Dynamic Stories for Dynamic Games: Six Ways to Give Each Player a Unique Narrative

I think in game writing circles my official title is the other Richard so today I'm Todd - are we good on the knife there is this one this one's good and this one's off all right so thanks everybody for coming to dynamic stories for dynamic games I always start off in saying a little bit of a spoiler warning people always complain when I maybe talk about endings of games during talks I don't do it too badly in this but if you see a game you're really intent on not knowing anything about I recommend you cover your duck your head and cover your ears quickly I don't spoil anything that's come out in the last couple of years also I have a lot of references to some other articles in this I will be putting the slides up on my website by this time tomorrow let's let's say that's my deadline so you can go grab them off of here if you want to get some of those references and some other links that I had to cut out for time so it's funny when I come to GDC and I swear it's like this alternate reality where

everyone is talking about narratives and ranching and player choices and empowerment and innovative stuff and to do with stories and games but I found the world outside of GDC is a little bit different you go back to your development team and like we should do this cool thing I saw in there like why would we do that dynamic stories that doesn't matter it's crazy get back in your hole writer so so why you know so I get this question a lot why do you care about dynamic stories why I do that why do you want to have multiple endings why do you want to change the story based on what the player does you know and for me literally it's why I wanted to get into working on games and making games is I love the idea of setting up situations where the player comes into a situation and has to make difficult choices and then sees the ramifications of those so for me it's so innate that I want to tell these stories so it's tricky to explain it believe me I've heard all the reasons not to do this from all the different disciplines so you have reasons not to from programmers programs always talked about how incredibly hard it can be to do interactive storytelling and they're right it can be incredibly hard if you take it to that sort of

extreme but programmers you just need to tell them look you guys are really smart you're able to you know simulate incredibly complex physics you're able to render things that look like the real world and you had to make concessions and trade-offs to do that and we can do this same thing with with any sort of dynamic storytelling you know I also here you know from producers and usually the answer there is not surprisingly it'll cost too much money and that's what but that's what producers say about everything you ask them for right it's it's it's just like when the artists go and ask for their fancy shader set or the combat designers want to do some crazy motion capture suit you know they're shot down by the producers too but they go back and they fight and they say no we need this to push the game for and this is what's going to make it great we need to use to get we need we need to get used to fighting you know to have dynamic elements in our stories if we really want them just like the other departments do but I also hear reasons not to from writers sometimes and probably some of the people in this room right now are saying well I came to this talk because I know I hate dynamic stories and we'll see if he can change my mind and and I think there's a lot of good reasons for that right that the history of storytelling you know outside of games is certainly very linear and is

used to certain structures that have worked well and things we've learned over many years of telling stories to which I say you're right those things are true those things do work and they have worked in some games but I think this is a new medium we need to try new things I know it's a little scary but I'm hoping it'll be better in the end and finally from from designers the answer I always get there is but players only play the game once so who cares I want to you know as an answer to that look at the types of games that are the most popular right now out in the world you know if you set aside sort of multiplayer games as sort of a separate entity we can look at a lot of the other successful games are ones with these big gigantic open worlds of one kind with tons of things you can do them tons of dynamic things you can do with the gameplay often still with sort of a linear story in the middle but they end up having a lot of dynamic storytelling elements around that and these games typically have you know tons of content to support all of this if we look on the other side of the spectrum though we'll see four small or sinker player games players have been gravitating towards you know a different type of style things that are procedurally generated

things like rogue likes or rogue like likes or rogue lights or whatever we're calling them this week these present games that provide you know an infinitely infinitely and interestingly most of them issue traditional storytelling opting Justino set the tone in the setting and then go from there with sunless sea maybe being a notable exception so why do players like these games despite their lack of narrative well you know players we have to have a limited amount of money right they don't they want to buy a game they can really get value out of that they can play over and over again that they can feel that is really challenged and they feel that is is reacting to what they're doing you know another reason that has recently arrived is the advent of streaming as a thing that is everywhere and you know everyone watches games being streamed now and if you want to make you know a narrative game but it's the same every time you play it it's not gonna be that interesting stream frankly you know and both of the consoles now have really easy clip sharing where you can share game pay clips with your friends and there's nothing more depressing than sharing a

clip with someone else and way to look at his clips and see his clips of the game is exactly the same as your clips of the game and to realize hey wait a minute maybe these games aren't as dynamics I thought they were so now we can't get away with that and I think we need to work at you know making games that are dynamic that people react to well in a variety of ways including streaming and sharing and I think you know as long as we're doing that we should also focus at how can we make the stories in those things dynamic as well so I you know I still wonder why these games are focusing on having such static stories in the middle of it but not all of it is and I think there's a lot of techniques that games are doing already that we can use that can really open things up and tell dynamic story so I'm going to go through a number of techniques that games are already doing today and talk about you know how those have been tells those have been done before and how you might apply them to your game the first thing I want to talk about is something I call the explorable story space so I'm going to go through this technique fairly quickly because it's pretty basic and it's used in a lot of games successfully already so we look

at something like fallout or the Bethesda games in general these are often you know these giant RPGs in these giant worlds that have that core narrative then they have a lot of side narrative around them little stories that you can do or not do and we can look at this in smaller games as well like gone home you know despite being a much smaller scale game still had a lot of pieces on the side of it that you might find or not find depending on how you looked how long you looked at the game and how thorough you were trying to be the important thing for this sort of side content that is be is that it has to be completely optional and that allows different players to find different pieces to come up with their own version of the story based on the pieces they find and players who want to put more time into it will find more of it and the other ones we just want to get through it quickly I can do that and then they when they compare stories later they'll be talking about a different experience they both had and this will increase their engagement in the in the work now when I say explorable story space I don't necessarily mean physically exploring the space obviously her story is a recent example where you were literally sitting glued in a chair in

front of a computer terminal the whole time in the fiction of the game but you were exploring the story through all these clips you were finding on the computer another one more example I want to bring up of this before we move on it's the excellent adventure game the last Express from a number of years ago amazingly well written and structured where you're sort of moving through this very last the last Orient Express right before World War one and seeing all these different stories and different parts of the train I don't think you can actually get the entire story through one playthrough because of all the layers of it that you can see depending where you are at what point in the game so the next technique I want to talk about is something I call the open-ended story and this is it's hard to call this a game technique because really this works in stories you might be writing you know in any medium so if we look at say novels we can look at novels like slaughterhouse-five or the Martian Chronicles tell stories that leave a lot out or that they're fairly convoluted and involve you you know trying to interpret what's going on in them trying to figure out what's going on certainly there are a lot of

movies like this as well if we look at something like The Shining or Inception a lot is left up to personal interpretation in these movies about what exactly is happening these tend to these these type of works that are still linear encourage repeated viewings just to try to figure out what's going on and fans come up with their own theories and debate them with each other and in so doing make these stories interactive in a way indeed Brian Upton in his excellent book the aesthetic of play talks about this where he says as we leap from beat to beat we choose a set of interpretive constraints to explain each beats presence these interpretive constraints generate expectations about the direction and significance of the story future beats either confirm or falsify these expectations triggering either satisfaction or frustration it's really fascinating book to check out and he talks about how just the act of reading and interpreting is interactive in its own way so if we extend this to game and we again so we look at something like you know dear Esther which is on some level completely non interactive in terms of the story because it's really just the

the voice-over narration you're sharing juxtaposed with the imagery you're seeing there's really nothing you can change about it but it's so sort of minimalist in its structure that you really have to you know you you can't help but interpret what you're seeing and hearing as you're playing and coming up with your own theory to figure out what exactly is going on and being said here another example of a way to do that is to make your characters you know really abstract and and cartoony would be one word even where you know maybe they don't necessarily say words that you even understand but that you get just a sense of them from these sort of broad strokes and you end up having to bring a lot of yourself to the experience to fill in those details yourself and get invested in the work that way or you know you can take another approach in Silent Hill they have a very specific very detailed story but they just show you only these tiny slices of it and you're just left with these big holes that you have to fill in one way or the other so a final thought about open-ended stories you know a word of warning I guess I've seen a lot of games have really great open ed stories that you're

interpreting as you're going and then at the end they feel you know at the end of the game they feel the need to resolve it all and explain exactly what happened and that can be very frustrating I think some of the best open-ended stories and other mediums like say the end of the shining or the end of Inception they kind of open it up again right they bring in some sort of resolution to help you understand what the heck just happened in the movie but then they have something at the end that twists it a bit and it's like well actually this and then you're left sitting there going oh wait maybe my theory is still valid the worst thing you can do I think is get to the end of one of these very open it experience and then shut it all down so the player feels like you know all of their theories are thrown out the window so the next thing I want to talk about I refer to as reacting to player actions you know this is one of the most common things that people talk about when they talk about interactive storytelling and branching so obviously I'm gonna go through this quickly as well you know we're all familiar with Bioshock that had the choice of the little sisters whether you harvest it or rescue them basically the same choice that you're making over and over again throughout the game and then Mass Effect handled

this differently with their you know blue for paragon read for renegade of choices but the game is balanced in such a way that you're really encouraged to pick one and just keep making that choice for the entire game sort of having that same shortcoming that Bioshock had some interesting other examples of this where it worked a little better if you look at the the indie game the novelist you have much fewer choices than you have in Bioshock or Mass Effect but they're not coupled to each other really you're able to make one choice here and make one choice here and it doesn't push you to any one of them as the right ending though or the right choice to make and it can have all these sort of interesting ramifications in the story and there is no no right answer The Walking Dead game certainly that first season had a lot of interesting choice points throughout it and never labeled which one was good or which one was bad or anything like that it had so many of them throughout the game that they started feeling like a real system and you never knew which one was going to be the important one that would be remembered and they were really good at setting up complex situations that didn't have a clear right or wrong

answer interestingly Firewatch is an interesting evolution of this right where Firewatch builds on that same idea of tons of little dialogue choices affecting things and it's interesting to see online a few complaints that are well all the choices you make don't actually affect the plot of that game very much but to me the heart and soul of that game is really the relationship you have with Delila the other radio tower operator and that does changed substantially based on how you talk to her and that's sort of what's most important about the game so it it is very impactful what happens even if the plot is is relatively similar of course there are other games that approach branching not through these just narrative choices but actually through the choices you make in game play like this is the classic way to do it in immersive sim like Dishonored sort of situation where the the way you go about playing the levels effect not only gameplay ramifications like those rats that swarm you but also narrative changes and this is sort of aggregating all the choices you're making throughout the game not dissimilar to papers please which does the same thing where all the choices you're making or during gameplay and add up to the different endings you can get there's been a lot of exploration of branching it's very

interesting in the interactive fiction space if you want to check that out at recommend googling patterns and choice games just a great sort of blog post by the sam Cabo ash well I hope I pronounced that right where he talks about all the different structures that have been done in different games and it can be it was a really fascinating quick read on the subject so next we have shifting story elements and this is something that's really not done as much in games and is really fascinating to me as a result so this is um we can think of this if you think of this in terms of say adapting a murder-mystery like one can think of the act of reading a mystery novel like one of these books it's really a puzzle of a sort right you're trying to figure out the more you know who did the murder before you get to the end you're trying to find a resolution to that puzzle yourself before the you know a private investigator or whatever gets to it by the end the second time you read a book like this it's much less interesting because you already know that you know who is who is responsible for doing the

killing you know even for books like this that are expertly well written it's definitely a different experience when you read it the second time so when they adapted this into board game form originally they made board game Clue which I'm sure you're all familiar with and here they randomized who was responsible for the murder and who did the murdering and voila you now have a replayable murder mystery interestingly when they turned this into a movie they said hey well so we've got this clue board game it's got random you know random people who are guilty what if we did something similar with this movie where we had different endings that would give you different people who were who were guilty of the crime of course this kind of doesn't really work very well because because they had to make the endings work with any of the possible you know they had only one version of the main film and then these different endings sort of tacked on the main version of the movie had to support any ending that happened to come up so now you had no way to actually figure out who did the murder because all of these different endings were possible so of course when you adapt this into game form you can do better than that so the example I like and there were a number of these in the 80s that did this

actually that had these murder mystery games where you were trying to solve a mystery but who actually had done the killing and and who was killed was randomized each time you play and because it was a game they were able to plant little clues all along the way that you were playing and you would be able to solve it unlike the clue movie another example I like to bring up all the time is the Blade Runner game where each character in the game could be a replicant or not depending on whether depending on basically a dire role to be any of the game and then this would affect what you had to do with them because you were you know hunting for these people and had to make choices about about what happened to them so some of the really interesting work in this sort of randomized story space is actually done in board games you know going back to the classic Arabian Nights game or more modern games that sort of randomize the narrative in different in different ways obviously these have the benefit of having multiple players in them but they still do good things to change up the story every time so it feels like a very dynamic experience and different every time you play an interesting example you might want to check out is this game the

black closet which is actually nominated for an IG F award this year and I believe you can go play it on the show floor this is sort of an interesting mix match of several different game types sort of like a dating sim mixed with sort of a adventure game and here you're you're in this all-female boarding school and there are these things like who's going to betray you and each game is randomized each time you play along with the other stuff making by sort of combining it with a sim in these random narrative elements you're really getting to a very different experience each time you play so my own game the church in the dark mint which is now in development is set inside I really just caught in the 1970s but the player doesn't know whether that cult group is is is merely happy socialists who want to go live in the jungle or something more dark and apocalyptic it's going to happen there and the key to the key to this is the leaders personality is randomized each time you play the game so you don't know what they're what know what exactly you're dealing with and you're sort of acting as a detective going into the int into their sort of jungle compound to figure out what's

going on so beyond just the the interesting factor from a replay factor now I thought it was also interesting because I felt it fit them Attica Lee with the hard to understand nature of cults where they can be really hard to tell if it's actually a good group or or something more sinister from the outside so it's my hope that that including this shifting story element you're able to see both sides of it can be a very complex story and I like specifically games that combine this sort of you know shifting story elements with player choices and like the unfortunate thing about player choices when you make them in a game if you ever want to go back and replay the game you know naturally you're probably gonna want to make the same choices and often when you go back if you want to see all the content you'll force yourself to just make the choices you don't want to make which is not really a great experience however if the context of the game is shifting you know so that it's different this time you'll want to make different choices the second time you're playing because the context is different because the reasons you're making these choices are different and that makes games like these hopefully replayable in a more interesting way all right so the next

thing I want to talk about a bit is character simulation so this is one of the harder methods to implement but also represents sort of one of the less charted areas so I think a great area for future experimentation with character simulation techniques each character in your story has a simulated personality with their own wants and needs and desires friendship of legion sees and these dynamically change based on your behavior as you're moving through this story sort of based on what you do this sounds super complicated and it can be but it's simpler forms are possible that still produce some interesting results so probably the most commercially successful example of this is the Sims right in the sims you're able to manipulate the personality as these people and play them off each other in these interesting ways the possibility space of the Sims is intentionally I think quite broad it lets anyone live out whatever version of American suburban life they want whether that's getting married and whatever fashion they seem they find suitable or having a goth family or even torturing their Sims this all feels like it fits because the game is so cartoony because

it's so broad any goofy thing that happens feels like it fits okay into the game another game that people don't often think of as a storytelling experience but I think it's doing this is civilization and the different sort of leaders of the different countries that you meet as you're playing and they'll come to you you know trying to negotiate peace or talking about threatening war or setting up a trade deal or whatever it is and they'll remember your past interactions with them and they'll remember whether you you know betrayed them before who else you're having a fight with and then try and negotiate and this was taken even farther in the sort of cousin of civilization Alpha Centauri which treated the heads of state even more like believable characters and they each had their own sort of political agenda and their own philosophical beliefs that affected whether they were weren't going to want to side with you and they sort of gave you feedback based on how you're playing the game whether they agreed with it or not if you consider these types of games really story games if there's no reason this sort of faction based reaction system of these games couldn't work too you know emphasizing even more and make feel even more like a narrative there's

a couple of games that used character simulation that we're definitely going for something more narrative when the first of those is versu this was an interactive fiction system that used a much stronger story framework and much more limited player choices than you would then you would find in the sims or something like that so this was a much more confined social social situation the player was able to engage with all the characters in a number of different ways and these choices would persist and then affect their relationships with each other and the first setting for the pursue stories was this sort of Jane Austen ish social settings where character and dynamics are everything prom week is another interesting example of this sort of social simulation in a very carefully chosen setting the week before high school prom a time rich with narrative possibilities and and pretend potential conflicts it takes a Sims like approach to controlling these characters where your only goal is not you're not just playing dollhouse like you're playing in the sims you're actually trying to get people to pair up with each other and following some pretty you

know more more confined goals the game uses a pretty elaborate personality model that replicates the social craft you might find in that hellish time of high school but to work both you know it's interesting to look that versu and prom we you know both had really elaborate AI systems so in addition to requiring you know a kick-ass AI team to implement something like this they also made very specific careful choices about the setting they were in that they would have a tone that would allow this kind of gameplay to work so you want something that allows for these interesting narrative simulations in the case of sort of the Jane Austen heir story or a prom type of situation and that you they both sort of chose a tome that was very deliberately silly even slapstick farci where where the player might try to do all these goofy things but they would still feel like they would fit in with the game you know there's many potential settings I think where if you tried to do character simulation would be much harder to do particularly when I think that the technique is in its infancy as it is now and the other thing these games were

really good at was telegraphing what was going on to the player and I think that's super important when you're trying to do something as complex than this you know for Sue used our rather prom week use the a menu system that was very much like The Sims that really gave you all sorts of bars and graphs and things about how everybody felt with everyone else and even with all of that it was still was very hard to try to understand what was going on the game if you didn't understand the simulation decently versu tried a more integrated approach where they had those little images at the bottom of the of the page and they would actually change facial expression based on what their social state was at the time and if you wanted more information you could tap on them in a little box would pop up it would explain to you why they felt the way they did and I know that over telegraphing something like this can feel sort of aesthetically jarring particularly as a writer you know I feel that way myself but I think a technique like this is in such its infancy that we really have to overdo it to give players you know a chance to catch up it's like people didn't learn how to watch movies you know they learn over the years and film to get more and more subtle on their teeth techniques

and I think we have to see games the same way when trying something this ambitious and different an indeed Emily short the co-creator of versu had this great quote she said my general rule of thumb is it's pointless to stimulate anything in AI that you don't also have a plan for communicating to the player otherwise they don't see how clever you're being or worse what it's actually sophisticated behavior winds up looking like a button and if you want to hear more about versu or prom week there's this excellent talk on the GDC vault it's up there for free you can just search for beyond Eliza and it pops up and they talk about both of those projects and how they worked in there now the final example I want to talk about is drama management you know I like to think of drama management as the flipside of character simulation or as character simulation the characters can sort of drift in sort of interesting ways but that maybe don't end up feeling as much like stories here this is actually steering things back towards something that feels like a story you can think of this as as akin to a D&D session we have a really good DM who's trying to shape things into a into a

certain narrative form he allows a lot of agency for the players but he is constantly steering things back to the story he wants to tell while still letting the players feel empowered so the most interesting example of this in digital form or the most well-known example is is facade in this game the player enters into the apartment of a couple that is having an argument before as of a dinner party and you're sort of there while all their marriage is falling apart the players here has a tremendous out of agency you can say whatever you want to the characters and they'll try to react to it and the game really does a pretty good job of steering it back to the main story it wants to tell so facade was always you know it was sort of interesting as an academic project as a PhD project like it was when it was released but then it found this interesting second life on the internet because it became very popular as a let's play thing and people loved streaming this game because it was so reactive to whatever they would try and they would make all these let's play videos of like oh let me try to get it to do this and it fad and this is sort of going back to what I

was saying before about streaming being a thing that's popular where people who are streamers really want to find games it will do crazy interesting things each time you play them the other example I wanted to bring up of something that I consider to be drama management is the Nemesis system in shadow of mordor you know this is obviously a much more commercial application of it but it takes all the interesting sort of simulation elements of an open-world game with combat in it and shapes them into these little player stories that they give you a lot of motivation to keep playing the game and indeed Michael platter who's sitting right over there has said that the Nemesis system is emergent procedural narrative that's neither linear nor branching which is an interesting way to think of what trauma management is really trying to do so just to give a quick summary of how this work the game turns the common experience of dying in the game turns it into a little piece of fiction basically where the player dies but the world keeps going without him and the player reincarnates and the enemy who killed him is sort of shown how he's moving up in the ranks because

he did this great thing of killing the player and indeed he might get a promotion or something like that and and you now are he's marked as a nemesis for you to go kill later in the game and and it's very good at creating these little vignettes when you find come and find these characters I'll show you a little cutscene of this guy and it has they have all these lines of dialogue that they'll say to sort of taunt you at this time they all have custom names and all of these little elements that went into the system that made it work well and it's interesting that it's just taking the completely the thing that you do in every open-world game and just giving it all this context that makes it extra meaningful sort of creating drama out of the simulation so that when you finally can find this guy and kill him it's a very satisfying moment all right so those are the six techniques I wanted to go through today you know I think it could be interesting to look at your own projects and try to see which of these six things you're doing already I'm not saying that every game should try to do all of these things though that might be interesting if someone to pull that off but you know if maybe you're doing one or two of them maybe you can add a third of some kind really depends on the needs

of your game but I think if you find that you're doing zero of these things I would argue maybe you're not making a great game story and I think you know at the beginning of the talk I talked about how it was important to look at this because of streaming and sharing and all that stuff and I think the reality is that streaming has revealed a dirty little secret that we've had in making single-player games for a while and that's that we made the play the experiences feel dynamic to the player but they weren't actually right that everybody felt they were having an eek experience but really it was just exactly what we wanted to happen and I think we can't we can't hide behind that anymore and I think even if a player never actually you know never actually streams a game or never actually shares it or sees it in any other form it's still important just for player investment it's still important for players to feel like they're involved in the story if it's dynamic in some way and that they can they can affect it I think you have to recognize that there been many games that are very successful using this technique whether it's you know finding that last safe that has the final clue to the side story in the part of the game you really

want to figure out and feeling that you found that or you know the abstract experience that really lets you invest yourself in it and feel that you know this becomes your story through that or even just hunting down that guy and finally killing him after he's been your nemesis for a while or that scene at the end where I remember all that time we had together and all the choices we made along the way I think if insured you know I think when dynamic storytelling really works you know players really love it really become invested in it and and feel that they've told their own story so thanks [Applause]