Until You Fall: Building Satisfying VR Combat on a Budget

[Music] hi everyone and welcome to our virtual version of this talk until you fall building a satisfying VR combat on a budget I'm Dave Bennett the guy with the glasses on the left and I'm Patrick shell bear the guy hiding behind swords on the right and we're both game designers for a studio called shell games and if you haven't heard of us we're a studio that works on a wide variety of different weird things both original IP and client work but our most well known VR title is probably I expect you to die it's a kind of escape the room spy v puzzler where you need to escape death-defying situations that only a you know spy with telekinesis would find themselves in but again we work on a lot of other weird stuff ranging from location-based entertainment to physical toys that teach chemistry to iPad apps that teach your toddler how to stop and go potty with the assistance of Daniel Tiger but

the most recent game that we solved weird problems on together is a game called until you fall which is a melee combat action game with rogue light elements where you are a badass glowing knight fighting your way through a neon hellscape that kind of feels like it was taken from every 1980s album cover and the core combat in the game is striving to be what we call like a rhythm light combat or gameplay that focuses on choreographed attacks and blocks and tries to maintain a much faster pace of combat than we were seeing in other melee games at the time and to give you an idea of what we mean by that we've got a clip from one of our top players [Music]

[Music] [Music] and a shout-out to kosis for giving us that so yeah kind of with the high level of until you fall kind of established we I also wanted to get take a moment and kind of give some context for why Patrick and I wanted to give this talk so when we first started working on until you fall one of the early goals was to make something that was not only high action and variant like you know high-intensity it was also profile it was kind of a stark departure from expect to die but with so many other VR titles kind of under our belt we wanted to you know try stuff we hadn't done before and even though Patrick and I are

both game designers at the studio we both had different roles on this project so I for example and the project director who's responsible for kind of the high level vision of the game as well as the theme and overall cohesiveness of the game itself so you can thank me for the weird synth wave and neon nights but Patrick is the design director and he's in charge of the combat systems the kind of player experience and the overall implementation of how we actually do combat in this game so even though this wasn't shell games first VR title we faced a lot of kind of difficult challenges I mean like we've worked on I don't know this is like number 8 or 10 or something like that there's a lot of VR that we've done at the studio and four until you fall there are a lot of useful lessons that we had learned that we wish we had known earlier on and kind of after talking with Patrick and trying to reflect on our experience making the game we wanted to take these lessons learned and try to give them as

takeaways to give to you guys so that hopefully you guys don't end up making the same mistakes so Dave and I talked about it a lot and cut a lot of slides and tried to boil down all of our takeaways from until you fall into kind of five major points that'll be the subject of the talk that we're going through here today one it's okay to reference non VR games for your VR game to be wary of adding secondary VR interactions three no your player fantasy especially for vr4 set the rules of your virtual reality and 5 find ways to incentivize your player fantasy in your gameplay so kind of getting started with the first take away it's okay to referenced on VR games and this might be something that's just more specific to us and our studio but there's a lot of things that with

our experience in VR it kind of really made it hard to trust our instincts so kind of taking a step back when we first started working in VR there's a lot of stuff that we had had to throw out a lot of experience or instincts that kind of thing and in VR things that seemed oh so obvious and so straightforward actually aren't and more often not can make players just super sick and that can lead to you feeling like all of our experience up to this point making other types of experiences location-based tablet PC whatever it feels like it leads you in the wrong direction and can lead you to becoming super shy about trusting those instincts that you'd honed over years of making games and kind of worse of all this actually leads to this feeling of things you can and can't do in VR these generalizations that kind of become ingrained in us and it becomes this vicious cycle where you

can't do that because it didn't work that one time so we shouldn't do it again therefore it becomes the truth that you could never do that and even for problems that in other mediums have solutions that are pretty straightforward it feels like you have to invite entirely new sometimes very expensive solutions for these problems so something that became a really important learning point for the team in pre-production was that our our instincts aren't always wrong and sometimes the non VR solution to a common game problem can still work especially when it came to game feel and until you fall because our game was a hack-and-slash game so our hit impacts were some of the most important things and and a big challenge early on for us was figuring out how to get the that same good meaty game feel that you could have in a non vr hack-and-slash or a non vr combat game to translate faithfully into

VR and non VR games would try to accomplish these things with screen flash camera shake hit pause sometimes an animation hit pause things that some of our developers were nervous about trying to translate into VR but we did end up studying a lot of these non VR games and even when we couldn't directly take the solutions from those games we were able to really use those as our starting points and as our tone targets and sometimes the solutions actually did just work which is a which was a happy accident we found out that screen flash and camera shake worked way better in VR and were way more successful in terms of getting a satisfying hit than we thought it would and we only needed like minor tweaks to get them VR ready but in every case we were always ready or capable of using those non VR solutions as starting

points and then iterating towards those VR specific solutions if we needed to and for until you fall this directly led to how we solved the game's movement system how we solved all of our impact in our VFX system and actually led a lot towards our final solution for weapon weights in the game since we made that as a direct response trying to come up with something to mimic the animation lag on a heavy hit in in a game like Devil May Cry or God of War so even though the non VR solution for some of these problems do work or at least serve as a good starting point we are still making a VR game and one of the pitfalls of working in VR is that it feels so real and you want to make it feel as realistic as possible and I guess our second takeaway for you is to beware of those secondary viewer interactions also known as don't drop the sword but we'll get to that in a little bit so when I'm

talking about secondary interactions I'm really talking about those high fidelity moments of VR that are super exciting and super seductive it's those those things where it's just you touch everything in the world you pick everything up you can interact with anything and everything because as a virtual world it should you know reflect reality and be realistic so I should be able to pick this object up and hit it against this other thing and it will work the way I expected to because of my experience in reality and a lot of early VR actually based their gameplay including I expected I on having this wide variety of novel interactions having the sandbox of different things you can interact with but if your game is trying to have a more narrow or a specific fantasy having a lot of these interactions can actually become a trap which if you haven't figured out already is when we fell into during until you fall is development so to kind of illustrate this I wanted to kind of walk through some of the earlier features that or you know the lost features of until you fall that you know we end up

having to walk away from so way back in the before time the pitch for until you fell actually started with a solid combat prototype so it was just a simple thing where you were fighting one enemy and it highlighted the you know fun nature of our choreographed combat system and you know obviously to take that from a prototype to a full game we wanted to make it a fully immersive fully featured VR dungeon-crawling experience so with that target in mind we wanted to we started prototyping all sorts of features that seemed absolutely necessary to not only hit you know stay in the space but also keep it entirely immersive in VR so we added a bunch of stuff like you know looting cuz there's treasure and dungeons consumables because you need potions and things to acquire and heal yourself which means I guess you have health and maybe money I don't know there's weapon holsters cuz you have a sword so you gotta put it somewhere the sword has physics so I should be able to throw it that makes sense also the weapons should be you

know stuck in trees cuz you know they're sharp and sticking them in trees or you know even better enemies is awesome and hey there's a big rock over there so I should be able to climb it because I can climb in real life and more stuff along those lines and for every feature that we added it ended up in spy to either add more features more depth to the existing features or more polish to make sure it met player expectations in addition to her own quality bar so we kept adding these features and we were sort of following this normal path of iteration which you know coming from the studio that made a game like I expect you to die it was not super surprising but we started to get some issues with our play tests especially as we got more into combat play testing we noticed that players were very frequently during combat dropping their swords and after they've dropped their sword during combat they'd look at the ground looking for their sword fumble around trying to pick it up and get beat around the head

and shoulders by even the most basic enemies in the game and it turns out this is actually the symptom of a bigger problem because by adding all of these secondary interactions in pursuit of that like VR immersiveness and that feeling of being a fully featured game we'd added a bunch of new ways for players to be bad at our game climbing introduced falling particularly during combat the whole concept of weapon holsters introduced the weapon dropping issue looting became grabbing the wrong thing during combat throwing was really just dropping with style and what been pinning was tough because it was really awesome against the last enemy in a room but mostly super frustrating when you were trying to fight and your weapon was getting stuck off-screen and each one of these problems individually was solvable like we could iterate into solutions to shave the rough edges off of these but

the iteration cost for all of those little rough edges would push us out of scope so ultimately these were all features that we felt we couldn't actually afford and be faithful to so if the worst offender is dropping the sword obviously the solution is to not drop the sword right I mean we actually did take a hard look at what was really important to her core gameplay and going back to our simple prototype its spoiler warning it was about hitting dudes and we asked ourselves in what ways is it okay to be bad at this video game and as we went down the list of all these things that we thought were oh so important to make a fully-featured dungeon crawling game a lot of the VR interaction polish that we had added was just straight-up removed or reimagined entirely and wherever possible these points of friction these kind of stumbling blocks that vexed players so often we tried to turn those moments and make them opportunities for the player to feel awesome instead and going back

to dropping swords I think weapons summoning our solve for that became kind of our best example so in our game instead of dropping swords and having physical weapons you have magic hands with magic swords and at a moment's notice you can summon that sword out of thin air so if you need don't want your sword you can just get rid of it and be able to interact with a limited number items or you know if you do it by accident at a moment's notice rhe summon that weapon and it's okay so these small moments were it's upside up now actually look and feel awesome and off of that don't drop the sword actually did become a mantra on the team as we went into our next round of iteration and a process that we run through that we'd recommend absolutely doing with your own games and your own features in VR is for every feature asking yourself does this feature do anything other than increase VR immersion and if not does it introduce

more depth or confusion does it add new ways to fail and is it supporting a core interaction in an important way that justifies it doing that and even if it does could another solution do that same thing in a way that doesn't add those points of failure and lastly is this a feature that it is supporting or undermining what your core player fantasy is or it does have anything to do with it at all and that takes us to our third takeaway which is knowing your player fantasy and really kind of understanding it and instilling that knowledge to your player and this is something that you know as designers we generally you know try to establish a player fantasy early on but for until you fall in for VR especially it seems extra important so once we had kind of changed the rule to you know you can't drop the sword in

this game it really kind of helped us establish what your fantasy is in this world you are not some bumbling idiot you are a badass magical night from the moment you put on the headset it's not a story of zero to hero this is a competent swordsman at the very least becoming a literal sword god and as a rogue light it's also a game about dying and learning from your mistakes while still feeling okay about it in this game death is expected so it needs to feel worthwhile and you know understood and you didn't die because you are a chump you died because the odds are unbelievable and you fought valiantly until the bitter end and you'll be rewarded justly so that put us in a place where we want to make the player feel empowered in a game that is still a roguelike knowing that you will die so to reiterate and until you fall you are supposed to start off feeling like you

are a competent swordsman in this universe and then progress towards being like this transcendent figure and preserving that heroic identity became the guiding light for a lot of our secondary features that we did keep in for example our rewards like every upgrade or item that you select in the game is actually a crystal that you grab and crush in your mailed fist and that sort of underscores that moment of winning our narrator after this pivot was redesigned and she was turned into a character that was much more earnest and there to pump you up and tell the player that yes they are worthy of this challenge and they can overcome it and what they are trying to do is difficult and worth cheering on and lastly something that actually helped a lot with this was introducing a new weaker enemy because we found that our our husks which were our basic enemy at

the time we're accomplishing a lot of technical goals for us they teach you basic combat they showcase our mechanics they explain to players the best way to play but players would still die to them in the tutorial and end up feeling weaker because of it so the purpose of the lesser husks that we added later on is specifically to remind you that you are not the bottom of the power rung in this universe you are strong and these are the weakest things in the world that you can dispatch easily and quickly and a lot of these features became ways to invite players to really embrace that power fantasy even if it was their first time using VR and we found that was pretty crucial to get people into the experience so they didn't immediately feel like it was something that they were bad at or should feel embarrassed of while they were learning and even if you're new to VR or a veteran it's

really important at least we found that having this strong player fantasy and teaching this to your player and having them internalize it it it helped really reinforce the kind of most important part of the game which was the combat and what we mean by that is that as soon as the player kind of started believing that they were yes a hero yes they were a badass magical night they started acting like it and it manifests in subtle ways but they're important so for one you know if you're being told that you're heroic and you look cool and you feel cooled you start standing up a little more straight you uh strike heroic poses whether you mean to or not and then when it really comes down to the combat you start taking bigger more exaggerated swings which not only plays better and works better with the hardware that you're playing on but it also looks better feels better and is just awesome it's like for people both in and out of the headset it's makes this you know kind of goofy thing into something actually heroic and enjoyable and also as a bonus it's really helpful

for scope control so for until you fall if the feature did not help make you feel powerful or worse made it made it undermine to that feeling we just straight-up cut it or in cases where we they couldn't cut it or really kind of thought that no this could be cool if we just did one more thing we actually just gonna try to change the rules for how it worked and this idea of changing rules or setting up rules for what our specific virtual reality was became another big mantra or or something that we tried to carry forward into production because really what we were asking is what is our games baseline fidelity of interaction and to really dig into that we found that this is something that is actually a pretty hot debate right now in VR development and in the VR gaming community because the nature of VR creates this bottomless pit of

expectation for my new fidelity and functionality like we suddenly now have this universe where we have hands and the hands can touch everything and we can pick things up and we can manipulate them and we as developers need to pick the point where no you you can no longer do that in VR because it's almost impossible to meet every request for fidelity of interaction and for until you fall during development we did start to get a lot of that and we got a lot of playtesting requests for features that we deliberately walked away from either because we didn't have the scope for it or because we knew that it just added a lot of noise or ways to fail to the game so it became this big sticky question of how can we preserve that sense of immersion and that that selling point of VR without blowing our game scope and one of the ways we kind of inadvertently

early on done that was by rewriting the rules of reality so you can't throw up the sword was actually our first stake in the ground at trying to solve this problem and as an unintended side effects players were no longer asking for holsters because they knew where swords lived in this world they were in your hands their magic swords with magic hands obviously so you don't need a place to put them and based off of that kind of initial change we started establishing other laws of her worlds to head off some more problematic requests that we had noticed so for example looting and fighting in our game is always separate and that helps solve the problem of picking up the wrong thing at the wrong time there's no way that an enemy will sneak up on on you while you're looting and there's no way to accidentally grab a piece of loot while you are trying to fight for your life we also kind of shrank the size of our combat arenas to make them small and focused and make sure that the exit is always clear and in our world because you got magic

swords these weapons have super abilities and what they really are these kind of catch-all grab-bag powers that actually let you do things that otherwise would have been too risky or too skokie to actually implement into our game so with all this combined it really changed the way that players started at requesting features and requesting things in our game so in our world rather than asking for a way to throw weapons because you can't drop swords in this game it became a request for super that let you throw the weapon and it was kind of great because it helped recontextualize a lot of things and make requests a lot more sane even if we couldn't entertain every single one and that was because we had established to this virtual reality where the nomenclature for weird things that you wanted to do was your supers and when we took a big step back from that and tried to figure out like okay well what's the what's the system what's the process underlying this iteration

that we're doing when Dave and I talked about it it was really that we were deciding on the fidelity of our world and we were deciding the smallest interaction or that was vital for the game and our recommendation moving forwards is to take a look at what that baseline fidelity is for your game and consider cutting interactions that fall below that line in terms of gameplay importance because things that fall radically below that line may not be missed and in cases where you can't do that we found other ways to remove wants from playtesting requests that we really couldn't support from playtesting for example and until you fall we had a lot of playtesting requests initially of players wanting to pick up enemy weapons and use them but as soon as we made weapons dissolve immediately after

killing an enemy without letting them touch the ground those playtesting requests disappeared and it wasn't that we deprived players of the privilege of looting these weapons it was because in this universe clearly weapons dissolve but and I'll throw this out this is still something like we said before that is kind of a contentious topic in VR development until you fall it gets critiqued a lot for being a VR game with a narrow feature set and some people really love what the game could accomplish because of that feature set but it's definitely an ongoing conversation so we're not saying that going narrow and and being very laser focused on your features is the right way to go about VR development it's just what ended up working for our game and for our team so with our limited interaction set and kind of our more focused core gameplay it becomes

really important for us to then incentivize the interactions that we do have and one of the ways that we really found helped us out was actually incentivizing our player fantasy and put it in another way we've really looked for ways to find how to make the fun way to play our game the mechanically the best way to play our game and that's really exploiting this idea that something's not dumb if it works and that was pretty important to us because we saw early on in our play testing that for especially adults and teenagers and especially players that are new to VR they can feel self-conscious while they're playing and that affects how much fun they have and the type of poses that they strike while they're playing and beyond that fatigue in focus will naturally push players towards doing whatever the smallest effective action is for the game aka we waggling and until you fall because we wanted to underscore our power fantasy

we ended up building a lot of detection to isolate and shut down these waggling strategies because we wanted to change our smallest effective action and this actually had a really obvious and dramatic impact on our play test and actually became something that play testers called out as their favorite part of the play test in hindsight they loved that they couldn't choose the mechanics or they loved that the game's tutorial wasn't lying to them about the best way to play the game and we found that for players that really internalize that and believed what the game was telling them about making big swings they were more likely to strike those powerful poses and get into that play acting and that was because to them it was no longer a silly act or sore play-acting with the game it was the effective smart and optimal strategy for playing and we

found that that actually carried over not just from our firm or more organic like our mechanical swinging features but into the game stats as well that once we had this trust with the players that the games mechanics mattered we found that we could use our weapons stats to enforce and create a number of these optimal play styles that players can aspire towards so when we were doing our weapon design we actually created for major archetypes and we assigned - two of these archetypes - each one of our weapons the bruiser was all about big hits and big swings and dramatic motions the spell sword is all about supers and nooks and big damage sometimes from a distance tactician was for players that really liked manipulating the space so dashing around or locking enemies in place or pushing enemies around and rogue was sort of our high mobility probably VR familiar

player who wanted the high-risk high-reward or height which skill strategies and once we had these play styles and places targets we then made it a point with the game's traits and weapons to ensure that these archetypes had multiple effective builds and we used built-in traits on the weapons to make sure that the way the weapon was being used during our play tests actually aligned with that fantasy of the weapon and at the end we found out that this really benefited all skill levels of play this kind of balancing and this kind of mechanical enforcement of your fantasy is not something that the only the hardcore players are going to appreciate but that a lot of our first-time VR users really appreciated because this isn't about balancing the game for the hardcore it's about giving players permission to play act with your game because that play acting is the

optimal way to play so we're just about at the end now before we sign off I wanted to just kind of go over the takeaways one last time so we remembered what we were just talking about so five big takeaways of this talk were it's okay to referenced on VR games be wary of secondary VR interactions know your player fantasy set the rules of your virtual reality and lastly incentivize the player fantasy that you've established just kidding there's a sixth one so if you can be sure to try to consult with others in the space or Jean right even if it outside of VR Jason Vandenberg kind of came in early he's you know well known for his expertise in melee bat games and he was able to give us a really helpful advice early on and we would definitely recommend for you to do the same in any case just find someone that you could rely on to give you good advice so big shout out to Jason

Vandenberg and that's it so we can't give any we can't take any questions here in the digital format but if you'd like to talk with us more about any of the stuff we've got our emails and our Twitter handles on-screen and please don't hesitate to reach out thank you so much for listening everyone you