Choice Architecture, Player Expression, and Narrative Design in Fallout: New Vegas

[Music] hello can anyone hear me everyone hear me some portion of people oh wow hi hi welcome welcome hi Steve excuse me are you guys ready to rock out about Choice architecture yeah wow it's a very exciting topic um welcome the name of this talk is really dumb and I apologize it's do say the right thing Choice architecture player expression and narrative design in Fallout New Vegas uh I'm going to be talking a lot about Fallout New Vegas but also a lot of other things my name is Josh Sawyer I'm a Project Director with Obsidian Entertainment before we start there are a few things that GDC would like me to go over for you all one please uh turn off your cell phones or set them to vibrate or telepathic link or whatever you know sort of cool features are coming out these days um basically just turn the ringer off also

you should be receiving an evaluation form via email uh please fill it out because it helps GDC evaluate things and do cool awesome stuff awesome let us start so just a little bit of background um I'm a Project Director at obsin entertainment and a lead designer my background is in design uh I started design at black is Studios back in the late '90s on the ice wind series I am primarily a system designer and adapter of systems uh I've worked on a lot of Dungeons and Dragons games so I've had to do a lot of work adapting rule systems uh even though I do writing my job is not primarily doing writing I work a lot on the mechanics surrounding writing which is why I wanted to give this talk I'll get a little bit into the sort of difference between the pros and the choice architecture but I focus primarily on system mechanics uh my role on Fallout New Vegas was as a lead designer system designer and Project Director uh just to sort of let it be

known there were actually a number of people that contributed to the dialogue on fall at New Vegas John Gonzalez was actually our creative director and he wrote the story uh we had a lot of great writers on the project that all contributed to making it a very open narrative experience uh I am a tabletop RPG player I been a tabletop RPG player for a very very long time uh I love playing computer role playing games but I considered myself to be a fairly broadspectrum gamer now uh I used to be very focused on PC RPGs but now I play lots of console games actually games racing games all sorts of stuff so that's my background all right hold on I've already screwed up this presentation this is awesome yes okay so this talk is really about branching dialogue and the sort of challenges that come up with branching dialogue branching dialogue is not necessary it kind of goes without saying but I I thought it needed to be addressed it's not necessary or even

desirable for lots of games there there are a lot of different styles of dialogue that you can approach in a game uh these are the three sort of broad categories that I identified When I Was preparing this talk the first is linear or cinematic this is arguably the most common that you'll see in games especially nowadays um sort of exemplified by games like the Uncharted series the story is written the characters are established you aren't really the player doesn't really have agency and how those characters interact with each other how the story unfolds it's a cinematic experience writing's Great Looks awesome and the focus is really on the gameplay and you're just the the Cinematic storytelling and the dialogue is there to serve a narrative function it's not about player agency totally valid cool thing to do uh another way of approaching dialogue is what I call keyword or subject dialogue this is also not really about player agency this isn't about saying to the player you get to shape the story it's about saying to the player hey do you want to learn more about this setting do you want to learn more about this world do you want to trigger this Quest now and so a game like the Diablo or the Diablo series is this where you have a

list of things to talk about and the player is really just saying like tell me about this thing tell me about this thing tell me about this thing so again it's not about granting player agency it's really about allowing the player to make a choice as to when they want to receive any bit of narrative content what this talk is about though is about branching trees this is dialogue where you have menus of options presented in a variety of ways where the dialogue can flow Branch out Branch out even more Circle back on itself ideally it's producing gameplay changes it's producing story changes uh it's not a whole lot of companies necessarily do this I'm glad so many of you came I wasn't sure how broad the appeal would be but obviously companies like BioWare Exel at it black Isle uh we did it back at black Isle we continue to do it at obsidian Bethesda also does it I'm sure there are other companies um that do it as well but it's um it's a tricky thing and that's why I'm having this talk so someone might ask why why even use branching trees what's the advantage um you know you can get all sort of cool

cinematic stuffff and cool character development with good voice acting and everything in a purely cinematic view why even use branching trees there are a few very good reasons for this first is you get to Grant story agency to the player you get to say to the player hey you know what you have some role in defining where the story goes and how it unfolds it's very important for a lot of people especially in role playing games uh for a lot of people who came from a uh table top or pen and paper background they really really enjoy the ability to screw with the story in some ways and make it their own you also Grant character agency and I want to differentiate that from story agency you're not just changing the story but you have the ability to express what type of person your character is so if you're a jackass or you're a really cool guy or you want to smooch some girl or some whatever that's that's why we build that in there to allow you to express the type of character you are not just How The Story Goes but how you make it unfold also it forces you to make tactical and strategic choices um again this is sort of ideally this is is what I'm going towards in the talk but good

dialogue is gameplay and the player should look at it as a series of choices that they're making short-term choices and long-term choices and when it's done very well the player feels very engaged they're paying attention they pause to think about what they want to do they feel satisfaction and the sting of of making sacrifices along the way and that produces great gameplay through dialogue and also it produces reactive narrative content this can take place outside of the dialogue itself this can be people giving what we call Bark strings in the world someone saying like hey man thanks for you know blowing up our town you [ __ ] or whatever you know like it's that all that sort of Rich World reactivity that doesn't really need to be deep uh it just needs to be actually more widespread and shallow but it makes the player feel that the world pays attention to the things that he or she does so this talk is not about writing Pros uh there are writers round tables actually one's going on right now strangely enough um they're great writers roundt that are going on obviously Austin GDC has a bunch of cool

writer stuff going on this isn't really about pros and writing good dialogue from the perspective of Aesthetics it's about choices uh the reason why I want to start with Choice architecture is because it's it's under it's it should come under and before Pros uh choices inform the pros that you write but choices present problems on their own a lot of problems and they need to be thought through and choice is the reason to have branching dialogue in the first place so it should be the place where we start to think about what we're doing and how we want to do it so I couldn't quite pinpoint the exact uh time and place where sort of branching narrative became a thing but one of the common things that came up was with an author Jorge Luis vores and he wrote a book called The Garden of forking paths uh and this was uh pretty radical at the time because the story could be read in a lot of different ways and the reader would arrive at different conclusions about what the narrative was and where it went um and uh someone

writing a few years ago about this uh had this cool quote that actually when I can see it on my screen I'll read out to you uh bores confronts us with the pullulating moment when we become aware of all the possible choices we might make all the ways in which we might intersect one another for good or evil his imagined Garden of forking paths is both a book and a landscape a book that has the shape of a labyrinth that folds back upon itself in infinite regression this is the definition of sort of the possibility space for the player uh when we empower the player to make choices and then reward them for making choices they become aware of the power they have in the story and they start to embrace it if we consistently reward that they feel like they have a very strong role as a player in determining the outcome and they engage with the story much more than if they're just watching it because they realize they're an active participant in it gez Louise this is when I I actually started

caring about it I was strangely enough I was not reading borus when I was 5 years old um I was reading uh the cave of time which is my favorite of the Choose Your Own Adventure series uh how many I'm curious how many people have read Choose Your Own Adventure books yeah there great series of books awesome yes yes so Choose Your Own Adventure books are they're great um fantastic stuff every book is a different sort of style of book or genre of book but they all had that idea of branching narrative um however if you were even a remotely smart kid what you started to do after you read a few of these books was you start to put your finger in the page before you made a choice and just to draw a parallel this is not really that different from abusing A Safe Load mechanic um which is actually a big point of some of the things I come to later so Choose Your Own Adventure books are the start of this but even they had underlying problems that still persist now in branching dialogue and role playing games then later on I really got into

role play games tabletop role playing games and I played with a lot of different DMS and I was the DM and a lot of different players and the stuff that I enjoyed the most and the stuff my players seemed to enjoy the most is when the story was constructed in such a way that the players could [ __ ] it up basically they could they could screw with it and they could push it and they could pull it and the they could smack the quest giver and like throw them down a well and and and the story could keep working and and then it rolled with it and that was so satisfying to them and often there were consequences for doing that sort of stuff but they felt very empowered in that environment especially because that was just it's up to the people there in their imagination so that feeling can be hard to emulate in a computer environment but that's what ideally we try to do and so this this had a lasting effect on me I played tons of these games and the most had its fine times were the ones where I was able to do that uh dialog trees in RPGs my first experience with it was with the original Fallout Tim Kan gave a great postmortem this uh postmortem this morning on Fallout I was in college when it came out and I was just enthralled with it I

thought it was great I loved how you could make choices you could be an evil character you could be a sort of neutral character you could do all sorts of different stuff and the game didn't it wasn't like a grand karmic wheel was there to smash you in the face after you made a bad choice people reacted to the bad things that you did but the game itself did not punish you the designer did not punish you for playing a certain way also when I first came to Black Isle uh I my first job there was as a Web Master on the web team and I made the planescape torment website and I was working with the development team and obviously torment is very very widely uh regarded as a great game with amazing choice and consequence huge amount of dialogue lots of reactivity um and so that had an impression on me because when I came into that environment I saw these guys that were just busting ass to give them you know all this choice and consequence to the player so that was the environment that I came into when I came to Black aisle so dialogue trees today I I had to get that ass shot in

there had to haven't seen enough of Mass Effect 3 I there might be a more upto-date ASA that I could have put in there but I didn't have time so dialog trees and crpgs today they're not really that different in a lot of ways um they're presented different the Aesthetics are a lot different but um kind of fundamentally on the under the hood how they work they haven't changed a whole lot so the changes have been primarily aesthetic not really structural um but there are persistent flaws and to me in my mind uh that sort of suggests a lack of analysis to this point in time uh I've been thinking about this a lot over the past few games I've been working on um and it's been difficult because I've been writing in this mode and this style for so long but I've made this talk because I wanted to figure out like what the hell are we still doing wrong like this is the way I want to write dialogues but there's got to be some there are some underlying flaws here that I need to dig at and get out we did it also it's just it's uh yeah it's not it's not just um you know

other other companies that are doing it and hopefully I will provide more examples of us screwing up in this talk than other companies although I will give examples of other companies screwing up on New Vegas um our writers did a great job John Gonzalez did a great job all the writers did a great job um worked very fast using an engine we weren't familiar with um but there were certain structural flaws to it um mechanical things that I I saw and other people saw and they're kind of like this still kind of feels stinky there's got to be a better way to handle this so the core problem for me is about Choice conflict um what I mean by choice conflict is when the player's goals and uh there the the goals of the player as a um game wise conflict with the goals of the character as you know an expression of the player so the player has a character who's like a sassy jackass or whatever the player wants to get a thing if they play their character as a sassy jackass they can't get the thing so when we do that as designers we're saying your character concept is

invalid for this game um that's an extreme example but that's really the root of the issue uh is that doing or saying the right thing meaning the thing that is actually going to progress the game in the most favorable way conflicts with the idea of a right character um the two examples I've listed here are dex's uh social boss battles which no offense I I actually thought DSX Human Revolution was a fantastic game I liked how they presented their conversations but the conversations felt like puzzles they didn't feel like things that Grant player agency because if you choose to play a character in a certain way in that conversation you will lose the conversation however similarly in Fallout New Vegas we have dialogue skills like speech and barter we also have dialogue perks and they become you in the conversation choices uh and the choice here talking with um I can't remember that guy's name Santiago I think when you're talking with Santiago if you pick anything other than the speech option you're basically intentionally handicapping yourself so if you want to role play a certain way we're essentially saying okay thanks for being a dummy uh you are going to lose

out in this situation so to repeat this this is the main subject of this talk that I screwed up yet again is that doing or saying the right thing to get a gameplay advantage or they gameplay advantage to move it forward the way you want to conflicts with being the right character and you may be saying well sassy man what's the right character that's your voice in my head by the way um the the answer is any character that falls within our supported range of expression for the player that's a right character if we allow you to play out a certain fantasy that's a right character some of this has to do with consistency if we consistently reinforce something for the player and then pull the rug out from under him or her bad news so when we look at what's a valid character we Define that expression range and we have to support it if we don't we're letting the player down we're saying yeah we kind of allow you to play this but not really it's it's mostly cosmetic and you know sometimes you'll get something out of it but most of the time you're going

to get screwed that's the problem it's a [Music] game it's not a puzzle puzzles are not really games there are right answers and wrong answers it is not a movie it is not something you sort of sit back and watch and it is not a book it is not just for reading you're making choices to go through it um it's a game and if we Embrace dialogue as gameplay and not things that are sort of ancillary to the game and treat it as like well yeah there's the game and then there's the story if you're doing branching dialogue I think it's important that and again you don't have to do branching dialogue it's not important for all games but if you're going to do that then you should do it as well as you can and you should incorporate it into the gameplay so my assertion here is that if you plan well you can avoid Choice conflict when all the ways that you can go through the game are considered validated ways and you give a clear response and reactivity Spectrum to the player meaning you make a choice you get this feedback you make this choice you

get this feedback when that happens the player benefits and the character uh roles align there is not a conflict because you've already planned out hey if a player wants to be a crazy jackass they're going to make these sacrifices but they're going to get this out of it and they're going to feel good about it but it takes planning it you can't just like decide to do it uh it doesn't happen spontaneously you have to look at three things when you're when you're planning this stuff out the types of choices you want to allow the player to make the range of player and character expression this is sort of like well what sort of characters are actually valid within this game as an example in Fallout one you could uh kill kids wowe um in Fallout 2 you could uh marry somebody and then uh basically sell them as a you know into like pornom uh so that's like a range of expression that says like okay the goal posts are really far apart here um so once you've sort of establish that that's what

you're going for and then finally your narrative goals uh you as designers you also have to accomplish certain things you have to communicate certain things to the players and those can be in conflict with what the player wants to do if you say hey player you can be a sassy jackass but if you do that you're going to miss all this critical information for playing the game that's uh that's kind of sucks so here's here's my systemic approach to doing this consider the types of choices that you're going to make Define the range of player character expression establish the narrative goals that you have for the game that you need to communicate then get some perspective on how these three things work together then you write your pros and again this is not about Pros that's a a separate sort of subject that can be a rich tapestry of interesting things um not here though and then everyone goes my God your game is so amazing thank you I got to play exactly the character I wanted you're the best so Step One is considering the

types of choices you can make so you want to provide tactical choices to the player tactical choices are what I would call node by node choices in the conversation every node where you're allowed to make a choice there should be some sort of tactical decision there uh those are immediate reactions strategic choices this is long-term planning this is not necessarily about I'm going to say this and this guy's going to get mad at me and then this girl's going to give me a kiss it's about um you're sort of like building reputations with people indirect reputation mechanics things like that you want to forecast consequences uh there's actually a pretty cool anecdote that Tim Kan brought up in the Q&A session of Fallout one which is about the ending of junktown in Fallout one so junk toown initially the endings were if you put kilan Darkwater who's like the good guy in Power then junktown gets screwed and if you put uh Gizmo who's the scumbag in charge junktown actually does very well but the consequences of that weren't really forecast it was kind of just it was a reversal it just had haha actually Killian sucks um and because those for

those consequences weren't forecast it felt weird to a lot of people even though it was arguably more interesting and then finally you have to validate all options to avoid the perception of win or loss the perception of well I guess I just made the wrong choice I'm playing the wrong sort of character I should go back and try to get something else out of it so some words on Choice architecture hopefully I won't screw this up yet another time as I go to look at it uh so I've been reading an interesting book by two uh professors from University of Chicago called nudge and it's about Choice architecture and a lot of other things and there's this interesting quote that one of the authors gave at a convention about Choice architecture so Choice Architects must choose something you have to medal for example you can't design a neutral building there is no such thing a building must have doors elevators restrooms all of these details influence choices people make uh there is nothing neutral about putting an attack reply at the top of a dialog tree

structure there is nothing neutral about putting a goodbye reply at the end of the dialogue tree structure these are consistent choices that you make how you lay these things out how you present them to the player that has a huge impact on how they perceive them and how they make choices um in obsidian games forever in Black aisle games usually the default option for everything is always the top one because we find that people tend to just gravitate there if you put an attack node there oo people get mad because start going one one one God damn it why am I in combat um so it it matters that goes back to this idea you really do have to think about how you lay these things out oh my God yet again what am I doing how do I use PowerPoint so node by node choices just to go back to this node by node choices are again you get to a point in dialogue the guy says yeah what are you g to do about it buddy and then you have three options you know you can say well I'm sorry let's work it out I'm a diplomatic guy or I hate your guts I'm going to not you in the face um that's a tactical

choice there's going to be immediate reactivity and feedback to that so that's tactical Choice it's shortterm these are things ideally that happen quite a bit in conversation uh that usually require a lot of of scripting so may they can't happen all over the place but uh these are the things that people really react to where you know was the conversation in Mass Effect one where the reporter is saying something to Shephard and he's just like you know punches her in the face um you know people like oh man million YouTube videos I'm the first guy to post this uh it's a tactical Choice um very cool there are also strategic choices to make though uh one thing I identified in here that that works generally very well for this are what I I just made up this term indirect reaction systems these would be things like Karma or reputation or loyalty or whatever uh you don't get necessarily an immediate feedback out of it but you're putting you're putting pennies in the the Rainy Day event reactivity fund so you know you you kind of sass someone and you know NCR gets a

little grumpy with you and that you know you get points in the negative pool and then you do some stuff for Caesar's Legion and that you get points in the happy pool and then eventually you know some NCR Ranger comes up to you and says you're going to [ __ ] die bro and shoots you um so it's a long-term outcome it's a strategic decision it's not about having a big scripted reaction it's about a little reaction that builds up over time forecasting consequences so you know I gave that example of the choice from from Fallout one so the idea is that you want to Intrigue the player you don't you don't need to spell everything out but you kind of have to forecast what's going to happen in a writer's Round Table yesterday one of the guys gave an example of I don't actually know what game he was talking about but he said he picked a line that seemed like kind of aggressive and then he wound up just like gunning a bunch of people down he was like whoa whoa whoa I didn't I didn't quite want to be that aggressive uh so Intrigue the player uh but inform them of generally what's going to happen you can surprise them but don't make them angry so if you surprise them in a pleasant way awesome great that's that's

the joy the joy of story uh as it unfolds but don't make them mad because then then they just feel like they were sort of tricked into picking something that they didn't want to take uh differentiate the consequences this is super duper important uh false choices not all players pick up on them but players are simultaneously smarter and dumber than we think they are and a lot of them do pick up on fake choices they sense it very quickly especially if a game uses them a lot so I would say let's just stop doing it everybody let's stop doing it thank you um lace consequences with sacrifice so I'll get into some differences between in Choice Agony but when a person makes a choice uh they should feel like they're making a tradeoff the idea with sacrifice is always that you get something more out of it than you lose because that's why you made the sacrifice but you should feel like you did make a trade-off there and ideally it's a bunch of different options where there's some pros and cons and they're

not easy choices but they don't have to be horribly intense uh and that Push Pull conflict produces Agony and that generally speaking is good when it doesn't go too far validating the options of choice and consequence again this goes back to when you make a choice you want to validate it say yes player this is a a correct way to play the game you're the character you chose to play we support that you can keep rolling with this if you make a similar Choice down the road it's also going to be supported keep on trucking good job this is in contrast to choice and no consequence I'll pick a little bit on Dragon Age here um Dragon Age had a lot of cool stuff in it this isn't one of them sorry um you have a bunch of options that people talk about and it doesn't produce anything there's absolutely no different ending result from you picking any of these options you can Circle through it over and over again and it's the same outcome it's kind of a waste of the player time and it's a waste of the writer's time so we've done this in the P oh sorry oh jeez we've done this in the past a lot of companies have done this in the past um I think we do it because we feel that

we want to give the player options but again players pick up on this stuff they get irritated with it if you do it a lot they get really bored of your dialogue because they feel like the choices they make don't mean anything versus win and loss so again fall at New Vegas we did this all over the place this is also bad um when you come into this situation unless you just want to fight which is you know you can still fight even after taking this option but this option makes it incredibly easy you can come in here you have a high speech and you can just say hey Manny I'm here to drop off some drugs see you and then you can just walk around and steal everything and then come back and see4 everyone or whatever you want to do um if you have a high speech you're kind of like shooting yourself in the foot if you don't pick this option so we're kind of saying like if you don't take this you're playing at a disadvantage not really that great or satisfying so I want to talk about Choice extremes so uh

that that's how to spell extremes by the way um officially so there kind of two ends of the spectrum here neither of which I feel are are particularly great uh one is the sort of Jesus Hitler option um again a lot of our games have had these I hopefully we've moved mostly away from these because we like moral ambiguity a lot Jesus Hitler is kind of like hey there's a baby do you want to give the baby a little smooch on the forehead or do you want to throw it into an oven um and and and those I mean we see those options and they're they're not uh they're not really compelling like if you've decided you want to play a crazy dude then there's no struggle to figure out what you want to do there's no sacrifice just like well yeah I'm a crazy baby killer there you go um or no like I'm I'm a super nice guy why would I ever do that so it's there's no real interesting choice to be made there it's very flat it is an option and it does it is differentiated hopefully in how it works out but it's not that compelling uh and then also then the other end of the extreme is Sophie's Choice spoilers for a uh 30 some year old book um so uh

yeah Merl st's character um she has to choose between which of her children will die no one wants to make that sort of a choice in a game uh there's no s it's not like wow I'm so glad I made that choice instead of that one it's just two shitty choices so a player should feel like there's something good and bad out of what they pick not just like well which limb do I lose or like which child do I sacrifice those choices they're interesting in the sense they produce some struggle but the player really doesn't feel good about having to make those choices so ideally the agony and in in the conflict is somewhere between these two spectrums it's not as simple as Jesus Hitler and it's not heart-wrenching total loss like Sophie's Choice uh just some examples from actually these are kind of nasty too um but they're differentiated in an important way so Choice Agony is a Greek Greek plays do this really really well and very consistently uh Greek plays Will spoilers for thousand year old plays

um so in this case the agony is not between two things that are to borrow an economics term that I learned yesterday fungible um they're not between like a child or this child or this arm or this arm in the case of arrestes and antigon it's a it's a choice between two kind of equally bad and equally good options but the sort of value arithmetic that goes into it is not equal um arrest father was killed by his mother and I think it's Apollo is saying yo you have to avenge your dad kill your mom and the Furies are saying dude you cannot kill your mom that is bad news he's like oh man and he kills his mom and then the furies chase him and oh my God how agonizing um Antony wants to honor her brother bury her brother I think it's her dad says you can't do that I forbid it and she's like wow this is agonizing and then she goes over over and does it so again the the value arithmetic there is not oh yeah well it's just 61 half doesn't to the other it's two different things and when a person reads that one

person might say well I value this more than that I value this more than that and that's what makes our individual choices more interesting as opposed to hey what color do you dislike more it's not to be confused with spice Agony just so you know choice Agony is not spice Agony arguably spice Agony is even cooler because it's from Dune so but it's not the same sorry guys uh so the second step I guess I'm doing okay on time even though I'm screwing around all the time um the second step is to define the range of character and player expression uh this is really about hey what sort of characters do you want to allow people to play and are you going to support through the choices that you give so you want to define a response and reactivity scope this means what can the player say and how are people going to react to it it's very important to figure out uh what sort of character prototypes I'm just using I made up that term because I didn't want to use archetype I'll get into that later what sort of character prototypes do you want to use support for character ideas and then create a range of impacts

that you can have these can be little impacts big impacts they can be physical impacts emotional impacts all over the place so figuring out the scope of responses and reactivity is very important um if you are making a game with very split uh morality good and evil that's your range is it about Shades of Gray where there's a lot of moral ambiguity are the responses aligned along things like stance so for example Alpha Protocol had stances there are number of games that have things like stances where where they're more like attitudes they're not really aligned along morality um and then uh what sort of reactivity are you going to have what sort of scripted or direct reactivity are you going to have and then also uh what sort of indirect reputation mechanics are you going to have reputation influence things like that so the player can get more frequent inputs into that system as an example Fallout new Vegas's scope and range it was really big um one of the cool things about the Fallout setting is that there's just so much crazy stuff in it uh and designers have

a lot of fun with that players usually have a lot of fun with it it doesn't always result in a consistent experience though um so there's a huge scope of responses anything goes at any scale so fall at New Vegas you couldn't kill children um because I don't think any games allow you to do that anymore but um you can kill everyone else like everyone uh what I told the designers is that you have to imagine that the player has a like Thanos kill field around them everywhere they go and they just can't help it they just everyone drops dead so so the whole game was structured so that you could kill everyone um even your one sort of fall through guy you could kill him and he died but he would then uh transfer his Consciousness into a different robot and come up to you so it was very important that we allowed that um also you can do all this crazy reverse triple betrayal stuff um because of our reputation mechanics and the way that they worked you could start out working for NCR and then kind of drift over to Caesar's Legion and then get on their good side and then oh we're actually going to help out Mr house and we're going to betray him and go independent so there's a lot of things and ways that

you could weave through the story um there's a lot of scripted reactivity like a crazy amount of scripted reactivity uh and all the credit for that goes to the the writing team and the area designers that made that stuff happen uh and reputation we had a reputation system for factions and for locations and that served as our indirect system where players would either perform actions or do dialogue things and resolve quests to feed into that system so now I want to talk about uh prototypes again I use this word instead of archetypes because I think that sometimes archetypes get uh that word kind of gets thrown around archetype is higher level than these guys archetypes are more abstracted than these individual characters prototypes to me in an in a role playing game are starting points for character Concepts they can be sole characters sometimes they're combinations of characters uh you know I want to play I I want to play Aragorn I want to play Snake pliskin I want to play Clint Eastwood's racist character from Gran Torino um by the way I want I really to make an RPG where you someone's like yes I'm

playing Clint Eastwood in Gran Torino um so the idea is that is that players really they like they gravitate towards these characters and players are creative but we all start somewhere and so some players will be like yeah I want to be the man with a no name I want to be Elizabeth cander from you know Girl the Dragon Tattoo I want to be blade whatever there's a starting point and then over time they usually expand from there but understanding what your game and your genre supports and where that's going to start is very helpful it's not just useful for def finding dialogue but other things like dress and appearance and costuming and things like that in fact in New Vegas I mean the number of uh clothing mods you see for people that are supporting the man like everyone is am the man no name great I'm Clint Eastwood in New Vegas great there's like a zillion different clothing packs that support that so understanding what people are going to want to play is very important what are the fall at New Vegas prototypes who knows because we didn't to find them Ahad of time unfortunately we knew we were going to have a huge range of uh influences and responses but we didn't really sit down and think of

this ahead of time unfortunately so yeah there's some stuff where you kind of you kind of feel like the man with no name uh or you also kind of think about you feel like Mad Max surprised these are the same character um one is in Australia one's in the United States they're kind of the same dude though um and like there's some stuff where you feel kind of like this weird spec ops guy that's that beard guy you know who I'm talking about it's that beard guy and uh oh I know he has a real name but everyone knows him as that beard guy um but you know but we didn't really fully support it and there's other things that we could have supported post-apocalyptic characters um like the all the crazy tribal tribal people from doomsday or even the like the tribal kids from uh Beyond Thunderdome and stuff like that but we didn't really think about it also there are other examples of like strong female characters like Alice from the uh Resident Evil series could be considered a post apocalyptic character um but we didn't really think through like what what are we trying to support unfortunately so the range of resp responses was granted in the Fallout Universe they're already all over the

place but they didn't really feel like you're playing consistent character necessarily you want to create a range of impacts um choices should have impact again this goes back to don't don't just kind of make fluff or meaningless choices you want to ground them and make the player feel like wow I'm I'm cool like that that actually did something it wasn't just sort of an aesthetic change you can also make that happen through indirect reaction systems when the player sees and some people complain about this tough [ __ ] uh when I say like plus one to reputation you know I know it's not it's my my immersion gets broken but you see hey that actually fed into the system great that's good feedback so as game designers even though some people get all worked up about it it's very good to reward people that way and then the important thing I think to remember is that narrative or emotional impact is the important scale there's examples of two big impacts from Fallout New Vegas one is aredes too this big space laser and that was in all the trailers uh I don't really think that choice had a big emotional impact on a lot of people um the only person that has a strong emotional reaction to it is

one of the companions arcade and it's pretty unlikely that you actually have him at that point uh it doesn't really influence a lot of things around it so even though it's like a big event there's not a lot of emotional strength to it so it's like wow the crazy space laser I don't care I'm just going to go on with my adventure um compare this to I think the DLC especially all the so I did one of the dlc's Honest Hearts um I'm only putting up there because this is the one that I did but um Chris Avalon was the Project Director on all the other DLCs other than Gun Runners Arsenal and they had great uh emotional content to them people really connected with the characters did a great job there um in Honest Hearts you know the sort of resolution of it people seem to respond very strongly to because it was a very tense moment between really just three people in a very small space uh but that sort of stuff can go a long way so the scale is really about emotional impact when it comes to change when the player feels like they're producing something that's really tugs at them I keep doing this and I shouldn't um that uh it makes them feel good about the choices that they make something I think was much better defined in terms of scope and prototype

characters was Alpha Protocol it was very clearly defined so um they had the professional and Suave and aggressive stances there were variants to that that would appear in dialogue but that's pretty much what was established for the the spectrum of responses a lot of heavily rescript uh scripted reactivity they also had a reputation mechanic that was used very frequently um and they had three prototypes it took a while to settle on this but it was the called the three JBS Jason Bourne James Bond Jack Bower equipment choice um dialog options all that sort of stuff kind of fed into that is like if I want to play like Jason Bourne I'm going to want to say this sort of thing if I want to play like Jack B I'm gon to smack this guy in the face and like get all aggro on him uh and that that really helped uh drive a lot of the choice people like oh okay yeah yeah I get those guys that's this is the space that we're working in which is very important so step three is about establishing narrative goals ideally you should fuse information and emotional content not do info dumps you should separate your critical information from optional information

this is very important when you remember hey players can go around smacking dudes and skipping over stuff and shortcutting all sorts of things in dialogue and you want to outline your narrative branching ahead of time you do a lot of people look at New Vegas um maybe not everyone but a lot of people look at it and they go like wow how did you plan all that stuff out and it's actually I don't want to say it's easy but when you do it from the beginning it is not especially hard if you try to do it 6 months down the road forget it um any that I've worked on where we've planned for this stuff ahead of time it's gone very very smoothly um once you start developing the game and you try to back into a bunch of freedom of play and movement and choice it doesn't go so well so don't do this um dialogue branching is I think at its weakest when it comes down to essentially being no more than a keyword list where you're saying hey tell me about this tell me about this tell me about this um I think there are much better ways to do it I think that Mass Effect putting stuff in the Codex is a really good way to do it other than the XP gain which I think should be gotten

rid of um also Bioshock does it really well they have like essentially three levels of interaction where there's the main story line there's the stuff you can hear through all the tape recorders and things like that thank you for reminding me of that and then also the um uh all the posters and things where you can infer and imply all sorts of other relationships by looking around so there are other ways to give information about the world and what's going on in it than just info dumping onto the player uh if you can give information about the world to the player try to sort of wrap it inside the emotional content of something else so a character is talking about something that's important to them and in the context of talking about that thing you're learning about the world if you just have someone sort of stand there and dryly tell you about the world they're just a narrator for an encyclopedia and that gets really boring you have to figure out what your critical and optional information is going to be in the game uh again very few players want information dumps and dialogue uh most players they do want to understand what's going on uh usually they're skipping through

things because it goes on for too long and tries to tell them much more than they actually need to know um if you write your story well the player should not need to know a huge amount of information to understand the meat of the plot and they really shouldn't need to know much about what they're doing in the next stage of their Adventure by the way hints hints and hanging references those are also Expos Exposition a person can reference something without defining it that's fine Let It Go like Let it hang there person will be intrigued look it up later maybe another character references that group or organization or religion later maybe it's never defined and the player goes huh I wonder what that is I've gotten so many questions for fall at New Vegas because Chief Handlin Chris christopherson's character says yeah all the veteran Rangers are hunting ghosts down in Baja and it's like every week I get two or three people saying what is he talking about I am never going to tell you what that is never never never like there something weird going on down there they're in Baja that sounds mysterious great just Let It Go that's a side thing cool that's

intriguing narrative branching uh obviously you have to guarantee that all the critical information is communicated if you allow the player to do something that breaks their understanding of the game that is very bad uh you have to have ways to handle that uh you want to isolate reactivity so reactivity is awesome and players love reactivity I think a trap people fall into is they want to make the reactivity really like there's a you know huge ripple effect from any like a pebble hits the water and you know like thousands of people are changed forever and it's it's very messy players don't actually notice it or appreciate it it causes a lot of bugs um you can actually do a lot of more like very localized reactivity and players feel like the world is very rich with the reactivity but it's very light scripting and bug bug uh fixing wise and then generally speaking more branching is better than bigger branches so it's it's more interesting to have like little little ways to alter the story and then feed into other things than to try to make one big massive narrative that branches out in a million different ways um just recombinant wise it it

becomes very difficult to manage that stuff and bug fix it so here's an example from fall at New Vegas that I think worked pretty darn well um they went that way was the first main quest that you go on so fall new Vega starts with you being shot in the head and dropped in a shallow grave and your first Quest is to go find the people that did it so your first steps are completely informational and independent of each other it's really just where did this dude go it's it's Benny the guy in the crazy Checker suit where did he go and you go from step to step and each person tells you something more that you follow and follow and follow because they're just information and they're independent of each other the player can actually go straight to New Vegas from good Springs without ever going to any of the other steps they can do things in different orders they can go straight to Novak through a pass they can go to Boulder City um they can do things in all sorts different order because the story still makes sense the story is find out who shot me and where that person is if you just walk into I guess the story still makes sense you miss little side details

actually you miss quite a few side details along the way but the story is flexible and by the way we just totally ripped off Fallout one for this just ripped it right off if you look at Fallout one the first critical step you leave the Vault find the water chip that is actually the first critical path thing that you need to do is find the water chip you can walk from vault 13 to the necropolis you do not have to do anything else in between that's the way the story makes sense because you're point at hey go check out vault oh it's not there oh I'll go back to this town oh what's that Raiders oh okay I'll go to the Raider den oh I'll go to the here so it unfolds in a very logical fashion but the player is not required to do things that way because it's all about information the reactivity is very localized to those places so reactivity is in that space and then later on maybe we'll have we do this all over New Vegas you do a quest hours later some random dude out in the middle of nowhere goes hey a reaction to the quest that you did and they're like oh my God this world is full of all this reactivity um but it's it's really just that one character doing that one thing and then it's

completely forgotten about but if you do that for a ton of quests then people go like wow there is really a lot of stuff going on where people remember and do things and this world is alive uh critical information is always the next step that's very important um supplemental info is seasoned all over the place and the player freedom is huge you can kill everyone you can go to the first town where are these guys Bam Bam Bam the guy's like I don't want to tell you bam shoot him take a thing off his body go to the next town hey who are all you guys cool go to the next town and it's it the story it it it makes sense because you all the information is there you go to the next town the guy and the the dinosaur Manny says well I know where those guys are but I need help and you're like yeah you need help you need a doctor boom you take it and I mean you can just you can kill everyone you don't you don't have to like you can just keep going and the story totally makes sense and you're just this Whirlwind of death and it worked very well uh so the fourth step is to get some perspective on this stuff um so this is your analysis where you've sort of looked at all these things and when you start to look at the choices you're giving are they satisfactory so outside of what you've

defined when you say like if a player were playing this are they going to like the choices that they're given and are are they going to like how they play out um look at the role playing versus the benefit if I'm playing a certain type of character do I get enough of a benefit that I don't feel like I'm being penalized for playing that sort of character and if there's a conflict you have to identify what that conflict is um look at your critical and optional narrative is the player getting everything that he or she needs to actually understand what the story is again this is separate from the optional stuff if the optional stuff they can skip that's fine players that want to dig they're going to dig you don't need to like force it in their face you don't need to bribe them to look at it that's fine then write your Pros no problem man super easy well the point is that once you've identified the sort of range of the stuff it's not that necessarily writing Pros is easy but that that Spectrum has been codified much more so when you start to make your choices about how to write it it's in the perspective of what the player is going to appreciate so writing Pros it's its own thing again but this is the setup

for making branching dialogues so in conclusion in conclusion I screw this up one more time um you are Choice Architects when you're building this stuff there's nothing neutral about the choices you're making you have to plan it uh RPG players want good choices and that's very subjective but the idea is if you're supporting a type of character and in a character range they want to feel like that's valid for them and the games that do this really well really make players love I mean they just love those games because they'll come up with C man I see the craziest character Concepts people come up with for New Vegas and a lot of it is in their head a lot of it is sort of self-narration about the sort of person they are and where they came from some of the craziest stuff um but if the choices support it then they feel really excited about it and they they really attach to it um designer and player goals can be in Conflict your narrative goals their desires as a player you have to resolve those conflicts or at least accept the consequences of that conflict you can't

resolve every single thing but you have to at least recognize it plan for it and accept it thank you very [Music] much I believe hey we have three microphones it's awesome hello so thank you so much for that great informative talk I had a lot of amazing techniques for for branching dialogues um I'm wondering about your your searching about players wanting good choices uh and I I totally believe you about most players but I it makes me think of players who look for hard choices because they're not validated right like players who want to play a pacifist front of a certain type of game because it's so hard and because it's unrewarding and I'm wondering would you would you ever make a game where there's a a set of choices that have no reward maybe until the very end and if you just

like went without reward you get like a single flower or something like that you know I I no I I think I think that's valid because what you're talking about is sort of um well there's a reward there um but you have to identify that reward is so people that play pacifist runs of because you can play Fallout New Vegas as a pacifist run um and you I think you have to actually kill anybody and uh people enjoy that and the reward is the fact that they didn't have to do that at all the thing is though inherently they know what's coming for them like they know that that's out there for example if we said hey you can do a pacifist run of this and they got halfway through and we're like psych you got to kill this guy they would be livid and pissed because we betrayed their expectations we did not forecast what was going to happen um there is actually an interesting thing where for example in Honest Hearts there's a part where you're trying to evacuate and you can play as a pacifist you go in there and you say like I don't want to fight any of the white legs these tribals that are just butchering everyone and um uh what

Daniel says is if you don't kill these guys they're going to kill a ton of the Sorrows this very innocent group of people and you can say like well too bad I'm not going to fight and you can do it but but the choice of the but the consequence of that is forecast ahead of time to them and I think that the as long as you sort of communicate to them this is going to be the reward this is you know it's sort of self forecasting the case of pacifism because that is ahead of time they know I can do this the whole game supports it I can get through it I don't have to do it that's that's fine so you don't have to give them like a system reward like you get a you know a special bonus if you were a pacifist I don't think you have to I mean I think if you want to that's fine again if you want to forecast that uh for a while we had challenges in New Vegas they they were too hard to track but for what I can't remember the technical reason but we had pacifist like sort of rewards for people that got to level 10 or level 20 without killing anyone so you can build that in um I think the important thing is one way or another you have to figure out what the player is actually going to enjoy because a lot of it is very subjective

um some people just like being a good person and like people responding to them very warmly in the world and so that can be a very good reward for them thanks yeah thank you that was a a great talk and um fantastic game as well I think I was just wondering um where you were presenting there with the example where you had the speech statistic and the notion that you can win a uh a conversation um and actually I found you know that was really effective in Fallout New Vegas because I could grow my character in that way because you know combat obviously is a gameplay system and you resolve it by winning and uh felt you could have that ability to express yourself by growing in combat or Express Yourself by growing in conversation so if you remove the ability to win in a conversation or to have a sense of having varying degrees of success within that system um how can you still able players to feel like they're growing in that part of their character or that that role um I think

it's a difficult question because uh when we say that the player wins in there there's not actually a game there like the game if if anything the game is the strategic game to put your points in your speech skill but there's no real way there's no real way to forecast that stuff um so if a player if a player's already put the points in it when they get to that point there is no gameplay there it's just combat you put put your points into your skill but then you still have to like you know move around and actually choose when to shoot and when to take cover when to use your stems so it's it is a very difficult problem and we've seen games that have tried to tackle this um Oblivion tried to game like sort of more make it more gamey uh and it I don't think a lot of people responded super positively to that um it's a difficult question I don't know what the exact right answer is I do feel that it doesn't have to be a skill that you put points into I think that dialogue can be thing that you feel like you navigate your way through and you build reputations I think there's a

lot more that we can do with indirect reactivity mechanics that make people feel like they're playing through that uh the important thing to me is that people stop and think and they make their choices and they're weighing the Tactical and strategic things speech skill is not in conversation is not a tactical Choice nor a strategic choice and that's what it comes down to for me and I I'm sorry I don't know where to go with that but I know that a lot of people commented on it and felt that it was kind of out of place MH cool thank you is this one one more three two two great shrugging uh can you talk a bit about the granularity of choices and and where you figure out where to fall on that Spectrum I think at the bottom of the spectrum of a game like BioShock 2 where you had like four choices over the course of the game they were all incredibly memorable to me they they felt they were very difficult choices to make and I remember every single one of them uh at the far end of the scale of course something like Fallout New Vegas where there constant choices uh and each one feels in a way a little less important because I mean

it's a design choice in a way but then there's something in the middle like Alpha Protocol uh where fewer choices but they're more meaningful uh what do you do with that scale and how do you sort of decide where to end up um I think I think a lot of it has to do with um again the range that you want to support the range of reactivity you want to support and the range of of expression that you want to support so um I think early on in BioShock you know I don't think the player under any illusions that they're going to be making a ton of choices um and so there's limited numbers of choices limited amount of reactivity uh in in Alpha Protocol it was I think actually in many ways Alpha Protocol is kind of the way to go as far as conversation mechanics go because Alpha Protocol is is every conversation feels like a scene it has a beginning and an end and it like flows and people are talking about things and it doesn't sort of Branch out into a a big info dump or anything like that and you feel like you're making meaningful meaning choices in there but I think ultimately it has to do with figuring out how much expression do you

want to actually support in the game what do you think players are going to want to do um if you think players are going to come up against these situations where they say I really wish that I could do something different in this situation then that's something you consider like well maybe we should have a choice here to determine how the player can go through it um but I think I think it's a stylistic thing on a game by game basis so I guess one more question at least hi um um you made a point like about info dumping in dialogue and also um your reference to one of the characters making reference to something that actually wasn't in the game um and I found my experience especially with Skyrim this didn't happen so much with Fallout cuz I think maybe the writing was better but um in Skyrim I felt the shallowness of kind of this big fleshed Out World at points because there were characters that I wanted to interact with and no more about that didn't even have that kind of um option to sort of engage further

with them in a sense or that just that simple option like who are you type of deal and I was wondering like do you think there are better techniques within dialogue to sort of provide that kind of closure or Bridge yes I I do so I I think that um the answers to things don't always lie in films but I think that we can take a lot of there's never Point well I shouldn't say never it's very rare there's a point in a film where someone says hey I'm this guy like every character comes in and I explain who I am and blah blah blah it's done in the context it's done in the context of the plot and so you learn about the character through the context of the struggle and that's one of the reasons again why I think that Alpha Protocol was did a really great job because every scene was about a thing like two people or more in Conflict they want something they're working towards it and in the course of that Revelation you learn about the character of that person

here's another thing um interesting conflict reveals tons about characters the sort of Jesus Hitler characters nobody finds those characters enjoyable but when you like a character like Sophie and Sophie's Choice that wrecked her whole character or even if you look at um I was talking to my friend Terry yesterday about uh James Bond about Casino Royale so in Casino Royale you kind of see young Daniel Craig um you see young James Bond as the sort of Reckless dude who gets shaped through these difficult things that he has to deal with these difficult choices that he has to make and um he seems I think a lot more interesting as a character because of those choices so when we when we instead of a character saying hey here's a quest if they become more emotionally engaged and involved in that and you become engaged with them I think that reveals a lot more interesting stuff about characters one more cool last nobody hooray we won everybody we won

thank you so again thank thank you all for coming and uh don't forget to fill out the evaluation thanks