Puzzle Game Magic Secrets

[Music] hi everybody uh this is puzzle game magic secrets let's talk about puzzle games uh first uh who am i i am brett taylor that is my name and my identity i got my master's degree in cognitive studies and education from colombia and i've been developing games for actually i guess it's over 10 years at this point i founded my dog zoro my company in 2015 and my company my dogs are made limelight uh so let's talk about limelight briefly uh because this is uh oh actually quick show of hands who's familiar with limelight okay all right about that cool all right uh so landline uh it's my dog zoro's first release uh it's a puzzle game not surprisingly uh and we released it in 2017 for playstation 4 steam ios and android uh in 2017 and 2018 after taxes it earned about a hundred thousand dollars so you can do the math of the annual salary of that uh it also did pretty well uh review-wise there's over uh 10 000 uh google play reviews 4.9 stars etc etc so here's the trailer so you get an idea of the mechanics if you're not familiar [Music] [Music] [Laughter] all right so that's limelight okay so overview for this talk it's broken up into four main parts first we'll we'll talk about working memory uh then we'll get into some puzzle characteristics like vocabulary for puzzles um uh so we can like have better conversations about them uh then the bulk of the talk is the lessons that i learned from making limelight and there were many lessons then we'll end with uh how i make puzzles so some actionable takeaways actual strategies like toolbox of strategies that you can use and take home to make your own puzzles and stuff all right but before we get started this talk prioritizes fun and non-random puzzle games so you might not want to make a game that is fun for specific reasons if it's like an artistic thing etc this talk um sways towards that just figured i'd mention that uh and you can also subvert these rules for alternate effects so if you want to like intentionally disorient the player for some reason you could do the opposite of what some of these like rules that i say are um and also this talk i cut i think like 300 percent of it um there's so much that i cut from this and there's way this is just scratching the surface there's a lot of so yeah all right uh okay so definition what makes a puzzle fun so in this in this talk i i define fun as when any time the player gets to exhibit mastery skill or wit this is a theme i'll kind of come back to um also it's a definition of mechanics when i say mechanics things like boxes that you can push and pull enemies etc etc things like that um so most game developers when they hear the word mechanic the first thing they think of is this but that's that's erroneous no no no today in this talk we're thinking about this on the right uh yeah all right all right uh part one working memory working memory human ram this talk this slide used to be much larger that's basically it this is the fastest way i can explain it to people um it's uh it's the processing power that you have to hold information in your head and use that information to solve puzzles and do other things in your life um so the way that i sort of um in the in this talk the way i picture it is it's uh this is like an abstract metaphor of uh if your working memory were a bar uh from zero to a hundred percent so things that you do um uh occupy your working memory so let's say if you're solving a puzzle you're just taking up a certain amount um you're talking to a friend and you're also eating each of these things uh it's into your working memory so let's say that you know if you're juggling or something it'll be harder to do math problems because the juggling activity is occupying your working memory and this um has big implications in what players are doing in a puzzle game so who here is familiar with with flow uh flow okay so most people uh yeah so uh this uh when i sort of like found this analogy i was really excited about this uh so um so with with flow is basically um for those of you who don't know or that familiar um if the player has like uh if the level's too challenging and the player is not out of skill then it's overwhelming and on the opposite side if they're too skillful and the level is too boring then uh then it's boring and that's no not fun so working memory wise we actually kind of like there's kind of a sweet spot to keep a player um uh in terms of like how much stuff they're juggling in their head at any given time and it's sort of like it's not quite in the middle because the human brain can hold a lot and too much is pretty uncomfortable um but like there's a sweet spot sort of um i don't know in that range again this is like this is super metaphorical like this is not literally this obviously our brains don't work like this but it's a it's really useful um at least for me to like to think of it uh think of it like this so like how much working memory am i overloading or like am i requiring for my players all right after repetition uh concepts or ideas in in puzzles uh they take up less space in a player's working memory so let's say that rule a initially will take up forty percent rule b uh takes up another amount a rule c takes up whatever twenty percent as the player gains experience over time these each of these rules takes up less and less space in the player's working memory uh and then there's all this freedom space which naturally is filled with new mechanics that's you know just sort of how uh how games work uh so here's a puzzle from the witness for a beginner who is unfamiliar with the mechanics of this um you know this would be let's say i mean this is this is metaphorical again but uh yeah like this would this is overwhelming for that player um there's too much going on they don't have enough experience yet but for somebody who has more experience and has played enough of the game they can visually parse through each of these components takes up less space and that player's working memory so for an expert this is much easier to solve and um where they are stimulated working memory wise is in a much more manageable zone and that's sort of the sweet spot that keeps them that's the spot that we it's the most fun as a player to be in all right so if an expert you know let's say that the purple rule and the yellow rule only take up that much space but for a novice it takes up this much space um what is all of this extra space great question i'm glad you asked uh this is what i considered possibility space so possibility space it's reduced as the player's experience grows as we you know as i showed in the previous slide um so here's an example from the witness you don't have to know the mechanics of this but an expert who looks at this knows that they must draw a line through the center of this it has to be horizontal there's that's the only option that they have but for a beginner who's not quite as well versed with the mechanics of the game might think that this is possible and that that uh extra possibility takes up a little bit more space in their working memory um this could also be possible and this and this and this and this and that and etc so all of these things um add up to the novice's possibility space so um so something that is very is relatively trivial for an expert for a novice it actually would take up much more working memory because they haven't developed they haven't streamlined i don't know the neural pathways um uh that would make it make them visually parsing it and understanding it uh more efficient and the expert is better at it let's drink some water all right all right uh moving on puzzle characteristics here are some traits of puzzles so first off braid doesn't equal sudoku what revolutionary statement i know why well i know everybody's very antsy they want to find out i'll get to it please um all right so the way so i sort of put this on a spectrum of handcrafted puzzles versus procedural puzzles now there are other types of puzzles um uh another tons of other characteristics and components of puzzles but the most popular um spectrum that i see that is easiest to categorize most of the puzzle games that i see um and that are on the market is um sort of like somewhere in the handcraft and procedural realms uh so handcrafted puzzles uh there's a finite possible amount of puzzles and because they're handcrafted uh the player is um rarely repeating skills uh it's much more novel like each solution is more novel and surprising that's why we've got these aha moments um that we you know strive for and talk about and the content is authored which is uh one of the most important parts for the developer um because you are like you know handcrafting each one uh and it's a much more fresh and varied experience for the player um that's sort of why they're playing the games to see what's you know one of the reasons is like you know what what on earth will happen next uh procedural on the other hand it's basically the opposite uh there are infinitely possible puzzles um for the most part uh the player is constantly repeating skills so if you can picture playing sudoku you're you know you're not learning new skills there's only f there's a finite amount of sudoku skills um um that um and strategies that any one player can have uh and each you know solution is not that novel relatively speaking you don't like i've never like gone to my roommate and been like oh i have this like sudoku bird and like look at the solution it's like it doesn't really happen uh and the puzzles are born from the system so that's uh uh from the designer's perspective you can theoretically make a random generator that would make um these types of puzzles i've worked on lots of puzzles on both ends of the spectrum and in between um and for the player it's a much more predictable experience like you know if somebody's playing sudoku or a game like that that's um generally it's like part of a routine like maybe a daily routine you know like when i played sudoku for a while i'm playing on the subway on my way to work etc because i wanted that specific experience so these these are um just two different types of uh experiences and then there's uh oh yeah so games like rush hour also and flow on the app store those are both great games um are on the procedural end uh then there's tons of stuff in between the witness was kind of tricky to categorize because it is by its definition it's handcrafted but um this system itself is lends itself to procedural puzzles there are points in the game where you know they're just randomly uh randomly generated levels uh all right oh and yeah this talk um leans towards handcrafted puzzles uh or at least like we'll talk a little bit more about that because limelight is in the handcrafted handcrafted side of it noise so if you take anything away from this talk uh this is like the most important thing uh if i if i were to give you anything this would be it uh noise the way that i define it is it's anything that occupies the player's working memory but it doesn't contribute to the puzzle noise steals working memory from the player so briefly here's a mechanic from limelight these are called flippers and moving streets so you move left and right on the little hamburger thing and the street moves up and down so here's a level in line lights that is all like this level is mostly noise so the solution to it um the solution this puzzle is get three flippers in the right setting to complete a path that's like if i were to describe what the solution is that's the solution um but because of all the noise this level is basically like a big it's sort of like a big slip knot uh if i if i pull the slipknot apart this is all the puzzle really is you just have to hit these two flippers and then you've completed the path but i've added a lot of noise to sort of obfuscate that as a as the obvious solution and this is really in the game and i use noise pretty sparingly but here's a it's a good example of how i use it all right so um so here's one an example from the witness so in this uh this this example the player starts here they want to draw a path to the end they can either end on the top right or the top or the top or to the bottom right um and that's it just draw a path from start to finish uh all of these gaps uh in here that is just noise this could be those could be replaced with ink plots um because you can't actually pass all the way through them uh they're there to make it more difficult to visually parse that puzzle uh photoshopped out this is what it actually looks like and it's so much easier um to uh to see what the solutions would be because there's much less noise um and mazes like in themselves are just like giant slipknots in the first place but anyway uh so on the left yeah like let's say that you know takes up theoretically this much player's working memory that noise occupies a lot of extra space but on the right without it uh it's pretty trivial you can see um how to get to the end much more easily uh here's another mechanic from my light these are called cover-ups if um all of them are being touched then the gate opens up so here's an example from limelight this is a level that nobody solves accidentally because there is noise in this level uh so in this example uh it is um no i use noise to prevent the player from accidentally solving the solving the level uh the player the solution is for the player to back up into these two cover-ups to light them up because they can only control the head of the character um and uh i guess i'll just challenge you folks and see where where could you see the noise um so the solution is to back up into the top area here to unlock these so you can continue and get those any idea where the noise might be i don't fully understand this one here yeah i mean it's it's tough uh yeah stuff if you haven't played the game uh so so the answer is i mean if you're not as familiar with the game it's a bit trickier uh but the uh the noise is right here uh so it's this extra loop right there uh yeah and you're celebrating that very very well done uh so with uh um without that noise so people basically would go down um because there's only a few things to do like that's a dead end why would i go there they would go down and then they would back up um and the puzzle would be solved and i didn't realize that they had solved it or what they had done to solve it uh and that completely subverted the point of the puzzle because i want them to know what they're doing uh unless you want to actually be interesting if you made a puzzle game and the whole thing was like just the player accidentally solving each really interesting puzzle but but i don't think they were here to do that so uh so anyway so that's that i added noise intentionally here um to uh to prevent that uh so here's an example from inside so here uh in this example uh there is a crate here that you can push to the right uh the goal is to get up this chain here the the player wants to continue up by climbing that chain when you push the crate to the right or sorry when you pull the crate to the left the chain goes up and when you push the crate to the right the chain comes down um so this uh like that so if you push it all the way to the right the chain goes down um but if you push it to the left the chain goes up uh i'll give you a moment to think about where the noise here might be you got it it's the fact that you can jump on this platform here it's a red herring this does not it's not part of the solution of the puzzle uh yeah you got it excellent i appreciate you uh um yeah yeah so it's basically it's a red herring it's not involved in the solution it does not have to be a part of the puzzle it does contribute to the experience but it's actually not it's not part of the solution and by that definition it is um it is noise um so noise makes puzzles artificially harder i sort of like to think of it another metaphor it's a sort of it's it's like the breading in around general styles chicken um like a little bit is okay uh but if it's just all breading if it's all noise um and there's no meat it's like where's the puzzle uh you know i like that metaphor whatever uh and it does it requires diligence to identify like i i've played so many games that have a lot of noise in them and um uh so please take this away and look out for your noise and remove it uh yeah if you're not aware of it noise will pollute your game i've played so many games that um unintended like they have noise unintentionally they like the developer will understand that it's making the puzzle more difficult but they won't um but the nuance of having it be because uh it is not contributing to the puzzle um they might be missing out on that and that is not a puzzle solving experience as much as it is a seeing through noise experience which is different from puzzle solving um and it can be useful as i mentioned if used intentionally uh by uh deterring accidental solving and also you know like using skill to see through it like in solving a maze or something like that or in a bullet hell sort of game that's that's satisfying [Music] all right seven lessons that i learned from limelight this is the bulk of the talk you ready let's strap in all right lesson number one simplify less noise leads to cleaner and tighter puzzles so what do we want cleaner tighter puzzles how do we get it exactly simplify oh yeah good call uh yeah so but to unpack that uh the method that i do is describe the solution of the puzzle and remove anything you didn't mention i'll let that hang hashtag cut the noise uh this gets easier with practice uh i got way better at this when as i was developing limelight like the early levels that i made which i thought were so cool like i'm just ashamed of uh at this point but uh yeah i mean you know it takes practice uh anyway so uh yeah as i was developing limelight i hacked every level down to the bone until there was really nothing left to remove there was no further extraneous elements and then i continued to remove uh extraneous elements because there were still some like i would make things symmetrical maybe when they weren't i would streamline this streamline that and it's very rewarding actually um yeah and looking back as i was making this talk uh i i had to i was like yeah i simplified it so much i could have simplified it so much more um so my advice is vet every level at least two times check uh you know are these components necessary if your goal is to remove noise which i would say it's a it's a good practice it's a healthy practice so that noise does not pollute your game um i've added each of line lights levels at least four times some of them i did up to a dozen times i would continuously just just constantly mow in the lawn and if it doesn't belong ultimately um it might just be worth cutting which i did a lot in limelight's levels uh and it's also it's just really satisfying uh to like i got kind of addicted to like just like fine it's it was like i would identify like oh here's a noise and then i would like remove it out and it just you know it's good it's fun uh so my voice is if you do want noise um try to start with none and then add it from there so that way you're at least doing it mindfully um because you know it can be used for good it's not all bad um it can enhance the experience then prevent accidental solving and change up the pace of the game and so on so here's an example on the left this is what the level used to look like and i thought that this was super elegant i i was like what else what's there left to remove um on the right it turns out that there were a bunch of things i could remove it's the same solution um so describing the solution they're identical between these two now here's another one so there's about as many components on the left as there are on the right but the right the level on the right is symmetrical which is it's much more pleasing to to look at it's much prettier and it's also easier to visually parse um so that's that simplified it there the one on the left here i was like i was like i'm this game is amazing i was like super excited at least about this mechanic that i discovered when i was when i was working on it but the solution ultimately is uh just on the right it's so simple and it took many iterations eventually i ended up with the level on the right which is fundamentally what the level on the left is trying to be but it's obfuscated with a lot of extra components and clutter which it doesn't have to be um the level on the left is general styles chicken with a ton of breading and the one on the right is mostly just chicken i like the metaphor i don't know why lesson two cut pointless levels this kind of ties into the previous uh they are there accepted um so for each level ask what is the purpose of this level and if you can't answer that question it is a road to mediocrity here's some possible purposes of a level uh maybe it teaches something um it could be to reinforce the skills so some redundancy is okay i do that a little bit in limelight there's a unique identifiable moment so it's just something that player hasn't yet experienced or it's maybe a palette cleanser or just a breakup pacing there are some levels in line light intentionally after a bunch of hard ones where it's just a big open space just to let people exhale i discovered that those were really important to have in there also here's a side note on the term mediocrity because i have to i have to say this you are good enough always i promise that's a vow from me and if you don't believe it come talk to me afterward and we'll chat a bit uh don't be afraid of not being good enough and make things anyway don't be afraid of making things that are mediocre please and also if you like every expert has to pass through mediocrity in the first place so yeah don't let that stop you from creating things all right back to pointless levels it is not easy to admit a level is pointless especially if you have nothing better to replace it which is how i dealt with limelight um uh in the beginning because i was like these levels aren't that great but there's nothing better to replace them with so i was like but over time i got better at making them um so my advice is that every level what do we not want real freeloaders yeah uh all right so here's an example of a level that i cut um so this level uh i cut because it was not educational interesting fun or unique uh most importantly was not fun it was preceded by these two levels so this one on the top left explores an idea the one on the bottom left unpacks that idea brings it a little bit further and then the natural trajectory would be the third level so i take one concept i take the next point in that and then what's the next what's the next logical progression of that uh of that level and it was this but it was just not fun it was redundant so uh i ended up cutting it i love water it's so much fun all right uh all right uh lesson number three you are the expert player controls wise so i had some complaints about limelight the timing being too hard even when they had like a full one and a half second window to complete like to do an action and i'm like come on like it's like the barn door is like all the way open and they're like hey help me into the store i'm like it's right there however um knowing the solution and not being able to execute it is not a fun experience yeah so if people are complaining listen to them my strategy is to play through levels with your feet with your non-dominant foot try that out uh and if you can't solve it with your feet make it easier uh this is specifically for puzzle games it's like obviously for action games this does not apply but if it's if it's a game where the promise is um that people if they can figure it out that they should be able to execute it um play with your feet uh i don't know foot play testing i feel like there's a clever term there i'll come up with it after the talk uh all right new mechanics from light lock steps they move when you move as you move oh yeah so uh here's uh me you know in 2016 thinking i've logged hundreds of hours playing this game so i'm pretty sure i know what the average player can handle um which was um erroneous uh the lesson that i learned from this uh is that if people are still complaining because i made changes i made it easier but people were still complaining and i thought to myself i made the changes already you know i've catered to the audience but people were still complaining uh they're right they're still right continue to change it um it's not them ultimately it is you it's the game uh serve your audience it depends like if your goal is to serve your audience if you're making a game for your players um don't lose sight of that priority lesson number four keep action and puzzles separate this applies specifically for single solution puzzles um so things where you're building a solution or if it's like a physics-based rube goldberg maybe solution um that it's inherent in the actual design of it but for for single solution puzzles or like you know few solution puzzles this is very important so action or puzzles is great action and puzzles is um that frowny face so you wouldn't serve tune in and ice cream on the same plate so why would you serve action puzzle solving in the same moment so that's a bulletproof argument right there i don't know maybe so action adds randomness and distractions but puzzle solving requires patterns and consistency the player needs an idea of what they don't yet know so action subverts the foundation of what the player needs the stability that they need to actually to solve the puzzle because you need patterns and consistency and action removes that uh yeah so when a player enters a level they should know if it requires either action or puzzle skills because you can do like you know limelight has like actiony moments um but it should be clear like are we in puzzle solving mode or action solving mode it does depend what you're going for but this was a lesson that i didn't realize was as important as it actually is um yeah and about 10 of players complained and that's that was too many um so make action levels optional that's um one obvious thing suggestion to mitigate this so let's say you know there's a 100 action levels on the left and 100 puzzle on the right we do not want levels in the murky area basically if you find that they have components of both and it's a single solution puzzle push them out the ambiguity is um it's it's no fun uh yeah which leads me into lesson number five make the solution unambiguous what's that mean let's unpack it here are some red flags the player says was that right or they say oh i solved it my players sound like that or a player is attempting an impossible action repeatedly uh because it's not obviously impossible and that's just not a fun experience i'm sure that like i'm sure most people here who have played tested games have seen this happen um if the solution isn't if the solution is ambiguous uh it damages the player's trust in the game and that um that sucks and you don't want that to happen um uh it makes toby maguire pretty sad so we don't want him we don't want him crying um so here's an example in limelight uh so this one was most like people ended up playing this through trial and error and it was not fun here's why on the left uh people would experiment they would uh so that the way to solve the puzzle here is to hit um to find one combination of these two switches so there's four possible combinations each one can be on or off uh and then the player would travel all the way through here i intentionally wanted that travel space because i wanted them to stop and think i didn't want it to be easy to excuse me brute force it to brute force it so i added all that extra travel space um and then they would experiment uh up at the top but players still brute forced it anyway on the left and they kept going back up top and it never it never was obvious oh this isn't the correct state yes i feel like this is right um uh and um and now i can execute the solution it was they they just continued to brute force until they got the solution and that wasn't fun so i ended up cutting this level and several others like it here's another example uh so he dies every time in this gif i hope someday he'll make no didn't make it that time um maybe the next one now can we make it it's so close ah damn all right okay so i don't think that's possible to make but it feels like it might be possible to make what i did to mitigate that was i expanded um this area here this is much taller and i also made this uh wider on there so it's clear that you can take refuge so players no longer attempted to so if you see the player come down there they clearly don't have enough time to hide in the in the bottom right line um so that that solved that problem so um one strategy as in the previous excuse me let's get some water again very fun come on water's fun you all know it never want water skiing yeah that sounds fun uh so uh as a previous example move elements farther apart that's a great strategy that's super easy got extra stuff cut extraneous elements that's like a theme throughout this talk if you're working on a platformer please make all of your jumps obviously possible or impossible this is one of my pet peeves yeah so uh i was um showing somebody um uh a platformer that i'm you know casually working on and uh realized that i had violated this rule very recently so this is the puzzle here and it's not totally clear if that jump isn't makeable if it's possible to ball kick and then bounce and get up to there uh it is possible if you time it like that and this is early in the game so the player doesn't have quite enough skill and confidence to uh um to to feel secure that they're doing it right so um the feedback that i got was i don't know if i'm a like i don't know if i should be doing this because i don't know if it's possible uh and that's just not a good experience because like is it a puzzle or is it am i just not good enough and not knowing that is not great you want the player to know either there's something i don't know yet and i have to figure it out or i'm just not good enough but i can get better you like if there's ambiguity in between those two the player doesn't know what they need to get better at and it's a frustrating unpleasant experience so the solution click drag and now way more obvious that you can make it lesson number six player trust players assume inherently that there's no agility needed in puzzle games that's sort of the promise that most puzzle games will set up from the beginning uh with like you know um whatever the whatever the game promises in the beginning um but if you break this rule once every puzzle maybe requires reflexes in the player's head like it completely transforms every other puzzle this is a lesson that i understood conceptually but i never really accepted i never really internalized it um to like but now i get it and it's because it was painful i didn't want to but i have since um so like my players assume that many puzzles required quick timing or quick reflexes because a few of them did um so their perception of it was that there were a lot of um yeah so like if let's say there's two quick timing puzzles in one section of the game they might assume that 15 of these levels require quick timing which they don't like for my perception i know what they do and don't require but the player's perception is different and the player's perception is all that's real um i'll just repeat that a couple of times because that yeah gotcha uh lesson number seven uh this is less of a lesson and more of a pick your preference so this is a an internal battle that i went through as i was designing limelight um so exhaustive design versus fun uh so fun as we've defined at the beginning is when the player gets to exhibit mastery skill or a wit and also i would actually say this doesn't just apply to puzzle games i would say this is just like a broad definition of one type of fun um exhaustive design as i would describe it is like there's one of every possible thing it's about the system it's like we've explored all the possible avenues of everything that we can do sometimes they're incompatible uh and i struggled with this for a year this was me back in 2016. i have since grew my hair back i could watch this gift all day i love this so much all right so breeding the witness so breed was a big inspiration in limelight uh and um it was very much like in the john blue design school of thought uh and those games are like very firmly about exhaustive design and the witness is um especially about exhaustive design to the extreme uh to frequently to the sacrifice of what would be considered classically fun because it's about the system we're exploring all the possible things about the system it's much more of an art piece in many in many parts of the game ultimately fast forward like over a year uh i decided what i immediately wanted i i went with player fun um so this is just a bear mind that this exists um and sort of you know decide where you want where you wanna uh what your priorities are then you can experiment too i don't know but but this existed and i didn't know that i wasn't allowed like i didn't know that i was allowed to go towards one of the other um because i was just very confused but i figured out all that it's all good now uh and also uh what's fun for the designer may not actually be fun for the player and that's um frequently true yeah i made a puzzle game last year and it was ton of fun to work on it was fascinating it was like really really like some really cool design um but it wasn't fun for the player uh so i was like oh geez i guess i guess these don't always go hand in hand all right so here's an example so here are a completely different layouts they're so unique um demonstrated by the names of each of the levels completely different names tons of diversity here um that's yeah sarcasm uh for me at the time they were so different uh but it's it's a pattern this was like a series uh i was exploring a series of concepts i started with one thesis and i continued building so in this level um there's a one-way street which is what it sounds like and then there's a flipper on it you once you get on the one-way street you're gonna hit the flipper so you need to time getting the lock step down to the bottom right so the solution to do that would be kind of just do it in sync with the lockstep boom perfect great unlock step goes on its own and you solve the puzzle uh next iteration of it just to make sure that the player understands what they're doing i'm in the uh the street much larger and move the made the made the moving street much tinier um so they really have to know it and this is actually an example of um this is technically redundant but i wanted to reinforce uh that skill here's another example of it where a mirror flipped the bottom totally totally new solution um uh here's another one oh this is oh no no this one's great this one's really different i promise yeah because because i i truncate um the the line so then you have to think okay like what would the oh it's really oh super cool oh my gosh oh how smart i am uh oh no this one no no check this one out yeah see this okay so there's like two of them and then and then oh no even better no no this one's even better okay so we've got we've got two of them but the line is like oh my gosh this one the thing moves so so super fast and we keep going all right players got tired of the same layout long story short uh even though what i thought was that was a unique concept um uh and the boringness was not obvious to me at the time which is funny looking back at this i'm like it's i can laugh at it because i'm like wow these are boring but uh making it i was like i'm exhausting the design like what if i do this what if i do this then what would it mean if i changed all these other things and i'm following that natural logical progression um but uh what the result of it was not a satisfying experience uh and i was confused i'm like but i thought i did i thought i followed the rules uh it gave me the design at that point it was my choice of what i did with it um so before and after um i probably should have cut this down to three people told me to continue cutting it down and i didn't because i was like no but they're so unique um uh in the future yeah i would probably just hack it down to three um even uh yeah here's another uh mechanic that did not make it into limelight um so the lock steps some of them can move faster than you um and some of them uh are about the same speed and why not have ones that are slower than you that would be um exploring uh like you know just once faster same speed slower why not explore it it was just not fun you just ran into the back of them they were super unpleasant all right time for a stretch [Music] [Music] oh thank you you didn't know there'd be performance uh performance dancing in this time that's an excerpt from a song i wrote called eating flowers aptly titled uh i thought it was a funny title uh all right so let's uh polish this off let's bring this home with how i make puzzles let's talk about some strategies first off a disclaimer these are my methods and i like games as i mentioned with handcrafted puzzles i've made all sorts of types of games but um these these strategies apply to games like limelight uh and other you know handcrafted stuff and also i don't know everything i'm uh i'm always learning things uh so this is not like the penultimate collection of it's not the perfect suite of uh of tools but it's a lot and it worked really well for me with making limelight and took a while for me to develop this toolbox so people would ask me where did you come up with this level uh i'd be like put down the surfboard uh i uh first i put down the surfboard and then i'd say i didn't the mechanics did as like unpretentiously as i possibly could i really do mean this um the mechanics came up with the levels my strategy is to create the mechanics that's my rule i will invent them and program program them implement them uh use them together combine them and then boom we got puzzles and that's the end of my strategies just kidding we'll unpack it how do we get the mechanics you all ask in unison uh experience intuition and luck uh those would be basically like the three three things that uh yeah um and okay so i like to create mechanics that uh are diverse so they're very different um they could potentially interact in many ways um and mechanics that excite you mechanics that you wanna work on it's like this is cool like i mean we're making games so uh hopefully some at least some of it is is exciting uh for you to work on um ultimately cut the duds and don't force it um uh there were many mechanics in limelight that just either were redundant or they just weren't fun or i couldn't do enough with them like the they had very there was only like maybe eight or ten puzzles total i could do with all the mechanics and um they just atrophied and eventually fizzled out um the more you experiment the more great stuff you come up with um one of my philosophies creatively is the more darts you throw the more bulls eyes you'll hit and that applies of course as well to mechanics and of course have fun making games people uh there's a dartboard so you can remember the dart analogy easier because it's when we see pictures we remember things easier or so i've been told this is like super boring so we've got like mechanic a it's red mechanic b which is like i don't know whatever like orangish and then it's like oh it's super boring but what if we took blue whoa whoa whoa whoa whoa whoa what ready great blue and then yellow what would happen and then we get green this is actually what the experience is like uh can i'm curious raise your hands if anybody like relates to feeling this all right yeah yeah like you know like when you combine things you're like wow and you're like surprised by it that's great it's very satisfying uh so the more diverse the mechanics are the more surprising things will happen uh out of out of them um so my goal uh with these this specific type of these puzzle games like limelight is to have very few mechanics that yield many puzzles there there are other games that you know can be just a constant new stream of mechanics but i like to exhaust the possibilities first um so i'll start with just a few and do as much as i can with them and that's actually uh this is easier said than done because as i'm making it you know as i was making line like i had these ideas for um yeah like i had programmed these really cool ideas i could not wait to explore i was so excited about exploring them but i realized like i knew if i started making levels with them i was going to shoot off to the moon and i was going to kind of forget uh i would not have properly finished the logical sequence of of puzzles that preceded them without those new mechanics basically like i need to finish milking this cow before i move on to the next um i like metaphors uh and also it's a great sign if you're surprised by your mechanics like there were moments where i was just like um kind of florida's to some of the cool stuff that happened and that's that's always exciting and that's fun as a developer it's very fun as a designer solution sentences so this is what i call it other designers have talked about this before um this is not like my original idea but this is how i implement this strategy so limelight's best puzzles revolve around a single moment it's not a series of moments it is one specific moment and my goal in developing each level is to make the simplest possible level fewest components with that concept as the solution so here's an example the the solution sentence is use length to get lock steps moving clockwise the level i ended up with looks like this um it is basically the simplest possible version of this level i mean maybe i could just have two of each but um it works uh the solution to this is you let you know one one of those bounce because they can bounce on your insides now they're moving clockwise beautiful you can pick up the key and head through that door so my goal was to design a level uh that had that moment as the solution and this is the solution that i came up with the moment that the player gets the lock steps moving clockwise that's the aha moment that's like oh man okay i've solved it that's where things click and uh it really it puts the spotlight on what's important uh in that level uh yeah with these uh solution sentences elegance is inevitable this is one thing that i like you could have told me this and i'd be like okay sounds about right but like experiencing it firsthand i was um uh just so uh energized by the how easy it is uh to get elegant levels out of this uh relatively easy strategy i never really had to clench to get to get this stuff and as somebody recently told me simplicity is intrinsic to elegance and that sounded fancy so i added diversity in layout will happen naturally that's another thing that i was quite surprised to discover is that uh like these the the layouts will be completely different uh whether or not i wanted them to it just happened naturally which works out perfectly because i you know that's cool that's good uh so i like to think of each sentence it's sort of like the secret to a magic trick like if a player knows the sentence going in if they knew that for you know the last example if they knew that they had to get the lock step enemies moving clockwise um then they could do that and then this you know the puzzle would be immediately solved but they don't know that that's the step that they're trying to get to that's the secret that they're trying to figure out and that's where the aha moment comes in here are some examples that i have from mine you can take pictures i'm not gonna go through these but i had a lot of these oh i have more i just i wrote down so many of these these were it was just a blast uh to make to make a lot of these uh here's a cheat to get extra solution sentences um so here echo with the player with length block flipper yeah so there's a pattern here i'm sure that you're seeing um but each level is still a totally unique challenge which i was also pretty surprised about um it's mostly because they're scattered these these five levels are scattered across the entire game uh all five of these levels exist in limelight um if the play if these were in sequence the player would probably pick up on the pattern that they're using some enemy of sorts the length of that enemy to block a flipper but but they didn't because they're coming into a blind they don't know what the solution could be it could be anything so we get five unique quite interesting levels out of it and astonishingly a small change in the sentence resulted in a huge change in the layout uh which is another thing that just blew me away i would not have anticipated that and i think especially actually now that i think about it it's because each of these um the the the mechanics on the left are diverse enough um that it's like the butterfly effect like one small change in there um ripples out so here are the levels actually from the you know from the previous descriptions the only things that these have in common are the moving street the enemy and somewhere there is length um yeah but these are like they're completely different i would not have anticipated that these levels look so different but these are the simplest versions of the levels that i was able to come up with um yeah so how do you get those sentences i hear you spiritually ask ringing in the air well there's a couple of ways i like to force the mechanics together just play together and see what happens i've thought of it before is it's like i'll just like invite them all to a party and just like be like hey what are you guys talking about what oh wait have you met you know my friend this mechanic and then like the three of them were talking like whoa you guys doing cool stuff uh i love metaphors um uh this is uh probably the most of the levels i'd say that came out of limelight um would be when i almost saw something uh as i'm developing a level maybe or a playtester try something and i'm like whoa um like that could be a solution that could have been a solution um this is something that i can't make happen on its own but i can certainly prepare for it by just writing down that idea of what i almost saw uh happened that is where so many of line likes levels came out of so just you know staying open-minded and keeping your eyes open um ask what if like what if i combine these what if i do that that's just a general creative cool strategy ultimately my advice on this is don't try too hard don't clench your muscles too hard we want to let the system do the work for you it's not really supposed to be painful because great puzzles will occur naturally with uh with good mechanics so uh here's another strategy that i used this is a systematic approach where i basically pitted each of the mechanics against each other and pointed it you know like what if i combine these two things together what would be interesting this was like sort of a map that just gave me places to look at if um if i to make sure that i had like i wanted to make sure that i was milking all of the whatever like 49 odors of this cow uh and like which ones have i missed and which ones are probably fruitful uh it's wonderful is it fun uh we all love cows cows are great um i was gonna make a pun i'm gonna hold it back all right oh so story time let's get a drink of water first water is so much fun that's the that's the real theme in this talk that's the real takeaway just kidding um so show of hands who's heard of the game steven's sausage roll and okay actually these amount of people so i played the game steven sausage roll uh and i was absolutely floored i was blown away and i was confused as to how any any human mind could come up with levels this elegant and this tight and this like um just just the the the everything like felt like it was accounted for i was blown away um i can't even put into words like as you can see uh as to how impressed i was by a lot of the design in that game um and then fast forward i'm developing limelight and people uh you know i get people asking me like how did you come up with this idea for this level like how did you come up with this and i'm like well it was what and i had a huge epiphany uh i realized for handmade puzzle games nobody's brain is big enough uh this system the designer created thought the puzzles up so it's not it's not the uh like this the puzzles in steven sausage roll i am fairly confident as well as the ones in limelight and probably most puzzle games the the designers not thinking them up from start to finish they're creating the mechanics and the effects of those mechanics the weight of the creation of the puzzles falls on that and that is not on you and that's great because it's easier to design them that way so i like to think of it this way designers don't give birth to the puzzles designers deliver the puzzles so you want to be on the right here so this is you and this is your game on the left please don't show this to john yeah so uh if you find yourself you're like oh where's the epidural very poor impression of a pregnant person uh make sure that you you step you hop out of the bed um clean yourself up put on the scrubs do your job as a as a designer uh as the ob gyn designer to deliver those puzzles um all right uh so if solving the puzzle takes like let's say like 75 of a player's working memory like a really tough steven sausage roll puzzle uh my mind like the perceived difficulty like the perceived working memory required to solve that puzzle would be like like just an astronomical amount like an inhuman amount of um uh of working memory capacity to account for all of these crazy um edge cases that are somehow neatly all solved but the actual difficulty in creating each one is much easier you basically like show up sit down do your job put things together um and finesse et cetera et cetera and um yeah it's much easier than i thought it was so that's good news all right recap first let's stretch and do my favorite activity ever i haven't done a spit take yet i'm not going to next year next spit take the the gtc talk uh recap all right so working memory uh it's human ram and it frees up with experience and it sort of looks like this this metaphorical bar it's just an easy way to think about it again it's metaphorical obviously not literal um but good zone is sort of keeping the player giving them enough but not too too much stuff to hold to juggle in their brain like let's say if i ask somebody to memorize for example ten digits um or like you know six to ten digits hold those in their head while they're solving the puzzle um it would make the puzzle artificially harder because they um their work memory is occupied by by that extra task so we've got handcrafted puzzles handcrafted puzzles versus procedural puzzles uh the differences being the handcrafted ones there's a finite number of these puzzles and the solutions are pretty novel and for the player it's a freshman varied experience on the other hand on the other hand procedural puzzles uh no novel solutions and it's a very predictable experience these these are just different types of puzzles and this is the most popular um uh spectrum there are other types of characteristics of puzzles but this is the easiest to codify um noise it's the most important thing i want people to take away here so noise again it occupies players working memory but it doesn't contribute to the puzzle so if you say what the solution of the puzzle is and there's something in the level that isn't part of that sentence that would be considered noise uh if you're not mindful it will pollute your game and i've seen so much of this and i feel like this is like my public service announcement like you've got to be the janitor clean up that noise in your in your games like you can use it mindfully uh you can use it um intentionally but if you're if you don't if you don't check it it's going to be there and it's going to really be at the detriment of the quality of your game um so make it your practice to find identify it and then eliminate it and it's okay in moderation it's not all bad um my preference like i personally swayed towards less noise especially after making limelight but noise is totally cool like the witness is very heavy in noise and obfuscation that's like that's one of the main features of the game recap for my methods to create puzzles home stretch here so taking mechanics making the mechanics and then combining them together you can do it systematically like in that spreadsheet or organically like when you see um you know something about to happen um or et cetera good mechanics will lead to great puzzles like garbage and garbage out that so that whole sort of thing um and if you're finding that you're not getting great puzzles just keep making new mechanics like in practice this is what we're doing you know i gotta get better at it uh and the system bursts the puzzles the designer delivers them so remember make sure you got your scrubs on that's just picturing myself wearing scrubs uh sitting down to work uh that's too much all right uh and have fun because what are we making here we're making games we're getting supposed to be fun for the most part uh so yeah you're gonna enjoy yourself uh uh thinking fast and slow was a big inspiration um for uh for this talk i got a lot of concepts from it um and also another inspiration every puzzle game i've ever played um uh yeah and that's um that's my talk thank you [Music] you can find these slides online at the link provided below so i think we've got time for a couple of questions about six minutes all right so i got two microphones it hit me anybody anybody wondering things until somebody asks i'm just going to do my favorite activity yes questions hi brad i loved your talk thanks heather you're fabulous um how do you decide when you're looking at like all the combinations of mechanics which ones are going to be filler and which ones are going to be the ones you keep what do you mean by filler like levels that you're going to throw out because there's too many of them um it's only after i've created enough of them the beginning process of making limelight um i didn't know when i didn't have enough content really to have an educated decision as to what was good enough as i made limelight i i think this might answer your question let me know if it does as i made lion light i was constantly like ripping out the bottom 30 percent like the worst 30 of levels uh once i had a level that was better than like you know the top 60 percent whatever it would go in there and then another one would get cut this was it was a very organic way i didn't i didn't do this like it was not a one for one trade but i um i would go through and i'm like yeah this one's not good enough anymore the puzzles that i have now are better is it does that sort of answer did you base it on like if is the player response or your own opinions of whether or not it's fun um there are a ton of reasons and it's totally subjective depending on the experience that you want to get across it totally depended through context there was redundancy there was well yeah i'd say for the most part it was redundancy and the fact i think most of the levels that i removed didn't revolve around one moment they um were like the player would kind of chip away the solution uh and it just wasn't satisfying it wasn't a satisfying experience and the most satisfying ones revolve around that single moments the ones i ended up cutting um most of them shared that characteristic thank you great answer yeah great question oh all right yeah let's go so you you mentioned the exhaustive versus fun approach to mechanic exploration and so at what point did you see that your mechanics were just kind of bloating or like when was the cutoff point because you know there's always room for more mechanics there's always room for more fun quote-unquote but like uh when did i stop adding mechanics yeah like when was that trigger point hit there was a yeah that's a great question uh there was a moment where i was like i this is it this is the final mechanic i need to finish the game uh it was entirely about like if i had another mechanic i am like the end is not going to be in sight i said okay uh it's painful because i have so many ideas that i want to add there were i had a i did end up prototyping a couple but mindfully knowing that i was not allowed to put them into the game um which was maybe kind of playing with dangerous waters because what if they were really cool and some of them were uh but i it was entirely about i need to finish this game um and i only have i'm gonna set myself a specific deadline as to when to finish it and if i add any more mechanics i will not be able to hit that deadline thank you i'm curious i'm curious at what point in the design process you decided to go with three or four platforms instead of just one um because one of the big issues we're having in my business is that we designed these puzzles in mini games and we gave it to the designers and said okay ignore platform come up with how you want to play it and this is kind of the framework and one came back with exclusively mobile style controls and one came back with exclusively pc based controls and we couldn't put them back together was this um so it was two different teams working on the same types of puzzles there was two exactly we said here's the framework ignore what console we're going to put on what do you think will be the most fun solution for this or control scheme so i'm just curious where how you decided that we're going to go for three platforms where you need to come up with kind of similar mechan similar controls yeah i uh so for in my experience uh it was originally intended i made limelight for uh for console and for steam and i released it in january 2017 for console and steam and it's controller recommended because it's much more fun to play with the joystick uh uh enough people kept telling me like this should be a mobile game it's not a mobile game and i'm like i'm like the whatever designer in me is like no it's like it's not a mobile game eventually i listen to people and i'm like okay fine i'll make it a mobile game and then i added the i added um uh very intelligent virtual joystick controls um to uh to translate the experience to mobile and i also re i reworked the entire game it took me about six months reforming it all the levels um and the feel and the timing and everything to make it feel like a natural fit for mobile um does that answer your question directly oh it does thank you and so i know this was more about the um the not procedural side of puzzles but i was wondering if your approach you said you made both types i'm wondering if your approach to designing puzzles would change if you wanted a more procedural based puzzle like if you wanted to be able to generate them like would you still go with mechanics the procedural games that i've worked on um not a comprehensive view of all the types of procedural games there could be the ones that i've worked on specifically revolve around very very few mechanics maybe like two or three mechanics um at any given time um i don't know if that's 100 true um uh can you can you rephrase that question so i understand um i suppose it's just if you were to make a procedural game now if you decided you wanted something that you could procedurally generate put rate puzzles like what would your how would you approach it as opposed to how you approach this like would you still try and create mechanics and try a few and then see if you can generate levels or would you try something else that's a good question um if i were to do it today which i have done like i mean i made a new one like a month or two ago um each one is its own completely independent idea um because it is procedural if the core mechanic of it doesn't work so like let's say one is it's a path drawing um if that if that itself isn't fun then i will scrap the idea i'll move on to something else but that's like the core of it um of the way that you interact with um with the game from there if it's working then i'll consider adding components and mechanics on top of it um but uh your question is how would i how do i consider combining mechanics together yeah it's just sort of um if you wanted something that you could procedurally generate puzzles like would you take the same approach of like having very few mechanics and just seeing if you could procedurally generate interesting levels out of them or would you try and make more mechanics or try something completely different there are two main ways that i could see to do it the witness does it in a specific way which is in um each each level is dedicated to one specific concept the other way is much more um mobile market friendly or it's like the most of the casual games on the market are which is it doesn't matter like there's very few mechanics and it's extremely repetitive and these games are pretty successful um so i've been i mean it's interesting because i've been learning about uh it's like to make uh either either one of these um uh i don't have a definitive answer um because i'm still experimenting with it but uh i'd say as advice just try it out like what i do is i'm i just i'll you know try out tons of random different ideas like little things and if the prototype works out then i'll continue unpacking it i have not made a game that is both procedural and is heavy on mechanics cool thanks very much all right all right uh that's that's it if anybody wants to talk to me after uh i'll be around yeah okay thank you

Channel: GDC

Views: 39,186

Keywords: gdc, talk, panel, game, games, gaming, development, hd, design

Id: B36_OL1ZXVM

Length: 60min 48sec (3648 seconds)

Published: Wed Jan 12 2022