules and the rest of the game in one of two ways either they were too cut off from other systems kind of acting like a mini-game or they were redundant with other systems and

ended up undercutting them we found that these kinds of design challenges can be especially difficult to address since things often seem to be working up close this not only leads to an urge to defend them from valid critique but it can also obscure problems until much later on after those systems have become entrenched confronting issues like this require you to look at your game holistically and evaluate individual parts of your design as gears in the larger machine in the case of the condenser in the soil centrifuge both of them had been built around testing out different sinks the condenser was a power sink and the soil centrifuge naturally enough was a soil sink in order to demonstrate each of these ideas we'd use existing resources to serve as their output mostly just as a convenient proof of concept but in each case this had led to the same unintended consequence

both functioned as resource gathering shortcuts that started to undercut the need to collect resources from the wild mining resources served as the backbone for exploration and it helped establish the alternating rhythm between exploration and base management the chem lab was the least problematic of the three and mostly it just suffered from being underutilized we still felt like there was a great opportunity to expand its use with new resources but this was stymied by the problems we were having with the other two with the strain that these modules placed on the rest of the loop the obvious question was why not just cut them and start over aside from the fact that we were a live game oh the problem was that from the player's perspective they had real value the instant gratification that they provided sidestepped a lot of potential frustrations that arose when needed deposits came up short

in a way they functioned kind of like player junk food in small doses there wasn't a problem but without any limits placed on them they soon took over the more balanced diet of the mining loop players gorge themselves until they started to feel sick and ultimately started burning out we saw this pretty clearly with the soil centrifuge i had actually anticipated the impact that it might have on resource mining and i tried to mitigate it by making it less efficient than mining deposits it also required that sample resource of whatever you wanted to produce so you couldn't just skip ahead without doing some of the initial exploration the idea was that it would function great as a supplement to mining since you could collect soil in the process and get a double return for your efforts but players started to just grind for the soil itself digging up the terrain around their base instead of continuing to explore for

deposits that they could just clone instead we'd made some attempts to tune our way out of the problem but it didn't really seem like there was a sweet spot we were kind of in a stalemate with our players so over time we really lost sight of why these modules were introduced in the first place the design intent was to provide new resource syncs and engagement with base building we paired those things with these potent resource shortcuts as an incentive for players to generate power and collect soil but over time in our minds the intent incentive became a single inseparable assumption about how these modules just had to be we felt stuck in trying to balance a resource cloner in this infinite rocket fuel generator was not working this is why i'm mining your own design and doing it with someone else can be so important in really decomposing how your design works and figuring out what parts are actually necessary and what parts are not and i brought the perspective to aaron

that we needed a new resource set that expanded over time with the crafting gameplay he started to look at the existing crafting system in a totally new light my proposal provided a new set of incentives for players to engage with crafting and prompted him to look at their existing modules and the instances they provided and asked wait a minute maybe these modules don't have the right incentives at all this is the sign we needed to know that we were on the right track with this redesign the experience that we intended the player to have started lining up with how we were incentivizing them to play astroneer if you find yourself struggling to align your designs incentives and intent challenge the assumptions that you're making on each half of that divide this can reveal a hidden or flawed assumption if you're still struggling to challenge your assumptions then collaborate with someone who has a very different vision for the design it's easy to get attached to the way things are and to take it personally but

let it go this can be difficult to practice but it's a powerful way to get through problems that have got you totally stopped in hindsight the solution to these three was actually pretty straightforward since all we needed was to prevent these modules from being redundant with each other then the answer was to make all of their outputs unique as soon as we landed on this idea it immediately clicked for us having more modules that made unique resources gave us a way to expand the resource table without adding tons of new deposits which was pretty disruptive to the game resource acquisition could tie into additional game systems in a more fundamental way and most pressingly it would expand the depth of the crafting system without even needing new modules while simultaneously reducing the cognitive load caused by the current disjointed system the result of this was a distinct purpose for each of the three modules that actually strengthen their existing conceptual identities

first up we have the soil centrifuge instead of cloning our mind resources we clean coined the idea of trace minerals that could only be produced from soil this leaned into the fact that soil could be collected anywhere making these resources more reliably available than their mined counterparts this also did away with the idea of the confusing sample mechanic which made it much easier to select what you wanted for the condenser we embraced the idea that you were collecting resources out of the atmosphere itself and going further with that idea and giving every planet a different set of gases that could be produced and since that meant it no longer took the place in of other systems there wasn't as much need to constrain it which allowed us to do away with the reusable canisters that players were finding kind of tedious and finally for the chemistry lab we could finally expand it into a

multi-purpose crafting station to differentiate the complex resources that it would produce from the smelter we introduced the idea of composite resources which would be made up for mixed ingredients this had a secondary benefit and that it allowed us to compact multiple resources into one which was particularly important in a game with a diegetic inventory where your inventory space was actually physical and could take up real space in the game taken altogether these changes not only addressed the existing issues that we'd had with these modules they opened up new opportunities for us the condenser and the centrifuge could provide new challenge factors for resource scarcity planet travel and large-scale terrain deformation and the chem lab created a huge multiplier on our raw resources adding any number of permutations between them for the first time it felt like our goals for this update actually seemed to be in sight

you can kind of see with the graph there how this would work the three different um sources for our raw resources could funnel through the other crafting stations into finally a complex one at the end so fixing these modules required us to take a step bass back and ask two questions vital in this process of mining our own design what do we want the player to do and why is the player willing to do it the first is about examining your intent as a developer how does your economy enemy ai and counter design all come together to make the gameplay experience for the player what possibilities and challenges are you trying to introduce to them the second question is about what your player is actually doing in your game players can come up with shortcuts you never anticipated that subvert your intent they can also put up with boring or bucky gameplay to accomplish their own personal goals motivations can vary wildly they can be pursuing mastery joy item rewards fantasy fulfillment

but what matters is that your player wants it insists have shaped the desire path that players take through your game to get to what they really want you're playing the game wrong as kind of a classic designer complaint but this is more about spotting negative player motivation that's enticing them to suffer for what they really want the goal is to alleviate player fatigue and make sure they're not putting up with clunky ux or in gameplay this is a factor in our big game lol that i didn't totally appreciate at the outset i kind of thought it was all about systemic design right but once you've identified these negative player motivations fixing them is going to lead you back to that question of what do we want the player to do every time you iterate those two questions you bring your designs incentives and the player's desires into tighter alignment this probably doesn't sound like a totally revelatory process to you but it can be easy to forget to do it at the right time and more importantly to consider both these questions fully

so at this point we had kind of assumed that the hard part was really behind us uh astroneer uses reward materials partly this is for the aerospace fantasy aesthetic but we also felt made the printer recipes more intuitive we assumed that this real-world connection would make it very easy to pick additional ingredients aaron did some research and came up with potential set and it was too large and he struggled to narrow it down thinking that maybe we just didn't understand them well enough to uh to leverage these new resources aaron invited me to take a look over the potential ingredients with them i have an interest in engineering and manufacturing these were real-world ingredients they had real-world properties that were useful in the manufacture of various goods things like electric wood conductor light structural material shock absorber we thought tagging all these new resources with these properties was going to make it obvious to pick instead

we found it no easier to judge it was no easier to judge if astronaut needed another electrical conductor or shock absorber versus if it needed tungsten carbide or carbon fiber so running around in circles like this was an indication that we didn't totally understand something about our design maybe an entire pillar of our design that hadn't been identified yet we had to do some mining to discover the missing connection between astronomers resources and kind of the recipes from the outset we assumed we need more resources was the problem that we were solving it was but mostly in service of adding a crafting progression to our game we kept asking what are these materials used for in the real world to try and understand them better but that's not used as a useful question as what are these materials used for in astroneer to print items is the answer to that question half of their purpose

was unexamined our exhaustive research on the new resources made us pretty sure we had a good grasp on them but maybe we didn't understand the anatomy of a printed item itself well enough it was easy to forget about printed items because we weren't really adding many to this new update so we took all the resources off of the board and put up all the items you could print instead we went through each of them and started asking what would matter if you tried to build this a shuttle would need to be light strong and heat resistant a rover would need to be tough and have good shock absorption the sign of questioning made us focus on the output of the crafting system and it generated the same set of material properties that would come up with for the resources this is an exciting moment up until this point we never thought about our printed items in this way we had such a small set of ingredients in the game that we'd made all of our recipes on intuition there was an underlying logic to the system we had just never articulated before unfortunately with astroneer late in its early access development there

were a lot of printed items to do this mapping exercise on it gave us a very clear road map of the material properties that would be most useful for printing astroneer type items knowing that made the set of resources that best match those material properties quickly emerge so part of the problem here was that we were really relying too much on our design intuitions especially on a small indie game it was pretty effective for a long time to just kind of run off of our gut feelings on what the game needed and how things worked in the early stages of development having good design intuition is an effective way to build up a foundation but as the game becomes more complex those same intuitions can start to get murky rather than relying solely on your fight on your gut feelings try to try to identify the underlying rules guiding your intuitions and make them explicit this makes it possible to quickly iterate on concepts with confidence instead of spending weeks trying to

understand why something feels off so we were in a pretty good place every new ingredient in the revised crafting system had a purpose except for the gases gases were tough to think about they are used in manufacturing but more often than being just like a direct ingredient in making something more like a part of the manufacturing environment that you have to control for for instance oxygen oxidizes materials and so if you're making something and you do not want it to oxidize don't make it in an oxygen-rich environment the only gas that made sense as an ingredient was hydrogen for making hydrazine but we needed six useful gases to spread across all the different planets that had our progression we considered using the gases as part of the recipe for the crafting modules to try and capture that concept of their you know component of the manufacturing environment but it ended up being a weak comparison an alloying furnace isn't made out of nitrogen it just kind

of needs it to run um more importantly than that we wanted players to build a lot of atmospheric condensers not just get a gas once and then never have to worry about it again as before it wasn't that we hadn't thoroughly researched and considered the gases carefully but that we were failing to consider how they were actually being used in the game we just rolled with the assumption that gases would be used to print items like any other resource but there was a question to be asked of what are the printer recipes for they're there to give players items yeah but they're also there to take the players resources printer recipes are the resource sink in the game but they don't have to be the only one we were chasing the powers into the atmospheric condenser we could just come up with a new sink that could keep pace with its output maybe the alloy furnace could require certain gases to run

thinking about it this way made sense for the gas to only be partly consumed and add like a fuel for many of our different crafting recipes reframing the role gases played in crafting led to a new sort of item a canister of gas that would get gradually consumed to run crafting modules this shift into contextual framing for the gases allowed us to think about them differently and break the constraints that we'd always had around printer ingredients where they were fully consumed and you know they were competing with all the other printer ingredients in their input slots it made us more comfortable with putting the gases in all kinds of recipes for making composite resources that maybe may not have been a major player in otherwise this idea of what's this for that we asked with the gases isn't just an arbitrary question we've talked about several design problems that were only solved for us after we challenged a constraint that we had previously assumed was essential

what's this for is kind of a thought exercise and it's actually one of my favorite design tools it's a simple way to effectively take a step back when you find yourself spinning your wheels and remind yourself of what it is that your ultimate goal is what do you want the player to do and why is the player willing to do it are really just specialized versions of this idea asking what's this for forces you to articulate what it is you're trying to accomplish without focusing on the particular approach that you've been taking to get there so for instance if we used it on the soil centrifuge it revealed that we just wanted a sink for soil and simply needed to provide something of value to do so if we employed it on the resource table it revealed that we needed to know how our ingredients would be used in estrogeneer not in the real world and of course with the gases it got us past trying to confine them to the same conceptual uses or other resources

from here on things went pretty smoothly we'd realize the value of mineral and design and we got into the practice of challenging each other's assumptions on the fly our last problem to tackle was how the player would make the view composite resources you see we'd already identified five or so manufacturing techniques that i had convinced everyone that we were going to turn into new crafting modules but once we parceled the recess of resources out between them a few were severely underutilized we collapsed the crafting manufacturing techniques into two generalized concepts both of which could justifiably use gases as fuel heat processes relying on inert gases and chemical relying on fossil gases this resulted in two composite resource machines the alloying furnace and the chemistry lab we prototyped these and places them with the team but they were so similar the team found it very difficult to guess which one they needed to use to make any particular resource

the only difference was the type of gas used volatile versus invert otherwise they have the exact same ux and basically the same thing people don't have a great intuition about volatile and inert gas as it turns out um it just wasn't a strong enough distinction and it was departure from how unique all the other crafting modules felt from each other in the end we combined these two and kept the more conceptually open-ended one the chemistry lab aaron had really challenged me and be like come on man what are these really for and i was like all right you're right this replaced the old chemistry lab and it brought us full circle um we'd always imagined that kim lab could be the station for combining complex resources and now it finally was it's left us with a final set of modules that kind of had a surprising twist to them not a single brand new module was necessary with all the changes to our existing modules that we already talked about we had everything that we needed for this crafting update

not only did this preserve existing save files which is a godsend and a live game but it made onboarding the player to the new crafting system incredibly fluid while the inputs and outputs changed the overall ux heard each item was familiar it had been improved somewhat but it was remained largely the same this was true too of their visual language which a lot of these modules were kind of older and so really when we visually updated it from up when we upgraded them for the update it felt more like that just visual updates not a total rework of the modules we managed to address both of our concerns about cognitive load that had plagued this at the outset of this project being able to keep that familiar ux and visual language kept the cognitive load manageable we didn't have to teach players how to use these things but redesigning their output totally added this depth of crafting system that gave it some more staying power it was as though we had uncovered the true form of our existing modules and

our ugly ducklings that proven to be beautiful swans so this had taken us pretty much to the finish line with this presumably all that was left was to actually implement these modules in game and ship the update so far we've mostly talked about this idea of mining your own design in the context of identifying the broken or incomplete parts of your game and finding solutions that still stay true to it but as we learned at the end of this update it's also an important piece of protecting the parts of your design that do work or critical parts of the new solution while the crafting update was on the whole a significant success for us it wasn't without some key mistakes that ended up compromising some of the goals that we had set out with most of these came right at the tail end of development at this point elijah had been pulled away from direct collaboration with me in order to start actually implementing these new features

ostensibly the design had already been locked but as the features started coming online ongoing feedback from the team made it clear that further iteration was needed unfortunately we had neglected to buffer the schedule sufficiently to absorb these absorb these setbacks resulting in some hasty last minute changes to the design and without the benefit of elijah's counter perspective that i had relied on throughout the rest of the development process we were missing one of the key ingredients to this idea of mining our own design i want to take a closer look at one of the mistakes that we made here and discuss how better using the concept of mining our own design could have prevented or at the very least mitigated it play testing the new version of the soil centrifuge had revealed two overlapping issues previously the overlap between the centrifuge and mined resources

provided a kind of release valve on the persistent need for the common resources without that overlap players started to chafe against that constant need which crowded out their limited personal inventory space and forced them to make multiple trips to collect all the different kinds of resources that they needed and as they depleted the local deposits of those common resources this meant tedious expeditions in search of a resource that had once been plentiful on the other hand this the exclusive nature of the centrifuge meant that soil collection was now an unavoidable part of the crafting system this was of course intended but the prevalence of the soil extracted resources in our new recipes meant that players had trouble keeping up using only the soil that they got while mining deposits we kind of had pushed into the soil grind instead of alleviating it with the clock running down toward our

release we needed to figure out how to address these issues fast the simplest solution at least on the surface was to just restore some of the mine deposits to the centrifuge and at the same time make deposit versions of its new exclusive ones this did solve the immediate problem but it also removed the centrifuge from the crafting progression and now the mid game rested solely on the shoulders of the chem lab in the condenser spread out across six different planets so what could we have done instead importantly the two issues that had shown up in play testing had a serious impact on the pacing of the crafting loop having a solid progression was definitely important but it wouldn't matter much if players tired of the loop before they had moved through it doing nothing wasn't really an option here as before the real failure on our part

was in not asking ourselves the right questions what do we want the player to do and why is the player willing to do it more simply what's this for asking those questions would have reminded us of the critical role the centrifuge played in our new progression and redirected our attention on other areas that were more flexible such as the exchange ratio on the centrifuge or the frequency with which its resources showed up in recipes but it had been all too easy to see the new paradigm that we were introducing is the actual problem and fall back on aspects of the old system that were at least a known quantity ideally we would have had time to iterate on these issues but even with the most forward-looking production schedules the reality of development sometimes means that you have less that less time than you'd like to make decisions mining your own design allows you to isolate the most important core principles of your game

and better adhere to them when you're making these kinds of tough calls ultimately the crafting update was still a major success for astroneer it not only filled out our mid game but it also paved the way for another year and a half of updates since release the number of crafted and items in the game has more than doubled and it's done so without adding a single new resource or crafting module we haven't touched this crafting system at all since we introduced it year and a half ago but more importantly the process that we went through on this update greatly improved our understanding of astroneer allowing us to make future updates with more confidence and efficient iteration earlier this summer we released the automation update something we've planned for the game for a really long time it was one of the most major features that we've released made a number of improvements to the crafting providing new ways to counteract the tedium of manually upkeeping a functional phase but thanks to the stable foundation we built we were able to layer these new features

on top of the existing crafting system without having to overhaul it at all so mining your own design is not just a technique that's specific to astroneer it's about asking the right questions about your game to avoid sticking to familiar design patterns that are failing you asking these questions that challenge even your deeply held assumptions can allow you to mine the texts for hidden rules and a deeper understanding of what makes your game tick bringing in someone from a different discipline can invite a totally new vision for the design that's able to generate the sort of questions that are tough to ask yourself when you're really close to it on a daily basis it's all about nudging yourself into asking those fundamental powerful questions what do you want the player to do why is the player willing to do it wtf what is this for all right that is our talk once again thank you very much for tuning in um we've put up our email addresses if you

have any questions that you want to send us and i hope very much that you enjoy the rest of gdc [Applause]