Dynamic Storytelling in The Novelist

[Music] can you guys hear me oh there it just turned on cool alright so I'm gonna go get started thanks for coming to my talk called woo called clicking too many slides ahead it's called the system is the message dynamic narrative in the novelist the slides are going to be available very soon at that URL down at the bottom and that URL is also on the last slide so don't worry about taking pictures or anything please silence your cell phones and also make sure to fill out your evaluation forms so a quick bit about me I started my career at iron storm Austin where I worked on a handful of games before going to Midway Austin where I worked on a game that was never announced and got cancelled while Midway Austin went down after that I went to 2k Marin where I stripped 2 Bioshock 2 and after that I went to Lucas arts where I worked on a game they never got announced and eventually left that game to go indie and I almost forgot my sign

to myself that reminds me to slow down so I'm going to put this up here because I really sometimes talk too fast so there we go now we're set so that's a little bit about me and three years ago I gave a talk here at GDC entitled player driven stories how do we get there and I highlighted a problem where games aren't taking advantage of the interactivity of our medium and I was kind of pushing back against the more and more linear stories more and more Hollywood copying type stuff and really pushing for the player to have more control over stories and games and the goal I set forth is a story in which the player makes influential decisions create that create a unique narrative that's not prescribed by a designer and by influential I mean decisions that actually matter not just cosmetic things or not things that close a quick loop after that but really have a lasting impact on the narrative and the second word I want to highlight is actually that word narrative this is a cohesive series of events that form a long-term satisfying story not just isolated micro level events and if you want you can

actually watch that talk it's free on the GDC vault so if you're interested check it out but when I gave this talk is pretty much theoretical it wasn't related to any specific game I was working on it was just more like a call to action I wanted to see happen and even after I went indie I didn't immediately start addressing this problem or thinking about a game that dealt with it I actually started working on a puzzle game that I pretty pretty quickly gave up because I realized I didn't care at all about it and so I decided I would take a crack at trying to do something with actual player driven stories so I moved to a new idea that just had a working title called ghost and it's what eventually became the novelist but it started off as a very different thing it was six people in an Oceanside mansion after one of their friends had died sort of like in The Big Chill and you did play a ghost in the house and you could creep around and you could learn about them you could manipulate their relationships you could create and destroy romances make people friends uncover past secrets so and so and the idea was that if I give the player lots of control over these six characters then a story would just magically emerge

and I sort of built this toy like simulation so I built a prototype in unity and got it up and running and there was six characters in this house I named him after characters in one of my favorite movies which is Halloween so we had Annie we had Bob we had Linda we had Annie's boyfriend Paul who never appears on screen so here's where Bob gets killed and I also had the two characters from one of the great videogame love stories Rebecca and Tommy from System Shock 2 so those are my six characters and I built it and it worked and you could you know make two characters fall in love you could create in the story romances you can make people friends and stuff but it didn't really have any impact there was nothing to it that really had any emotional resonance there was no emotional hook there were too many characters and you could manipulate them in too many ways and so numbers moved around and you could check a status screen that said all these people are now friends or these people are starting to dislike each other or whatever but there was no on-screen meaning there was

no story emerging from this and I realized that without enough backstory without enough context that this was basically just a group of systems from which no kind of actual meaningful narrative would emerge or as I called it in my 2011 talk system soup which is basically the idea that you have a lot of stuff going on and bunch of stuffs happening but it has no real form and it has no real meaning so next I decided to drastically reduce the number of characters in the down to three and provide a really clear context so that players would have a framework for understanding their decisions so three characters is kind of the lowest you can go and still have interesting push-pull kind of dynamics and I picked a family because that's something that's universally known to everybody you know what it's like to be a kid you know what it's like to have a romantic relationship a lot of people know what it's like to be a parent and so even if you haven't done you know a couple of those things you still understand what it is so I didn't have to spend a lot of time providing with context and backstory you understand immediately what would make a good or bad relationship will make a good or bad

dad so I kind of tried to do that to address the shortcomings of that first prototype and to explain a little bit more about the game at a high level it's about a family name the Kaplan's and dan is a writer his book is not going very well Linda is his wife who's a painter and their marriage is kind of on the rocks and their son Tommy is a kid who just went through his first year of school and he had a really tough time he got picked on by some bullies he struggled a little bit with his reading so they've kind of come to this summer home to get away and try to figure things out and you play as a ghost and you explore the house and you read their thoughts and read their memories and discover what they want and you uncover these dilemmas and you actually start to help them decide how to resolve them so I'm going to quickly play the trailer for the game just for people who haven't played it just to it gives a pretty good idea of actually what you can do in the game it looks like the play it so I'll just start that now [Music]

[Music] [Music] [Music]

[Music] so that's how the game works and I am really really sick of that song so the I wanted to talk about a couple of my goals for the game the the high level goal was to try an experimental form of storytelling where the relationship where I take the relationship simulation of something like a sim style game but manipulate it to present it as a recognizable traditional story so you're not necessarily managing meters between characters and looking a lot of

on-screen HUD but you're taking those same types of dynamics and trying to turn it into a story so the first step there is the abdication of authorship I wanted to advocate authorship and put the play the story in the players hands and this is really important because this story is about trying to balance your career with your family and that's a really big question what's what are the right answers there and I'd be lying if I said I knew what was important so my goal wasn't for the game to be a platform for my own beliefs instead my goal was to make it a question game not a message game the player doesn't give or the game doesn't give the player a specific goal or judge whether the decisions are right or wrong it just lets the player make choices and try to give them reasonable consequences and kind of feedback the effects of their decisions so what I really wanted to do was create a dynamic narrative not necessarily a branching plot I wanted to create a narrative that was based solely on the players input as opposed to a branching plot where every player has the same series of events and choices so

I think it'd be useful to quickly establish what I mean by a branching plot a branching a branching plot is created by a writer or designer and though the player does make some decisions the events come in the same order every time so the plot starts moving along and at some point the player reaches a decision point and they're given multiple options maybe they have to choose between different missions or align with one of two fashions or even decide if a character lives or dies and this technique is used very frequently in games like Mass Effect or the Bioshock series very familiar so the player decides what they want to do and they're in two branches of the story and these are both authored content that are based on that choice and the player might be given another major decision along one of these branches so they decide again and the first choice might fold back into the branch coming from the new choice but eventually things probably fold back towards the together towards the end and you're presented with one final decision that once you choose there it splits the eye splits out two different endings for different players

and this is frequently done on a sort of good neutral or bad ending that's that's not the case all the time but it's a pretty pretty common trope there and before going on I want to be clear that this is a perfectly valid structure for a game plot I'm not my talk is not a screed against this kind of plot I wanted to do something differently but these these plots can always obviously worked very well in games and give player really memorable experiences where they make choices that are really powerful to them so I just want to make sure that this doesn't come off as me trashing this kind of branching narrative because I think it's fine but in my game I did want to impose this little pre offered structure as possible from from my angle so instead of writing a story and even a story with a bunch of choices I said wanted to create narrative systems from which a player driven story would emerge so actually a systems based narrative so I made a bunch of structural decisions to try to make this happen and the first and maybe most important one is that the chapters in the game come in a random order on each playthrough so I actually sort of

burned the boats and literally prevented myself from being able to author a traditional branching plot like I could not if I tried because when this chapter comes up I don't know if this has happened here or happened here or what the player did here or what the player did here so I actively presented prevented myself from being able to create that kind of narrative I simply don't know what comes before or after and so that provided a lot of constraints on the actual chapter design but each chapter became sort of a self-contained building block for the larger narrative the second point is that I wanted relationships to be the sort of core thread and the core connective tissue between chapters not a series of plot events and finally I wanted to create a framework that's based on player values as the authorship of the game I wanted to create something where the player has to make purse decisions to create this story so I tried to take each chapter and make each of the decisions the player has to make equally sympathetic from the different perspectives in the game I didn't want players to get to a chapter and say oh this is clearly the right answer or this

is the good versus evil answer obviously I should do this I kind of wanted them to have to be sympathetic to all characters I have to say wow what do I really think is right here what do I think is more important here so again it's a question game not a message game and more specifically it's a game about trying to balance your career and your dreams with your family so all of your decisions are based on making choices that prioritize one of these things here and the high concept is essentially a game with nine chapters each of which asks you to make a decision on these axes in a slightly different way and the hope being that you'll end up with a set of relationships at the end of the game that reflects your own choices on that subject and that the story is almost like a mirror to help you process your own values on this specific question so I want to walk through an example chapter just to make sure it's clear how the game works so that we'll have a vocabulary going forward in each chapter you spend time sneaking around the house and trying to remain hidden and so you read things the characters have written you look at things that Tommy is drawn you read the character's thoughts to see

what they're thinking about you even explore their memories and relive past conversations and past events and all these things help you piece together what's going on in each chapter so you have a discovery phase where you uncover this type of stuff and each character has their own chain of clues and so what you want to do is basically find enough clues for each character to discover how they want to resolve the dilemma in the chapter so to use an example of a chapter called the show in this chapter Linda is going to be putting on an art show in town and she's really under the gun trying to finish a last piece before the art show goes live and she's asked Dan to help her with some promo materials and also asked if she could maybe if he could help her find an agent but meanwhile Dan is a behind on his own deadline his agents riding him he's dodging phone calls and he's got a big deadline to Monday and he's super under the gun himself and doesn't have the time to do it and finally to make matters worse he had also promised their son Tommy that they were going to go to the beach that day and look for a conch shell and just have a nice Saturday at the beach so

everything has to be done that weekend and there's not enough time to do it all and the game is told from Dan's perspective so it's presented as a series of choices like these where he can't do everything he wants to do and you have to pick what's more important so after you uncover all the clues you basically have to decide whether Dan should help with the promo forms and so Linda can focus on her last painting and kind of support her in her career so you see there she's thinking you know he could take care of those promo forms for me and that it would obviously help the marriage or you can look at him just basically going into crunch mode over the weekend and chained himself to his typewriter to work on his book which of course helps his career and helps his novel or he could blow off both of those things and take Tommy on the beach trip which is good for his role as a parent so in this case let's say the player chooses to prioritize Dan and Linda's marriage and work on the promo forms these these decisions are fed back to the player through a series of short scenes that kind of tell you the impact of what happened so here we can see that Dan and Linda are working on the forms together and it kind of

explains I don't know how easy that is three but it basically goes on to say the art show was a big hit the ad in the paper did a lot to help that happen and it was good that he helped her and it was really a big boost for her and it was awesome that he was a supportive husband and that's really cool however Tommy didn't get his beach trip so he spent Saturday looking out the window being super sad and that's kind of a bummer and then Dan also misses his deadline because he spent the time supporting his wife and he gets chewed out by his agent and his book kind of goes a little bit more in the gutter so again the narratives constructed from the history of the players choices on these specific topics and so the starting point of the next chapter in this case would be a better marriage because he helped her with the promo forms and unhappy her child because he broke that promise to his kid and a worse book because he missed his deadline and these relationships are the spine of the story so that I decided that I would build a really simple simulation of them so each of those relationships is scored from zero to

nine with zero being the worst and nine being the best and at the beginning of the game they're all rated at five so everything comes in even and in the example we just looked at where Dan supported Linda at the expense of his work and his relationship to his son the mayor score goes up to seven while the crew while his career score and his relationship to Tommy each drop to four and these numbers are never revealed to the player they're actually fed back through a variety of other means that I'll go into a little bit later but it's important to know that this is all under the hood so the relationship scores like I said are the connective tissue that carry forward from chapter to chapter in the game and the next decision that the player makes will be based on starting with these new values so if you plot that change over time it basically looks like this each decision makes a relationship get notice Lee better or slightly worse and so this little pac-man mouth is sort of the atomic unit of the story every decision functions on these relationships one way or another either making it pretty pretty much better or slightly worse so the way this works over the course of

the game is that if you look at a game with nine chapters and they all start at five this is this is the sort of possibility space of any given relationship over the game these are the possible decisions that could happen so by the end of the second chapter any relationship in the game can actually already reach its maximal value because it's five plus two is seven plus two is nine so Dan's book can be the next war in peace or he and Linda could be on a perpetual honeymoon or he could also be the greatest dad in the world so if we keep going a little further you see that from each one of these points additional decisions can be made and continue to build out and by the fifth chapter any relationship that's been completely neglected can hit rock bottom it can actually get to the worst shape possible if you have ignored that you know either ignored his book or ignored his wife or ignored son the whole time so we follow forward and the decisions pile on decisions pile on decisions and this represents the full possibility space for any single relationship in the game and if you compare that to the branching plot diagram it actually looks pretty

impressive right there's so much stuff you can do but it's actually not this is kind of a mess it's too granular for most of these data points to mean anything and since the numbers are never shown to the player a lot of this detail is irrelevant so while there might be a big possibility space it's very hard to tease a lot of meaning out of this and in fact it's just a different form of the thing I was trying to avoid which is system soup so when I realized that I knew the next challenge was to basically take these numbers that were under the hood and very granular and not fed back very well and overlay a lot of sort of human readable context to let players know where they were standing and where these relationships were so here to some of the methods that I that I used to feed this stuff back I'm first off the player can basically read the characters thoughts and just see what the relationship statuses are so here's one where Tommy is bummed out about his dad they don't have a very good relationship and so he's saying Oh mommy said daddy's just too busy so he's you know things aren't going well and you can kind of

read oh wow that's kind of a that kind of hurts that maybe I should give Tommy a little love in this chapter you can also explore the characters memories and buying Diaries where they react to previous decisions so here's a diary where Linda's really excited about how things went with her art show and you kind of get that payoff of how elated she is that things went well for her Tommy also draw does drawings represent his relationship with his father so this is when things are going really good he draws a happy pictures dad's got his arm around him here's when things are kind of just okay they're just kind of saying like hey we exist and then this is when things are going really badly his dad is like this screaming monster and he looks sad he's given this what do I do things so basically shows you the different range of drawings that Tommy can make I also did a thing where their objects left around from previous chapters in the house so then this newspaper clipping on the wall only comes up if you do the art show with Linda and if you go read it it's basically a notice it says you know local artists or visiting artists Linda Kaplan put on a great show bah blah blah so it kind of gives you a little bit of payoff to remind you of your past decisions and

over the course of the game these objects build up across the house and sort of create a unique landscape of the decisions you've made represented by these physical objects another example would be if you had actually taken Tommy to the beach that conch shell would now be on his bedside table for the rest of the game another one is the one player or when characters pass each other in the house they'll greet each other so if things are going really well with Dan and Linda they'll have a really happy greeting that maybe talked about taking a break together or going into town to get something to eat if things are just going okay it's a lot more perfunctory just like hey oh hey and if things are going bad they barely acknowledge each other and it sounds really icy and and pretty tough so just ambulance you can kind of hear some representations of those relationships and if things get really bad characters will actually start locking themselves in their rooms so this is Tommy in his room alone playing with his toys sad because his dad's neglecting him and Linda can do the same in her studio and there's also some scenarios where Dan can do that in his office and then this is my favorite one in which I don't think anyone's ever noticed in the world

but I like it dance Trash Can if things are going really good on his book there's almost no crumbled up paper in his trash can but as things start to go worse it fills back in and each chapter it kind of represents the status of his book by how many bad pages are curled up and thrown on the floor and I don't know now at least some people know that that exists so these things are all good at giving the player status updates that's good it's information the player knows how these relationships are standing but we still don't have a story we still have nine flat chapters about any highs or lows and you just sort of manipulate these relationships but it's just the same sort of flow over and over again and one of the key benefits of an authored story is that the writer can make sure the story has a beginning a middle and an end they can make sure to modulate the intensity of the events and create some dramatic urban flow so that the story has some structure and players are kept more interested and you kind of have these big events and then your release with some pressure and release the pressure and then build up to another big event and basically modulate the flow so players don't get fatigued but they also feel structure and there are a lot of ways to do this but one of

the most basic is the traditional 3x structure which is a very simple way of creating rising and falling action over time I'm building to a climax and resolution now this is super super basic it's not innovative at all and a lot of writers actually advocate moving away from it and going to more advanced forms but in the case of my game given the difficulty of trying to create a narrative out of these really unpredictable systems where I don't know what order the events are going to happen in I don't know what the player is going to choose each time I actually felt that having something that's very identifiable and very recognizable and a form that people get would actually help so this this did not feel like the place for me to try to do something weird I actually thought the game benefited from having a really simple defined structure based on the three acts so the more familiarity and form I could add to the story I felt that was that was better so if we look at our map of the possibility space again and slice it to have three and a climax then we can start to see like specific points in the narrative and a timeline that we start to build up

- and if we slice it into three actually can also zero in on where those relationship scores can be at any of these points in the story so we start to identify I want to I want to punch things up here and this is sort of where things can be at this time and we can start to isolate places to put bigger and more memorable moments in the story to try to give it those peaks and valleys and I used the 3x structure as a guide to introduce more drama it known intervals and the way I did that was I made chapters three and six and eight higher stakes dilemmas they were generally more I don't know emotionally involving chapters and each one of them focused on a different character and then chapter 9 is the finale so it sort of has the highest stakes of all and since the game takes place over one summer I felt like there was a natural rhythm to it where I could just say well June is Act one July is Act two and August is act three culminating with them leaving the house and going forward into the future so so I started to do this but I realized you

know maybe so so the bottle is an example of one of these high stakes chapters it's one where Dan's drinking problems can is getting a little bit out of control his wife's worried about him his sons upset because he's not paying enough attention to him and I decided to set all these high stakes chapters at night the game has different times of day so I figured why not make it at night make you feel more ominous make it feel more heavy and so I made sure to put these at chapters 3 6 and 9 they're still randomly selected there's a bucket of these and I put a random one at each time so it's not like I really structure it that much but I just make sure if something big happens at these key points in the story and I also make sure that after that you know it pulls from the bucket of lower stakes chapters so she kind of get some peaks and then a little bit of release of that tension and in addition this was another thing that I added after a lot of play test feedback was after each one of these high stakes chapters you get an additional set of these narrative wrap-up scenes that give you sort of a roll up on the overall relationship status of all these axes that aren't related to any specific chapter so they

take completely off screen take place completely off screen and they serve as a summary of all the choices the players made up to this point so it's not just here's people reacted to the choice you just made for this chapter but here's where your relationships stand based on all of your choices so far so it wasn't feasible scope wise obviously to create scenes for every possible relationships core it might have worked okayed the first because there's not that many options but by the second act and certainly certainly by the third act there would literally be you know 10 options for both because every relationship could be everywhere and that's completely you know unsustainable from a scope especially from your one dude making a gay man's apartment so that wasn't necessarily a problem though because these would be so granular anyways that I'm not sure the player would know the difference between what's a what's a six relationship and what's a seven relationship so what I did was I broke this up into bands that basically denoted different different relationship values and then broke those down into you know eight or eight and nine or good six or seven or eight 9s great six or seven is good three four and five are

bad and zero one and two or awful so I get down into categories to kind of make them work more manageable but also because that way they would actually make sense two characters you can tell the difference between good and bad a lot more you can then you can tell the difference between four and five or something like that so it basically preserved the granular relationship scoring under the hood while compressing the dynamic range of the player feedback into something that was more meaningful to players so here's an example of some of those scenes you start off with a little intro that says hey the first month is done and you know a lot of stuff had happened and you then you get three scenes to tell you the status of these relationships based on all of the decisions up to this point so for example after the end of the first act Dan sends his work in progress to his agent and his agent comes back and says oh wow this is really great this is awesome so you know that Dan's book is going well at the end of the first act of the game dan and Linda go into town for a date and the date goes really well so you say oh man I guess they really are hitting it off I'm so glad I'm

saving this relationship as the player based on the choices I made but here's the story a sad story where Dan tricked Tommy out the woods to look for arrowheads and basically as soon as they got there he said okay come on we're going home and Tommy was crying dan didn't care and so it basically says that yeah he's up to this point being a very neglectful father and there are four different versions of each of these scenes so you depending on where those relationships are in those zones then you get one of those four scenes that's a lot more manageable than ten scenes each time and arguably it's more readable to the player and they really explicitly tell you how things are going so you get these at the end of you know you have these big high stakes chapters and then you get a roll up that says here you are in the story here's where it is you're like wow end of a month and it kind of gives you a good place to to stop like a lot of people will play an act and then put the game down and play the next act with the game down so it really did a good job of starting to split the game up until more definable structure so that there was some rising and falling action so once I got to this point the

system was actually in pretty good shape players understood how their decisions and actions were creating the story I was starting to get to a discernible narrative flow and there was a beginning a middle and an end and the players had an emotional connection to the experience and I was happy for a while but unfortunately while the core system was pretty solid it started to reveal some limitations and these started to come mostly from players sort of natural expectations regarding how human beings would actually react in some of these scenarios so let's take a look at a scenario where the player focuses on Dana and Linda's marriage with some focus on Dan's novel and puts no love towards Tommy so this graph is going to show all three relationships as they progress through a given playthrough so you see them developing series of choices +2 or down one depending on what's going on and if the player makes a series of choices to get to this point then the marriage score is at an 8 so Dan and Linda are madly in love their marriage is going great they're really reconnecting but then you look down here

and Tommy's at a 0 his dad is being terrible to him he has ignored him through literally the entire game he's neglecting him Tom Tommy is starting to get traumatized and then I realized that you know I'd put into these feedback systems as I showed earlier right that said how things are going and so the player actually really knows what these relationships are and I started to see these incongruities when when things started to reach the extremes of this dynamic range so it simply wasn't realistic to have in the same chapter Dan saying oh man I should make Linda breakfast in bed and Linda's saying I hope we didn't wake up Tommy last night so obviously their marriage is going really well but in Tommy's sitting in this room with the door closed saying why is my dad such a bully it's hard to imagine Linda being madly in love with her husband at the same time that he's being this awful awful father so once you move to the edges of this range these things can start to conflict with each other in ways that really started to jump out so since the player has no control over Tommy or Linda then they have to have some level of autonomy in the way that I deal with these

relationships and so the solution I came up with was that in this case Linda would actually speak up for Tommy and if the roles were reversed Tommy would speak up for Linda so let's roulette freewind us just a little bit and get to the same decision point here and apply these same scoring zones that I was using for the app the Act break wrap-ups so we can use these inflection points as triggers to know when the characters to speak up for each other so for example Linda sees that Tommy has just crossed into the awful zone like things are going really bad and she decides to do something about it so in this scenario instead of her wanting her desired outcome in a vacuum which is what we saw before she says wow which Dan could take care of this promo for him so I could work on my painting instead she says he shouldn't worry about those promo forms he really should take Tommy to get that she'll she speaks up for Tommy and tells the player very explicitly hey this is what I want but I don't want you to do it I want you to take care of your son so this is like a feedback mechanism in the game and it was originally just intended

to make the characters more believable but I quickly realized that I had to incorporate those intentions back into the simulation itself it wasn't good enough for them to just speak up so in a normal case obviously if none of the relationships are in the awful zone and nothing's been triggered then each character just wants what they want in a vacuum like we saw with Linda before but but here Linda has you know changed her mind and says she wants Dan to pay attention to their son not do what she would actually normally want so the player listens to Linda's request and goes and plays with Tommy Tommy gets the normal two-point boost that pulls his relationship back out of the awful band it's still bad but it's better than nothing but the key thing is that Linda doesn't take the normal 1 point hit which stands to reason she was very clear about what she wanted and she said you know my desires for you to put Tommy and front of my own needs and it would be a really bad simulation if the player got such a clear message from Linda did what Linda wanted and then she still turn around I got mad at him for not doing what she wanted so in essence the simulation allows Linda to forgive Dan

in conditional cases where she had actually said hey you know what what I want isn't the most important thing right now and on the flip side if we look at the same situation and Linda says you should take care of your son forget my promo forms and Dan says shut up I'm doing the promo forms anyways then Tommy's relationship score still takes that one point hit it still gets worse because he still got ignored but Linda also takes the hit she gets angry about it she says I told you you jerk you know pay attention to your son and you just when did this anyways how deaf are you and there are actually conditional outcomes for all the chapters that address this if you go either way I'm in this scenario then in the next chapter it's fed back and she says you know I really would have liked to have helped with that but I'm so glad that Tommy and Dan had a good time at the beach and blah blah blah so and this works both ways if Linda's relationship gets really low and Tommy's is high then he'll say I don't want to go to the beach go make up with mommy so they kind of regulate for each other because since you know the player doesn't have direct control over them and the way they do over Dan they have to have some level autonomy so I had to modify the

simulation some to account for edge cases and also to make sure that the simulation stayed believable and I was happy with how the players or how the characters that the player can't control started to behave in these scenarios and in general the system seemed really tight but there was something that still bothered me and it's already causing a lot of complaints for play testers as well and that's the fact that and this is probably the biggest one that that really took a lot of thought is the relationships system is zero-sum the player starts with 15 points in the system and every choice just reassigns the same number of points so 5 plus 5 plus 5 is 15 so does 7 plus 4 plus 4 so over the course of a game you know you start everybody's at level 5 which is at the top of the bad zone that's why they're here things aren't going too well for any of them and then and a zero-sum system the player is just spending this entire game rearranging the deck chairs on the Titanic so to improve one relationship you have to make the other two worse and so as you

go through the game you're just moving around the same 15 relationship points making one score better and the other score is worse and so in this example for Dan to write a great book he has to be an awful father and he ends up in the exact same spot look Linda they're both out of five it's kind of bad and nothing has happened so it's this really really nihilistic viewpoint from the system's design right you're just saying like wow you just have this set reserve of emotional energy that you can't spread around and you can't make any extra effort and life is just sort of this I don't know ugly thing and so players started responding to the essay I was like oh well cuz in my head I was just like oh this is so simple and beautiful as plus two minus one every time and it just say you know and and but but then I ask myself you know when I'm mapping this to relational context does that represent real life is there really no way that you can have a good career and have a good family I was like I know friends that do this like I know it's not impossible and and I didn't think that that that was the right message for the game to send and I

had this realization where I've done everything I could to advocate authorship in terms of the narrative content I tried my hardest I'm sure I didn't succeed everywhere but I tried my hardest to make all these situations very sympathetic you in every chapter you said oh I can see why you know Linda would want this I would see write-down would want this and I could see why Tommy wanted this and it's up to you the player to decide I tried to make that as balanced as possible but when I hit this problem with the zero-sum system I realized that I couldn't abdicate authorship from the game as a whole I had only relocated my authorship from the narrative content to the actual design of the narrative simulation and further I created a game those sending a message that I never intended I'd build this relationship simulation and I felt like Clark Griswald when he plugs in the Christmas lights and the house doesn't light up the system design had become the implicit message of the game and I was basically telling players you can't have it all life is zero-sum you cannot win and it was this really brutal bummer of a game unintentional and so I'm sitting there like why are the lights

not coming on but but yeah so it was it was really upsetting and the the biggest thing is like I don't believe that perspective to be true I believe you can put in the extra effort and have a better life so people were taking this message from the game that I fundamentally disagreed with sounds like I make I made this and I don't agree with it and I realized I was editorializing through systems designed like the actual design of the system had become the message of the game so I knew I had to make a change and I started talking to my friends and thinking about people who actually do seem to have it all people who are successful at their jobs and have a love and wife and have a great relationship with their kids and I asked myself how do they do it and the answer is actually pretty simple they just try harder if a parent misses their kids soccer practice then they might take their kid out for ice cream the next day or take them to a movie to make up for it or if you can't make time to have dinner with your partner maybe you buy them a little gift somewhere in the week or maybe cook them a surprise meal later to make up for it and say hey you know I know I couldn't make it when we wanted to but you're so important I want to show it and you know everybody knows that you can't please everyone all the

time but if you do those extra things you actually can have a good relationship over time by sort of you know a friend of mine said a really a really great thing which was the goal isn't to live a balanced week it's to live a balanced life so it's just an acknowledgment that's you know you can't please everyone in every situation if you miss a deadline it will work maybe you stay later to work on it so you find ways to put in the extra effort to make these make these things go a little better so that's where the compromise system came into the game and the the goal I had was I wanted to map the concept of extra effort in life to extra effort in skillfulness in the game and so for extra effort I decided that the player would have to go down a character's entire clue chain and find their clue and understand what they wanted to do in order to earn a compromise with them and for skillfulness they had to also remain unspotted so they had to make it through the entire chapter it's a stealth game without being spotted so you can't upset the character too much they've still got to be in a good place and if you do both of those things you earn the ability to

create a compromise with them so remember the scene of Tommy sitting by the window where he's sad because his dad decided to the promo forms instead and broke the promise to him well now I implemented a system where if you discover his clue find out what he wants to do and you earn you remain unspotted then you get the ability to have a compromise with Tommy and so even if he can't take Tommy down to the beach for the whole day he can at least work late and make a little extra time to take him for an hour and a half and it's not all that Tommy wanted but it does show that he's loved and that you know Dan is trying and he's trying to you know do better for everybody and just like Linda speaking up for Tommy and the example before this needed to be reflected in the simulation itself so in a relationship system as it stood you know choosing dan and Linda's marriage would cause a one point drop on the novel and on parenthood but in the updated simulation if the player finds a compromise that lets Dan make it up to Tommy in some way then the Parenthood score would actually stay the same and not decrease by one which if you notice all the sudden adds a point to the

system I just think of them as happiness points because the more you have that sort of happier you can distribute them through the system and so whenever you add start adding points to the system and using compromises effectively you can actually please everyone and get a happy ending to the game so remember our atomic unit of every decision in the game or something you to goes up by two or down by one well now we add it in a third option where things could also stay the same and if you spread those out over time you can actually have a much much different picture of the story you're telling so here's the possibility space without compromises here's the possibility space with compromises as soon as you get to get something to a level you can keep it there for the rest of the game and you can always sort of preserve and again that's not to say that this is better in and of itself but it does give the player more opportunities to impact the narrative so in the previous system again the player was just rearranging the same 15 points throughout the game so at the top you can see the totals it's just the same amount of points but

in the new system if you use a compromise in each chapter then you see that here for example the Parenthood rating of five stayed the same instead of going down and so if you use these carefully throughout the whole game then you can slowly increase their happiness and by the end of the game you've got 24 points in the system and you can actually have a happy ending for everybody because you manage your points and you're a good person and you love them all and extra effort pays off and life is awesome so the compromise system I would say is definitely my favorite piece of message through design in the game because it Maps a real world inspiration which is making the extra effort into game system a game system of making the extra effort and being careful and being skill all in your gameplay and it also and it pays off in the narrative layer so it kind of all comes together to create a message that I was much more comfortable with in the game so at this point I had a simulation that was in line with the message I wanted to send players had good feedback about the status of these relationships the story had a definable shape the characters responded in ways

that were believable in the edge cases and the gameplay was mapped to this idea of relational effort and it wasn't easy but with careful and skillful play you actually could get a happy ending for every character so I felt like I finally gotten the lights working and was looking at my beautiful house and felt proud of what I'd done but it of course didn't quite work out that way because I worked so hard to hide these numbers from the characters and I always thought that was a good idea or from players I always thought that was a good idea I never wanted you to be thinking of these meters when you're playing I always wanted it to be focused on you know story and characters and not the simulation and that ended up working against me in this case so let's look at an example so here's a playthrough where so far the player has chosen a novel twice and he's done a compromise once for Tommy and once for Linda so he's got you know up to 17 points in his system so so far so good he's he's playing it the right way but what if he chooses the novel again oh you can't go higher than 9 but the marriage and Parenthood scores are still

subject to the same rule so all of a sudden the player is done were they perceived to be the right thing what the game is saying the right thing is the right thing and yet you've actually lost a point from the system and the problem is that even players who are engaging with the game and playing the right way and trying to make the good choices and even when the on-screen feedback was saying they were making the right choices they were under the hood sort of getting failed by the system itself because there wasn't enough dynamic range in the system and even when players became aware of how the system worked at a high level and really tried to game it you know people are like okay I'll do dance Linda Tommy Dan Linda Tommy just try to cycle through the game and try to keep everything in balance I kind of got sort of caught off guard by the fact that people got so emotionally wrapped up in the content of the game to just say okay cool it's Tommy's chapter and they get ready to go in the back oh man I got to choose Linda here because the actual just subject matter of the chapter would make them feel oh my god I can't not take her side even though I know I'm supposed to give this one Atami and so they would end up sometimes this

will lead them to bump into the ceiling and and without knowing it basically take points out of the system and make things worse because in a weird way it was bad that they were connecting emotionally with the material forming from a purely number standpoint I guess so if you did this only twice in the game you lost four points from the system because every time you hit the ceiling you lose a point but then it takes another chapter to get that point back so if you do that twice you can only get up to 20 instead of 24 and if you see there with only 20 points in the system one character still gets a bad ending so this is a huge bummer because players often played the game well and yet they still ended up with one character who had an unhappy ending which was never my intent and in fact the game had kind of gotten back to our sending this unintentional message that is not what I wanted to say about life but unfortunately this is one that I did not realize until after the camp was released so I started seeing reviews and getting feedback from players saying that there's no way to have a happy

ending someone's always letdown it's this depressing thing and I was confused I was like what are you talking about you can make everybody happy I built the game that way it's supposed to work that way what why are you guys playing it wrong which of course is idiotic on my part but in the end I realized I had left this glitch essentially a glitch in the system that made it possible to play the game well and yet have the game send the exact message that I had tried to avoid and the hell of it is it wouldn't have even been that hard to fix like it's it's a very small thing but I just didn't understand the scope of the problem until after it was released like I sort of I knew what could happen but I was like oh yeah but if you just get to 9 just don't pick that one again but then that was because I was sitting here thinking is the person who understands the system as opposed to a player who this has none of this is exposed to and I'd sort of assumed a little familiarity that wasn't actually realistic that said I can rationalize this away you can argue that that scoring system is actually partially fair because to get there you do have to really zero in on one of those relationships for like three chapters in a row and if you do that you are kind of neglecting the

other people and maybe you do get what you asked for and so maybe that mitigates some of the problem even though ultimately I do think that's still a flaw in the game so in the end of mistake there was just assuming that players had knowledge of the systems that they didn't and did they understood it to agree that they had no way of understanding it and that was just bad on my part and so despite you know going to great links to to hide the scoring and making an analogue that actually worked against me so that's a bummer but uh frankly the more I think about it because this is something really I've only been like experiencing with the game since it came out I think I've given the choice I would rather err on the side of having choices be a little more difficult and having the consequences being a little harsher because frankly I don't want it to be too easy to get a happy ending because meaningful choices are the root of memorable experiences for the players and if they're a little too harsh that you know I wish they weren't like that but that might make players think more about those choices and and put a little

more thought into it and connect with it a little more and in life it's hard to have it all and sometimes there is a system that works against you in ways that you don't even know so I've kind of halfway talked myself into this bug being a subconsciously intentional feature and not a book but it's it's it's a book but but but the more I tell myself that the more I'll you know say was my authorial intent and you guys just don't understand my brilliance and all that so anyway I'm here I want to finish with a couple other lessons and conclusions today that I got from working on the game creating a dynamic narrative is really really really hard I thought it would be hard when I set out and it was even harder than I expected because it's really really difficult to take away some fundamental assumptions about how game stories are authored and then work backwards so that the player still perceives it as the same type of narrative they're used to getting because players don't give a crap how it works they just want to have a good story and a good experience so it's on you to try to make as much of that happen as possible while still doing all

your weird aspirational stuff under the hood to try to make a different type of game so it was very very difficult and I felt like I was constantly battling uphill to take these under you know under the hood simulations and really feed them back to the player and and construct a story out of it instead of it just being this idea of system soup it also doesn't save much work I had this night naive idea when I started that since it rooted in simulation character relationships there would actually be less work than authoring you know even a branching story even a linear story because I was like well you know I can take a lot of you know advantages you know procedural stuff and systemic stuff but the way that I built my story I had to account for all of the outcomes and all of the chapters so every chapter I had to have Dan's good middle and bad outcome and Linda's good middle and bad and Tommy's good middle and bad and with Linda and Tommy when they speak up for each other I actually had to have their conditional ones too so they had five outcomes per chapter so I had to build 13 outcomes for nine chapters and so it didn't really actually save in the amount of time and I bet it was a wash compared to doing this as a as a

traditional branching story I do think it has other advantages but it definitely didn't save me the kind of work that I did I thought it did and this last one is important to the the the feedback burden is really really really high with this kind of thing when a game has a non standard narrative loop and when the story reacts the player choices based on these personal values you have to use every possible channel to tell the player what's going on this is a battle that I had constantly throughout development and every every builder you know the first build I sent out people were like I literally don't know what's happening I do not know what's going on with this game I can't figure it out it's a confusing mess and so I added one set of things and a next set of things and I added act breaks and then I added memory stuff and then like I constantly was sort of backfilling to try to create enough feedback so players would even just literally understand the game itself you know when you're doing an authored plot even a branching one you can control what the player season when you can you'll guarantee they're

going to get to this point and get this piece of narrative information but when you're trying to build this out of a simulation like I was trying to do you can't take a lot of stuff for granted and you can't really funnel the player too much so you have to just basically tight take advantage of every channel you can to try to get get your point across to the player and you basic have to scrap and claw for every piece of feedback that you can find so I heard a litter it all over the game and it took a lot of work to get there but the last point is I think it was all worth it I I don't regret building the game this way players got really invested with the subject matter and they had real really personal experiences with it I know it sounds like a cliche but a lot of game reviewers talk about how they the game made them cry multiple times I got I get emails from players sharing their personal experiences with the game and people who have reconciled with their partners over playing the game and making them think about how they treat other people I've had people reconsider how they interact with their families and change their behavior based on seeing how their

choices affect others and so that really means a lot to me and and I think that those emotional responses from from people really come from the fact that when you give players the ability to put their own values and choices into a story in a real way I think they recognize that and they value it and they understand it and when they get that kind of personal experience they they know it's not something that can come from another medium and that really means a lot to players and so I think this is really something that's worth pursuing and finally I just think it's important to take risks and try to find new ways of doing things like my game certainly is not perfect but my hope is that people will take some inspiration from it and and and learn maybe even if you just learned one thing from it it'll be cool you know to see people take some of this stuff further and continue to take games into places that only games can go because it's important to take risks even if they don't work out the way you thought you can still learn a lot and and I learned a lot over making over the course this game so you know in the end I'm glad I made it and I'm very proud of it and I hope that what I learned in

making it will be helpful to you all as well so thank you very much and I don't I don't know if we have time for Q&A or if anybody has questions but it's you know oh look they were you know my name is Clint I have one question with regards to the compromise system to build on your example you know JB Chase has to like I forget his name in the movies but whatever he has to like learn to better empathize with the people in his family and learn to see why they're upset or why they're not getting along with each other and he has to improve himself by by examining himself changing his behavior and working harder I think right as you put it yeah in order to fix those things and get a get a happy outcome in the novelist Dan doesn't go through that process

you go through that process as a player but then the way you work harder is instead of by trying to be more empathetic or trying to help people with their challenges you do it by sneaking from lightbulbs and trying to not be seen and trying to dig down an exploration like tree to find hidden puzzle pieces mm-hmm so do you think that there's something dissonant there or something that could be better developed in order to to put what the player is doing when he's bringing his relationship management skills to bear right in line with the skills he's actually leveraging as a player I think it's a good point the the one thing that I would say does mirror that process a little bit is the fact that you don't have to find everybody's clues to go through the game you can if you show up in a chapter in black oh man I just want to do what Dan wants you can just mainline his Cluetrain and go straight ahead so if you decide that you want to empathize with other people you have to go through the steps of finding out everything about them so you actually have to discover what they want you don't just get it handed to you so I think that mirrors it some and so but but that's

probably fairly metaphorical within the game like I don't know how to make Dan have an empathy process but but I do think that as a player recognizing that you have this ability and then having to do that extra work and say okay I'm not only do I feel bad about what happened to Linda last chapter I'm gonna really do the work to find out what she wants this chapter I think that mirrors some of that one of the weirdest things that I noticed is that watching people play some people will earn a compromise and then not use them which is weird because I put a lot of training to like remind you that you have them when you can use them and people are like now I'm not gonna do that and it's like why don't you understand you can make people happier like that it's very that's a very weird thing I don't know how to explain that but but I do think that that clue discovery process is some of what you're talking about hi so I was wondering after you've tried to create this system and decided to

unless like not going to the authorial route and you know how when you go or you're like like putting together story where the main character really changes and has his weaknesses and goes through like many things to like have the final changing is my final decision that's very difficult to put together something that's good that the audience were watching will read and it will have closure at the end so after you've tried the other route when you see yourself working on your next game I use gonna still push for that other route are you gonna come not compromise but like try to blend both routes how will you see that for your future games I'm sorry so just to be clear your contrasting the more authorial types of linear or branching stories where you do see the characters specifically go through that arc versus my this this style which that doesn't happen is what you're asking

oh the thing I have no idea what a game I'm gonna make next it's definitely not going to be nearly as heavy because this is too hard to work on but the the name the novelist is is a little bit and and some of the visual metaphors of the game where you see the words come up and be typed out describing what happened it's a pretty it's a almost unveiled does not even veiled metaphor for you the player being the person who's going through that process and so to me it wasn't so much Dan it was never important to me for Dan specifically to go through that process because the thing I was hoping for and the thing that I found was that they were the ones who that happened to when when they went through they came in with one set of assumptions and then we started seeing the way people reacted to it they actually had they're like oh my god maybe I should think this way or oh my god I never really realized how strongly I felt about that or whatever so I think that the same process happened just maybe in two different places and in this game it's a little bit more personal and more metaphorical as opposed to going along for the ride on a

linear story which can be awesome as well but yeah does that answer your question at all I mean yeah right but but but going forward I honestly don't know I mean my bias is always towards this type of thing I've just always been a very systemic like I'm not a writer I I came from a systems design background I had to learn how to write to make this game like so this is not natural to me at all so that that's why I did a story from a nerdy systemic standpoint thank you you're welcome I was wondering if you so in the simulation it seems that in each chapter you have a decision and regardless of whom you pick like the numbers kind of going the same well I mean barring compromises but what I mean is that the weight of the decisions like maybe did you consider that there could be a decision where if you miss this deadline you might really get screwed for instance so maybe like the the score would fall more than that and then at the same time like instead of having particular you know author chapters that it could be sort of like a

randomly drawn card event that would define some of these things yes so so to the first question which was basically about should different chapters have had maybe different or more weighted scoring I I considered that and I thought about doing it with the with major chapters those act break chapters but as you see as I explained like the the dynamic range of the system is actually pretty small and so I was like you know if you like if I had given you a three point boost and you were at a seven and all of a sudden you would even bust further out of out of the system so I ultimately shied away from that there's one example are one exception to that I'm the final chapter of the game is structured a little bit differently than the previous eight and it actually does have different scoring and I actually the scoring was a little too harsh in the shipping version so whenever I an update in February I silently made it less harsh so so yeah I do do that one time it's the final ending that projects forward in the game so it is some I thought about but I just I was scared

that it would be too disruptive during the middle part of the game and then I've forgotten your second question the second question is I suppose I could probably answer it I imagine it has to do with authorship like that you would have had to create a lot more content but that for replay value you could there would be an option of like each chapter could be like a randomly drawn situation that would define some of these yeah I thought about some of that too and people suggested that because the game is very rigid it's it's a very rigid game you do some stuff and you get a decision you do some stuff you get a decision and it's the same process over the over time and that's one of the valid criticisms of the game and I thought about trying to that kind of thing and actually I think that I think Zac might have suggested something like that at some point I think in some of your feedback yeah I'm putting this on you now if you want against ascribes no but but but there some people suggested that and I don't even remember why I didn't do it but I think I was just scared that I was trying to get away from that process of system soup and like having really number Z and really

you know really tweak e stuff I was trying to get away from that and it was very hard even with the rigid structure to create a story so I was a little worried that doing that kind of because that's the stuff that works great in the sims or other games like that but I I think it was just more that I was trying to do as little bit as possible for the sake of it being a narrative as opposed to being a little bit more oh I'll I've you know I've got I've got this compromised point my back pocket or just owe it to you or whatever so I just think it didn't fit with this specific game but it's not a bad idea overall thank you no problem so this is we have two minutes as apparently the last question all the ghosts that are standing in line I have a comment and a question regarding people who don't take up compromises I did that once when I played through and there was one client I know what just I was like I thought like you guys are stupid I don't want to do that for you it's just I don't know well that's your experience and that's that's fine yeah I'm a Christian you Corps I'm just saying like that was my experience when I was playing once in a while just like why do you want that you

know so that that's good we made your own decision I was getting great I was getting into it or I was getting emotionally invested so so it's yeah it's a compliment not it's not a bad thing but but that's an answer to why's that might happen oh that what happened for me the question I have creates a little bit broad I was curious if you were I know you said you want to back away from something like this for your next game but if you were to explore what would be like the next step and dynamic narrative what would you kind of try to poke at well I would try to find something that was a little less rigid structurally like and and some of that one minute and so and some of that might be I am one minute I just do a blank I had a good point too so some of that could just be relaxing the structure of the game obviously fixing some of the issues with the dynamic range of the game I would also want to have a little bit more change in the game like one of the Knox's like the game takes place in one house for the whole summer so that gets really boring so I think you could do interesting stuff if you put them in different situations or if you had a scenario where someone's out of town and there's only two characters there you

kind of had a little bit more variation to the experience but um overall I I think I would I think we're just keep play testing and I'm a big iterative person so I would probably just keep play testing and trying to find different little bugs and quarks with the system and try to open it up more and make it a little more I don't know a little less rigid I think would would be a good thing so yeah I think that's it zero minutes all right thank you very much for coming