The Holy Grail of Multiplayer Level Design: Casual and Competitive Maps

[Music] uh well my talk is the holy grail of multiplayer level design it's levels for competitive and casual play briefly about me my background comes from the unreal tournament mod scene um especially unreal tournament three where there was the kind of hardcore old school low poly map pack um and also i'm a level designer high res studios where i do multiplayer level design mostly on paladins but a little bit on spite and some other stuff so before i get too far into the talk uh some definitions because i realize competitive and casual are kind of loaded terms when they're used uh in like vlog circles and by competitive i'm really talking about esports i'm talking about tournament play and by casual i'm basically talking about everything else i'm talking about recreational play but these definitions do get a little bit fuzzy because i'm also talking about the motivations of competitive play and the motivations of casual play so like you could be a very skilled

esports professional and choose to play to goof around with friends to explore to have many of the different sort of casual motivations for play and likewise it could be a new player who's still learning the game and are playing to win and playing to improve in a sort of competitive mindset and when i'm talking about the holy grail i'm talking about maps that really serve both of these audiences serve this huge range so with that in mind we all want to make kind of the greatest multiplayer maps right so what are the greatest multiplayer maps of all time and for some of you what may come to mind is facing worlds from unreal tournament 99 this is a map that for those who haven't played it it's captured the flag it's these two bases out in space on this asteroid they're frantically going across this sort of narrow pathways and there are sniper shots going the way and there's camera guided rockets it's just ridiculous and frantic some of you may have thought of dm 17 or longest yard from quake 3 arena this map

also happens to be in space it's very open it has jump pads as portals and it's also kind of dominated by the hit scan railgun which means that the only way to avoid that is using these movement mechanics on the map and it's very fast some of you may have thought of blood gulch from halo this map i guess it's kind of in space but it's more of an organic landscape and it's one of two maps in the original halo multiplayer that had vehicles in it or for multiplayer at least um so some properties of these greatest of all time maps one is that they have these big open environments that makes them pretty easy to learn pretty easy to navigate they also have pretty iconic themes or really recognizable themes there are bases there are sides to it or it's in space and they also generally offer sandboxes of play these maps have different mechanics that you don't get from other maps like with facing worlds you have the redeemer rocket and with the halo map you have the

vehicles but i also mean sandbox of play as these maps are really good foundations for custom games and for mods like probably if you played facing worlds you did slow mo with really dumb ai and instagib rifles and all sorts of nonsense uh but these maps also kind of create their conflict through their simplicity these are maps that on facing worlds if you're running to the other side and another player is running to get to your side you're going to run into each other it's a pretty simple map but somehow through all these properties these maps aren't really played competitively like they are played at lands but they aren't played in tournaments they aren't played for cash money um so what maps are there sort of competitive counterparts for unreal tournament it's maps like dm malevolence this is a sort of dense interior space with a sort of three arena format where each of these arenas has power weapons and power items that you want to control and really it's it's a lot more complex and there's a lot more strategy to

moving around this map for quake it's maps like lost world or arrow walk or blood run these maps also have that kind of dense interconnectivity and these a lot more complicated pathing for halos maps like chill out and damnation which are also this kind of three arena structure very dense this one has portals which is a main source of its complexity and about controlling items so if you think of these properties again they're complex the themes kind of stay out of the way and can even be disorienting with their simplicity and it's a focus on core mechanics you don't see the nuclear rockets you don't see vehicles on these maps and instead of the conflict through simplicity it's it's about item control it's about moving around the map for power weapons and that means if you're playing these maps for the first time in sort of a 1v1 with another new player you might run circles and never see each other and as a result some of these maps aren't played casually or they aren't the maps you play when you're first learning these games um

so what's going on here because none of these maps are bad and in fact all these maps are awesome at what they do one way of thinking about this is kind of an overlap of audiences and this model here is from david surlin in 2009 talking about fighting games where he argued that for a fighting game to be successful you need to have that strong hardcore competitive scene and really cater the design around that and then the casual audience will kind of follow and congeal around that and if you fail to serve that competitive audience then the whole thing kind of dissipates and it won't be very long-lived game we also see this with mobas where by really serving that competitive audience and creating kind of the virtuous cycle of esports viewership ranked game play we can get players engaged and retain for a really long time but this isn't something that's true of all games like if we think of platformers if we think of speed running as kind of the hardcore competitive equivalent um that's kind of an outlier right like that's not the more extreme version it's it's

almost a completely separate game and if we as designers are trying to create a game that really serves the maximum of these audiences and we do that with platformers we're going to end up in this really weird niche that doesn't really serve anybody so that's something to keep in mind also this isn't just first person shooters although that's what all my examples are from um we see as with action rpgs we see this with mmos where there's a difference between the hardcore players and with um with new players we also see it in the real world with for example this is from paintballs the example on the right here is a speedball variant of paintball where you have two teams of five and each player fills a role and it's very very competitive structure it's symmetrical and it's balanced you're playing to win and on the left is scenario paintball where these players are reenacting historical battle and because it's a historical battle you know who's going to win and who's going to lose right like these players aren't playing to win they're playing for very different motivations

another way of thinking about this that's um kind of the uh the way that casual and competitive and casual and hardcore gets thrown around is often terms of oh those filthy casuals those low skill players and it's actually a lot more complex than that because at least for first person shooters and with most most games there are these kind of games within a game and it's layers of skills that players are developing and for us it's the first layers mechanics where it's just moving and shooting and being able to navigate 3d space and being able to even use a controller and a mouse and that's tens or even hundreds of hours of gameplay there that sometimes we undersell as designers and once you have basic competence in that which again is maybe tens of hours then you can start learning map specific and game specific knowledge and tactics and like how to use items and how to move around the map and what angles you need to control and then from there once you have competence and again that's another huge arc of mastery then you can start worrying about sort of tournament level strategy which is the meta game character picks bands maps that

sort of stuff and all while doing this you're developing some parallel skills that are your ability to communicate with your other players your ability to handle the psychological pressure of competition your ability to not tilt in competition but the key thing here is that as designers sometimes we have a bias towards sort of the tactical strategy side of things and there's a huge amount of gameplay we can serve just by focusing on mechanics another way of thinking about this briefly is in terms of psychological models and traditional psychology the one that i found most useful for thinking about this is self-determination theory which basically argues that there are you can categorize intrinsic human satisfaction into competence autonomy and relatedness which in game design terms is kind of like thinking mastery choice and teamwork you can also think of relatedness as kind of relative competence or relative autonomy and the important thing here is that that this doesn't line up one-to-one with some idea of casual and competitive and that's kind of deflating the

the more mainstream filthy casual use of casual um but still taking all those different ways of viewing this problem in mind what are some ways that we can support casual play one is with sort of rebound and chaos mechanics um so the blue shell and mario kart is a really good example of this because it's an opportunity for a new or lower skilled player to catch up and subvert the experience of a better or more experienced player but with first person shooters we can also do this with random weapon starts random weapon timing objectives that are random we also want to make sure that the maps are easy to learn so this example here is kind of a rough abstract diagram of chill out which i brought up earlier and the green dash lines are the portal route so even though it's a very simple layout that's not too hard to understand those portal routes make it very confusing for new players and can be very disorienting we also want to kind of put a ceiling on tactics and strategy and this is what facing worlds and dm-17

do really well through the simplicity of their layout there's not really room for tactical play or at least not not at the same degree that we see in tournaments but sort of in conflict with this or something we have to balance with this is offering up a sandbox for players because not everybody's going to care about the same things and everybody's going to learn the different mechanics at the same pace so halo does a really good job of this with their vehicles because some people like driving more than they like running and gunning some people hate driving and they don't have to use those tools if they don't want to it's a really good way of exploring different play styles and it's something that the battlefield games have done an excellent job of serving on sort of the more technical level designing side of things we want single scan arenas basically what i mean by this is xero scan is kind of like walking into a closet that's small enough that your field of view encompasses everything and all the information is known as you walk through it a single scan is a little bit like walking into a room where you have a blind spot on one side

in order to see the whole room you have to scan and reveal that information a multi-scan room if you think of that sort of shaded square in the middle as a pillar you have to keep moving and keep looking in order to maintain information so we want to keep things simple we want to keep single scan that's basically how facing worlds and dm17 and blood gulch operate so some ways to support competitive play kind of in contrast we want multiple options we want complex options and even what i've drawn here is pretty simple but like the examples to the right of that are they look a whole lot like facing worlds actually um and that doesn't generally lead to a lot of tactical variety for players and can turn the competitive game stale we also want to make sure these options are resilient and basically when you get into some state you can recover from it and that means like you don't want to spawn spawn camp for a whole match but also means like in moba's you don't want to fall behind on the farm and

never be able to recover and this is really with esports in mind because we want dramatic games we want stories and one mistake early on is if that means the whole game that's not very exciting to watch we also want those multi-scan arenas again this is kind of the foundation for a lot of that tactical play if if multiplayer maps are just the single scan arenas and we're all very simple spaces then it would just be about who has the better mechanical aim who can hit their target most reliably so what about the holy grail maps what about these maps that do both of these things um you probably know that there are a couple of maps that i haven't mentioned and maybe to some of you that seems really weird so dust2 for example is a map that uh yep dusty is a map that does boast these things very well and serves both a competitive and casual audience

and that's partly because it has just the right amount of complexity that you can form different tactical options but it's also very clean geometry it's 90 degree angles it's very open clean spaces there are a couple curves in the map but for the most part it's on 90 degree angles for halo there are maps like lockout which it's a very dense space it's multi-tiered space there's a lot of verticality to it but it's also these kind of additive structures floating in space that are connected by bridges which makes it much easier to understand where you are relative to other structures and very easy to navigate it also has the added bonus of exploding barrels and like this jump pad elevator thing so like how can you not have fun with that um for quake it's maps like campground um and there are a couple others but campgrounds is the one that uh really fits this the best and that's it still follows sort of the competitive format where it's this three arena structure and these hallways interconnecting them in sort of dense ways

but doesn't rely on portals it's not overly vertical which makes it pretty easy to understand each of these spaces and how they connect and lets you navigate the level pretty quickly it's not hard for a new player to pick that map up so some properties there's kind of an intermediate complexity to these maps they're readability focused environment art they're really aiming for some really strong thematic but aside from that it's kind of like there's not a lot of overlap these maps are all kind of specific solutions to each of their game's designs we can't take dust 2 and put it in and halo and expect it to work we can't put lock out and counter-strike and expect that to work these maps are specific solutions which is if we're trying to make maps like this then what can we take from this and also some of you are probably wondering why i call them holy grails to my knowledge this is just kind of the term i personally use for it and part of it is this idea that the holy grail grants eternal youth that it's um in

terms of multiplayer games you don't need a lot of these really good maps in order to keep your game alive like there are still 24 7 dust 2 servers that's incredible and you know especially with increased fidelity expectations and the cost of delivering that fidelity being able to create a couple of really good maps rather than 50 okay maps or 50 specific audience maps that's really good if you can do that but i also call them this because the holy grail isn't real and like my example with platforming and speed running you may end up kind of serving a niche or running into some problems with that so another way of thinking of that is kind of the ultimate vanilla where basically it's a flavor that doesn't offend anybody and everybody can bring their own personal flavors to vanilla and it still works with vanilla you can you can have peanut butter vanilla you can have sour gummy vanilla and it's still yeah so some cautions around pursuing a

holy grail map so now i'm going to talk about paladins and sort of my quest for a holy grail map so this is paladins we are a team-based first-person shooter we have two teams of five each player is choosing a new a unique champion with their own abilities and um each character also has a whole bunch of customization that happens in this kind of out of match loop we also have objectives uh there are capture points and their payloads are pushing and we're also following the sort of free to play games as a service model so we're constantly trying to update and make it better and it's not like we've shipped it and we're done and we'll do better next time it's like no we're still trying to find ways to make it as good as we can so some history just for context for what i'm going to cover um paladins really started as more of a moba with the first person perspective and then moved into more of the class-based first-person shooter it is now and part of that is there were a lot of perception problems when the game went

to closed beta as this kind of mobile with first-person perspective where um there was just genre conventions around first-person shooters that weren't quite being met and part of that is that we moved from a high time to kill to a medium time to kill we basically had a time to kill around 15 seconds which is insane and moved to somewhere between like three and ten depending on what characters um i forgot to mention but we also have classes we have like healers and tanks and dps and flanking characters so like the 10 seconds is like two tanks brawling versus the three seconds as two flanks having a duel uh we also through the process of closed beta went from this sort of chaotic unpredictable gameplay to something a lot more deterministic and this was really in service of our competitive scene and there was a whole lot of what just happened to me moments in our play testing that we were trying to resolve and basically with all these design changes for level design we were trying to sort of keep up with the game design and that meant going from these sprawling

moba inspired maps to a lot more traditional tight fps style maps so basically we went from maps that look kind of like that and this is from the summer of 2016. this is after we shrunk the map down multiple times and this is also a view of the map that's only showing about a third of it and we went to maps like this which is much more recognizable as a first person shooter style level for perspective that's about the size difference that's enchanted forest on the left that's kind of the old style map and again that's after we shrunk it down multiple times and that's our kind of current approach to level design on the right um so through this process of taking those mobile maps and shrinking them down and taking and making new maps that style we developed a set of principles one thing is that we want strategic variety we want to make sure that every time you're going after an objective or doing any kind of play on the map you have multiple options open to you

we also want to keep the maps pretty simple and this is mostly speaking to our character's abilities we have larger than life dodge rolls and aoes that can fill the capture point so we want to make sure that players aren't tripping up over little bits of geometry that don't matter for our kind of gameplay we also want to center on the core mechanics we found with some of the maps that they ended up being kind of sniper dominant and our players didn't respond very well to that or at least our competitive z and our competitive scene didn't so we kept it focused on that sort of healer tank dps holy trinity um and we also want to limit zoning and allow recovery this is going back to the idea of resilience in our sort of competitive principles and specifically with these older maps the most dangerous locations was the route to get to the objective because of the way those maps were designed and just the constraints around that kind of design is that whoever won the first fight for an objective could then push forward and zone and prevent the other team from even be able to contest and that was the optimal strategy and

you'd see it in competitive play and it was just oh it was it was rough another principle that we have this is a little bit more on the environment art side is that we want to make sure that what you see is what you get we have characters that can fly and we want to make sure that as you're going over rooftops you're hitting invisible walls and that really defines a lot of kind of the level design problems we run into if we rephrase this sort of thinking in terms of our casual and competitive and holy grail divide most of these things were in service of our competitive players like even the idea of simplicity sure that helps casual players that helps that motivation but a lot of it is still like just the nature of the game so some results of these iterations as a team of level designers we felt a lot better about our maps they were better a lot easier to make they're more fun to make and they're a lot more fun to play test and our metrics were up we had better retention and engagement from our closed beta feedback and we had a much healthier competitive scene where it was actually engaging and exciting to watch our

players compete because there were multiple strategies available to them and it was exciting but in terms of subjective sort of anecdotal player feedback it was a little bit more lukewarm than we had hoped so basically what we thought we did was something like this we thought we had that kind of moba or fighting game situation where if we served our competitive audience we'd be set and we could just keep making the best competitive maps we could and we'd just keep doing that forever and the game would be awesome but i think what we actually realize we have is a bit more like this where a competitive scene is kind of an outlier to what a lot of the more casual audience wants from the game and with our change to these maps we ended up missing the mark a little bit and losing some of that audience or at least we still have players on our forums saying i kind of want those old maps back which to us as designers is like oh no so what some of those players want is really more variety they want bigger maps they want asymmetry which is another way of getting variety and they want the stronger themes one of the best

things about those old maps is that you're besieging an enemy fortress with a big siege engine thing and it's like the theme of it is really rich and that's something that we've kind of moved away from so a way for us to try and understand basically this casual competitive divide better is with public test queue this was a queue that we set up for this last year we only recently shut it off but it was running for basically a whole year of public gray box maps and if you're a player who was at level 14 i think it was you could play testing these maps and after each match you'd get feedback forms saying what'd you think of the map and sometimes we had other specific questions we were asking as well but the main idea is that we wanted to better understand our audience better understand kind of that venn diagram and we also want to test specific experiments which i'll go over in a minute we also want to reduce the risk of creating sort of lukewarm maps or these sort of false grails so this is sandbridge this is a test queue map obviously um this is a map

which i designed specifically to challenge some of the design principles that we had set up over this course of iteration i created these really elaborate flank routes and these really dominant sniper towers and there are portals in the map and i thought players would hate it but it was actually one of our highest rated maps in the test queue at the time but also we had competitive players coming to us and saying this is why i don't play the test queue this is terrible and that just kind of highlighted this divide that we'd started to realize so the map i made after that and trying to and trying to understand it's better was sewer which uh hint if you call your map sewer yeah um so i designed this map specifically for healers and tanks and i want to really boil it down into that calculus of just healer tank math and really limit the options for flying characters this is kind of no clipped into the walls so it makes it look like it's open space but it's actually all enclosed so you can't really fly

anywhere you can't really flank that much and yeah on average players didn't like it so much and this is corey this is after we went through [Music] the sort of gray box pass this is this is the final art as we released it and with this one i wasn't really testing any specific experiments i wasn't running any wild ideas this was just me trying to make the best map i could and use the test queue as kind of a vetting process to make sure we weren't releasing a lukewarm map and while it was popular in the test queue once we actually released it with the final art we found that it was a little bit lukewarm like we did get some weird feedback that we hadn't expected players were confused and disoriented they felt claustrophobic and partly that's speaking to most players inability to evaluate gray box maps but it's also speaking to that divide in our audience and that we can't do competitive and casual both so that's the case what's next for us with this realization

that our casual and competitive scenes don't totally overlap and we can't just make awesome competitive maps forever we've created new game modes and we have a team deathmatch mode and we have sort of king of the hills style mode and these are opportunities for us to create maps that are really in service to casual audiences we can do a lot more map dynamics we can have asymmetry we can do all sorts of wild crazy things and our competitive players won't care because it's not for them it's awesome um but of course we're still moving forward with that to try and figure out how to best answer that question so kind of the main takeaway from all of this through this process the main thing that i learned was this kind of audience check so basically every time i am designing something new i ask myself this series of questions who all are we designing this for what are their needs and again think back to like the self-determination theory competence autonomy relatedness what what are we really offering up with

this map of this game mode and how does our design serve those needs and because these questions are kind of hard to answer there's kind of the follow-up question of what steps can we do to answer these questions and that's really what we did with the test queue so basically i want to rephrase the idea of us all wanting to create the greatest map of all time or a map that belongs in the ranks of facing worlds and instead think of it as designing maps to serve our audience that's my talk you