Press Y to Cry: Generating Emotions in Videogame Narrative

[Music] hello and welcome to press y to cry generating emotions in video game narrative my name is andrew walsh a writer director and narrative designer who's worked in tv theater film animation and comics i've also got 85 games credits to my name including prince of persia fable legends the division ii watchdogs legion and i'm currently working as a senior writer with grill games as well as a lead writer on an nda project i'm also involved with the writers guild of great britain's video games committee and i hope you can come along to some of the events that we organise so today we've got half an hour to look at motions in video games i've got a lot to cover so i will be moving quickly at the aim of this talk is to give you a set of toolkits and uh to analyze emotion and games and see how techniques from other media need to be adapted to fit into interaction please note that i said toolkit and analyze i'm not giving you a strict rigid process what i want to do is give

you a set of things that you can use at any point in the process to analyze the emotion that you find in games okay so let's get started how do they come to choose this subject and title well way back in the midst of time ea published this as an advert it was intended to start a debate about the creative goals for the industry and to set down a challenge could you make video games that made players cry by this stage the medium was already decades old and yet while music books films tv and theater knew how to create all sorts of emotional changing people to some it appeared that this new immediate immersive medium wasn't able to explore the same range of emotional responses despite it being immersive why well two sets of very clever people tried to solve this problem but humanity being humanity they each tried to solve it from opposite ends often without talking to each other and sometimes in direct conflict

on one side of this divide were the programmer designers who are forging the code verbs and models that modern games rely on today they were defining interactivity but many of these individuals saw emotion and story as problematic games were about verbs and interactivity emotion and story where it was to be found should and could only be emergent and even though many leading luminaries on this side of the divide continue to talk about emotion and immersion the question can a computer make you cry for many more from a challenge to a statement the game can't make you cry stories don't need game games don't need stories was to cry all of which is not sure you don't need a linear narrative in a game but all games are emotional triumph loss joy frustration relations we define the games that we play by the emotions that make it that they make us feel mean meaning that emotion itself is

interactive in fact when games fail to create an emotional response in the player that's because they've ceased to immerse or challenge them and in that case they've ceased to be a game yet despite that that many people on the designing a programming side rejected or ignored emote emotion in fact things got so bad that one chap had to go away and write a book reminding people that games were supposed to be you know fun now remember there were two sides involved here and on the other side of the divide were writers and storytellers after all reasoned many an executive if we need to tell stories and explore emotion we should go to the experts in doing this we'll get some screenwriters and novelists and they will bring a whole understanding and toolkit for emotions straight into our industry simple right well the problem is that many of those talented individuals and the execs who employed them brought problems as well as solutions

just as many early filmmakers had simply set up their cameras and film theater plays people seemed to think that they could make games in just the same way that they made films they thought that players should come to them and find their meaning just as audiences did in cinema they were sure that if only the people making games fitted the interactivity in a round or to the side of the incredible filmic experiences they were making then the door to emotion in games would be opened and it's not like either of the sides were completely wrong no one had come up with the solutions at that time and they were trying to build a whole new industry and it made sense to embrace emergence and interactivity as that's what made the medium different but it also made sense as a medium that games should explore and generate a wider range of emotions than they were the two sides needed to solve stop solving the problem from the end and started from the middle they needed to focus on the player now the legacy of this divide continues

to today you will still hear people that were proposing the use of heroes journey in games you will still people to hear people telling you that the only true game is emergent whereas actually the player is often sitting in the middle to get there let me ask a strange and difficult question who is feeling the emotion in your game and it sounds obvious i just said the player right well yeah but how often have you heard developers talking about war as hell and how they want to put the player right into the realities of combat but is that really what they want to expose the player to some of the worst horrors of humanity to tear them apart with a set of emotions that cause ptsd and an increased chance of suicide what's the question really can games tear your soul apart in the name of entertainment at the heart of the problem is a confusion between the emotions we're exploring in the game and the emotions we actually want the player to feel and when we fail to align the emotional

content of the game with the emotional reaction of the player the player will be left either feeling confused abused angry or feeling nothing at all the dreaded ludo narrative dissonance will eject them from the game so how do we fix that well for each game that we make for each moment in the game and each emotion that we're exploring we need to ask if it is a simulation a representation or an abstraction for simulation the emotion in the game matches the one that we seek to trigger directly in the player here we want the player to feel exactly what's in the game for a representation the emotion in the game doesn't match the one that we see to trigger in the player it's triggered indirectly or both here we put the uh we put some distance between the emotion in the game and the player's emotion and to bridge this gap the player will need to interpret what we've given them abstraction

in an abstraction we present the emotions only as a concept we don't intend the player to have an emotional reaction at this point okay let's look at each of these in turn simulation in this vr simulation game the player is asked to walk over a plank that's suspended over a big drop here the emotional reaction the developer seeks to create calls on the player's own natural response to heights the game intends to generate fear directly in the player in a formula one driving simulation the intention is to create the same emotional experience for the player as if they were driving the real car exactly what sid mayo was asking for but remember though when we're talking about emotions we're not necessarily simulating the environment or the subject pong is not a tennis simulator there's no tennis court and the rules are different pong is a skill game what's important here is that the emotions the game intends to create success win loss are the ones that the

game intends to generate directly in the player the emotions match the mechanics such emotional simulations can be so powerful that they're being used to change how people behave in the real world some companies are putting people into virtual situations to help them recognize racial or gender bias in themselves when interviewing doctors are being trained how to tell patients bad news in each case the participants enter the scenario as themselves and are then asked to respond to it directly the emotions they're intended to feel are the ones that they would be feeling if they were in the real situation but when it comes to representation the experience is no longer direct the game places something between the player and the emotion in the game an obvious example of that would be a character if we were to take the same vr training in the same doctor but we were able to ask the doctor to play a character this time then the scenario would switch from being a simulation to a representation

same doctor same scenario same emotional content but the doctor wouldn't be playing and feeling as themselves their emotion is no longer direct we've changed the way that we've asked the participants to react to it to change the experience similarly when the plant game puts the player in a place where they're asked to feel the terror of crossing over that divide when we put a character in the place it's no longer the player who's in virtual danger it's the character here the game doesn't ask the player to feel the same emotions as their character instead of terror the game is asking us to feel thrill or excitement or to feel scared for the character rather than from ourselves so we're putting a distance here between the players own feelings and those in a game why would we want to do that well for a start interpretation isn't necessarily bad interpretation is itself an active verb it's a way to make emotion interactive

another pulse with representation is it can allow us to present the player with a wider range of experiences than we can in straight simulation we don't want to expose players to the true horror of war but we can represent it we can use a character as a barrier to this negative emotion this can give the player an emotional distance to explore such difficult subjects using this technique the character can be put through all sorts of potentially negative emotional states such as loss rage grief hate neck revenge without asking the player to experience those those emotions themselves directly as the character feels lost the player can feel sorry for them they can connect to the character but without feeling exactly the same emotion that's in the game and it's not just emotion that representation can filter just as pong distills tennis into a simpler interactive and emotional experience we can disrupt because we can deconstruct other scenarios into their emotional components to find a way that suits our

game or the emotions that we want to explore when we portray love in a game for instance we're generally not presenting the player with the whole experience just as we do with games with war in a lot of dating games we don't deal with the choice of what to wear the stress of another person taking 20 minutes to text back generally games seek to shorten and focus the experience the design wants to find the active verbs not the moments of sitting and waiting for the text but in doing so we are taking the player away from that simulation we're taking them out of the real experience and that requires them to interpret and believe in a shorter time than it takes in a real day to cover how did you get here tonight fancy your drink oh my god you're into parrot topping two shepherd in the game is already asking you to bang x for some trans stella interspecies frolicking they're getting shortcuts that experience to give the player the representation and with it comes the

risk that we eject them out of that experience is that what happens well anyone who's been stuck on a table next to someone explaining their complex physical gymnastics they'd like to perform with garrus while the person next to them keeps shouting but he's a fascist we'll tell you that representation can indeed elicit very strong emotional responses in players we have to be aware of the change and build around it and take it into account but we can still create emotion abstraction though is where the emotion is represented only as a concept we don't expect an emotional response from the player here a good example of that would be player stats when a stat changes there's not an expectation that the player will then suddenly feel an increase of 10 affection trust or suspicion in themselves they may experience a sense of achievement but the stats not there to provoke a direct emotional response okay so simulation representation and abstraction can help you bridge the divide between between emergence and

authored narrative it defines how you're aligning the content in the game and the player's emotional response to it please note too that this isn't a set of three distinct options it's a spectrum each decision we make about our game can move the player closer to simulation or to abstraction so we can fine-tune the player's relationship with the game and emotions as we go note too that many games contain a mix of all three elements sometimes even in the same moment role-playing games for example are generally representations that find moments to move the player closer to simulation whilst at the same time defining elements of the character's emotional makeup through abstract stats all right so we've talked about who is feeling let's discuss what they're supposed to be feeling the emotional content of an experience is often called its theme to some people or for some projects theme can come quickly and naturally for others it's hard to distill the whole experience down to a few words

so here's a way that we can look at themes for games you can have a simple theme or you can have a complex theme a simple theme is just a topic and that topic can be a subject or an emotion for more complex themes we take the topic and select a lens through which we'll examine it essentially we take the topic and we ask questions about it now when we're asking those questions if we're decreasing our own project we often start with d what do i want to say then you can find the best mechanics that to explore it but chances are if you're hired to work on someone else's game that a lot of the mechanics art and even story sections might pre-exist your arrival so in those cases you need to look at what exists and find a theme that suits them and a good place to start as my role role i've shown you is with genre so here we are back with the plant game again genre is the type of game that you're making and it can be great you can make great games with simple emotions and simple themes all that this game asks for is the simple theme of fear

it works by delivering that emotion directly to the player so the chances are if you try and use crossing the plank as a way to examine the fall of western civilization you're probably going to be overdoing it and adding such complexities can actually push the player out and away from the emotions that you want them to feel where you want to remember where you want the player to enter an emotional state swiftly or die or feel it directly try to minimize interpretation and that means keeping the theme simple other games like pong or fifa or formula one simulators set out challenges of skill instead of focusing on a single emotional response but the player is still responding directly in a simulation so a simple simple theme of challenge can work here the problem is that whilst these are really great games they're all quite similar their verb sets are all about having about timing and reaction as a result they only explore a narrow band

of emotions this is where many designers got stuck at the time of the great divide and it meant that the range of games produced only appealed to certain players to widen the pool of players they needed to widen the pool of verbs and some verbs don't fit simulations as well kik run drive can work quickly in simulations because they don't need interpretation but you can't talk romance explore interpret or negotiate to win on a button press not without understanding the wider context in which that action is set and context means a more complex theme a lens takes the topic and asks questions about them those questions can give our gameplay verbs context take the love as a topic you can look at it in all sorts of ways are we saying that love is good or bad are we saying it's good or bad for the character things gets us to a theme gets us to ask important questions why does the character hate love what situations will provoke the character to question their

feelings and what role does the player take what do we want them to feel about the theme these are the very questions that the character and the story need to ask about the theme as we progress through the experience is where the emotions come from now if you find short uh descriptions like this don't actually convey the emotion you can always write them out as a sentence and then bill out as mind maps mood boards whatever works for you if you're stuck for theme you can start by with having the list of questions and then boil it down to get to here just make sure that when you get there the central theme is clear and concise that way it's easier to communicate to yourself to the rest of the team and the player and don't forget the player or the game verbs just because the game can allow a more complex theme doesn't mean you can log in anything that you want without consequence the theme still has to be one that can investigate with the genre setting and verbs that you've been given or that you've given yourself aligning the verbs in the theme is the

biggest step that you can take to align the game's emotional content to the player's emotional reaction now does that mean that we can't challenge the genre absolutely not spec ops the line takes war as the topic for its theme that meets the player's genre expectations the game also embraces a lot through the expected gameplay tropes from the the gameplay genre that also meets the player's genre expansion and expectations for a shooter in doing these two things the game produces a satisfying experience that keeps the player engaged and immersed so it's a good shooter so what's the fuss about well having brought the player in spec ops then turns the lens on the topic of war and begins to ask questions about the mental impacts that conflict can have on the soldiers that fight it and in particular the character that we're playing interestingly one of the ways that the game examines war's negative mental impact in the character is by moving the

player between simulation and representation in changing the player's distance from the character it simulates the way that the character is mentally separating themselves from the actions in the game now by inviting the player into what appears to be a recognizable situation it's a shooter then aligning the player with character the game's theme is able to contradict the gameplay experience the player is expecting to have in fact it tells them spoiler that everything they've been doing is wrong and it manages to do that while keeping the player in the experience now an important way that spec ops achieved this is it did it over a matter of time game designers design for immersion creating immersion is a balancing act increase the complexity of a game too fast over time and it can become frustrating for the player increase it too slowly and it can become boring both risk breaking immersion and ejecting the player out of the game

and it's the same for emotional immersion introduce new context emotions or information too quickly or too slowly and the player can be dropped out of the experience the good news is it also means that new context emotions or information can be used by game designers as one of the things to keep the player engaged within the game over time nine hours in you're shooting guys again but now you're doing it to rescue your buddy okay so we've talked about what we're saying but how do we communicate emotion in the game let's go look at some of those verbs that we've been talking about we can tell emotion we can show emotion and we can do emotion tell okay what's this guy feeling no idea oh so now we know that the character is happy telling is a quick and direct way that avoids the need for interpretation when we tell you're simply informing the

player what the character is feeling as fact but this can be emotionally unsatisfying without context the player is likely to take the information and not feel a response we're close in this form to an abstraction we can show so here we can show that the character is happy in showing we're asking the player to interpret so there is a risk that they might get it wrong hopefully not with smiley but on the positive side we are making the player active and we're also asking them to respond as they would on seeing a smile in real life that means that showing is closer to a simulation it also means that showing is more likely to generate an emotional response in a player than telling it's a stronger method and we can do emotions can we well this can work but even though we've added interactivity this is really just telling with a button the problem with press y or air to cry

or x to be happy is it assumes emotions are turned off and on in both the player and the character although by adding into activity we change something into something greater just by making it a doing action okay it does shift us on by involving the player a bit but we're still telling the player so let's write this as telling plus note also that this is not just representing how the character feels it's representing a change in the character's emotion and that's because interaction is better at action than passivity this is one way that games are different from films and books it's not that just that we can do things it's that not doing things can be confusing or difficult to represent and that means when we talk about interactivity we often have to combine tell show and do to communicate certain emotions and story points in the strongest way possible we can tell or show the context so that we can do the moment of change

okay so if this is only telling plus how can we improve do to match the things that i've been talking about well remember our theme we need a lens we need context by shifting to pet the dog we've created an emotional context it's exactly the same mechanic press x the same result same representation but because we've asked the player to interpret not just an emotion but an emotional situation the player is more likely to respond with an emotional response they're more likely to feel we're closer to a simulation and one that's been proven to work rather well but can we do this with more complex emotions than happy well the more complex the emotion the sillier it can feel to communicate it by tell and the harder it can be for the player to just read the emotion from show is this character morally conflicted or are they constipated here is somewhere where perhaps do can

come into its own what we can do is give the player a moral dilemma pet the dog or kick the dog such situations can create moral conflict not just for the character but also for the player well they may but remember we've added another layer of complexity and that means there's an increased chance the player won't feel what's intended part of the issue here is that the choice is binary some people would never kick the dog meaning that this was never going to be a choice so they never are going to feel morally conflicted at least not about the choice they might need might now feel conflicted about the character what an for even considering kicking the dog other people will kick the dog because they find it funny even if the character doesn't and in both situations we divorce the player from the character the situation and the emotion the same can be true if we had an outcome pet the dog for a 10 increase in trust kick the dog for a 10 decrease in trust here we need to ask ourselves is the

player now making a moral decision or a game decision in each case a big part of the problem is the fact that this choice is happening in a vacuum to really make this count in a linear case or even in an emergent case because i know this feels like it's linear it's also true in emerging cases to really make this count to align the player with the theme character and emotion they need thematic context let's imagine if the game's theme is one of the ones that we discussed earlier love is hate here the character is faced with a loving action and a hateful one and so when the player chooses they are using that choice to interrogate the character's feelings and if we communicate to our theme the players should understand why the character is facing this choice and what the emotional response is from the player at what from the original response of the character are after the player has made the choice context is vital to bring us into this because now we're attached to the character and can understand why we're making the the choice it

allows us to attach more emotion to the situation so that brings us to the question why does the player feel emotion well in simulations this should be obvious as we've said the player is responding to the emotion in the mechanics there's always a chance that they won't understand but as the mechanics are the context the player should really know what to feel and why they're feeling it i drove fast i'm excited for representation though we need context and that context can come from story that's not to say that every story needs to be linear or even authored but it does need context and when we think of story in games it's often helpful to think about setting up a series of emotional boxes each box contains an emotion and a set of questions that are in a question about it that relates to your theme box emotion question for each books you need to think about the context that the player has when they approach it what will the genre have told them what elements of the game

will have they experienced will the player have the context to interpret the emotion in each box just as they did with the moral question over the dog where the answer is no or unclear you can think of each box as requiring its own mini three-act structure what i mean by that is an introduction which sets the context the middle which is emotion the emotion and how it's split between tell do and show as a result simulation representation or abstraction and at the end some sort of emotional payoff or state change for both the player and the character at the end now that can be as simple as i have the ball i kick the ball i score a goal yay or it can be more complex so let's look at an emotional moment in the game that's all about simulation there's toll through representation and see how it deals with one of its emotional boxes or in this case a cube in the first portal game the player enters a chamber and encounters the

weighted companion cube in essence this is the same as the functional storage cubes they've encountered elsewhere only this one has a heart drawn on it the player is told to pick it up and to carry it through the level as the player encounters the cube and continues through the level the new emotional content comes from context comes from a handful of lines from the antagonist glados the cube is not a threat but you're not to care for it now up to this point the player's only contact with the world with any other being has been with glados by this stage the player is beginning to doubt her our suspicion gives context to what we're doing so the telling here becomes more than that we're suspicious so we doubt so we interpret we interrogate the fact that the things that we're told are contradictory also invites us to find our own truth in it deepening the telling when the player reaches the end of the level the player is asked to euthanize the cube essentially to press x to kick

the dock but it's not a dock it's a cube a box that says and does nothing yet player after player reached the incinerator and looked for a way to cheat the test so that they didn't have to destroy the cube only to realize there is no choice they had to drop it into the pit what the team does brilliantly in the space of one level is make you care for a cube in fact most of the dialogue in this level is written after this moment with glados insisting that you have to destroy the queue the team had constructed something where they knew they could predict the player's emotional reactions to the character's situation they knew that they were going to have to cope with the fact that the player hadn't dropped the box in there now to achieve this portal ticked all the boxes we discussed earlier just like spec ops portal embraces the gameplay genre allowing the player to become immersed in the mechanics so they can add complexity elsewhere portal builds its theme from the game's abstract repeated gameplay what would otherwise have been a series of excellent but disconnected puzzles

becomes something deeper as the player completes abstract task after update abstract tasks they begin to wonder who they are what their role is in this world and whether they have any power over it the team created a story of free will the cube like the player is an innocent trapped in this bizarre environment and we're used to being bossed around by gladys we now want to reject what we're being told so when we reach the end when we reach the incinerator we're identifying with the cube the companion cube in this world is the same as the player and this pushes the game from abstract mechanics through representation and into simulation the result being people developed feelings for a cube can a computer make you cry yes it can so to sum up when you're examining emotion in games there are four main things that are useful to think about who is feeling the emotion in your game simulation representation abstraction

what emotion is the player supposed to feel theme genre how do we communicate emotion in the game tell show do and why does the player feel emotion context structure by looking at emotion as interactive as something to interpret you can go forth and help the player to fall in love cry hope dream and explore and another thousand different emotions all while trying all while keeping them in the game good luck thank you very much for coming along today i hope that's very useful for you have fun and be emotional out there