Gamasutra: The Art & Business of Making Gamesspacer
View All     RSS
October 30, 2014
arrowPress Releases
October 30, 2014
PR Newswire
View All

If you enjoy reading this site, you might also want to check out these UBM Tech sites:

Future of Games: Emily Short Talks Building Interesting Gameplay Around Dialogue
Future of Games: Emily Short Talks Building Interesting Gameplay Around Dialogue
April 15, 2011 | By Christian Nutt

April 15, 2011 | By Christian Nutt
More: Console/PC, Design

At UCSC's Inventing the Future of Games symposium today, narrative designer, interactive fiction expert and writer Emily Short presented a short talk called "The Listening Player," where she discussed a game project in which dialogue forms the core of the gameplay.

Short recalled a brief conversation she had with Chris Crawford at GDC in which he critiqued her and others' interactive narrative games. "You're not giving the player enough of a chance to respond; the game is doing all the talking and the player is doing all of the listening."

Said Short, "I assume what he wanted me to do was give the player more of a chance to respond, but being slightly ornery I wanted to embrace that, and create a game that is all about letting the player listen."

Since she was at GDC when she had the idea, she immediately thought of having the player control an audience. The player would input a reaction on two axes -- bored vs interested and confused vs comprehending.

"What I wanted to explore with this is the experience that the speaker has as reflected through the audience," she said.

"I became aware in my years teaching college how my content was shaped by the way my students were reacting. Wouldn't it be interesting to explore through a game what it's like to listen to someone telling you something very personal, and what it's like to be that person?"

It "wasn't too hard" to put together a mechanic to offer feedback to a speaker, Short said, "but the real problem is just the question of gameplay. What is it that the player needs to be doing in the process of listening? How do I make something that feels engaging and challenging?"

In commercial games, players often skip dialogue, Short noted, and many games that use dialogue as core gameplay in the indie space use it for combat -- interesting, but not right for what she was attempting.

"What I wanted to do was more of an exploration model -- what is it I want to know more about? What do I want to understand? How do I seek interesting discoveries?"

The game would not just be about the goals, but about exploring "the terrain" of the speaker's presentation, she said. The player could learn "how the person who is speaking thinks about reality; how she responds to feedback."

As the player, you'd understand "your immediate options, but also distant things so you know where you want to go." The game would have to offer "intermediate rewards" to remain interesting.

"What is the fictional structure I can put onto this that will create the exploration I need to do?" Short wondered. She eventually devised the idea of a speech with an awkward intro -- the kind that starts in the manner of "this is a topic some people don't feel comfortable with."

The topic she came up with is werewolf safety. The in-game speaker has had a personal experience with werewolves, but will avoid revealing whether or not she's a werewolf if at all possible, using methods including unrelated conversational tangents, Short said. "This would give it a shape where there is some gameplay dynamic."

Short said there's still "a lot of thinking to be done about the gameplay aspects of conversation. It's very tempting to focus on the simulation aspect," but "there's also a great deal to be gained by backing off of that and thinking about the design issues, and how to make conversation gameplay, and not framed around other gameplay."

Related Jobs

DeNA Studios Canada
DeNA Studios Canada — Vancouver, British Columbia, Canada

Analytical Game Designer
Blizzard Entertainment
Blizzard Entertainment — Irvine, California, United States

Senior User Experience Designer, Irvine
Blizzard Entertainment
Blizzard Entertainment — Irvine, California, United States

Senior User Experience Designer, San Francisco
Avalanche Studios
Avalanche Studios — New York, New York, United States

UI Artist/Designer


Alfe Clemencio
profile image
Phoenix Wright is a great example of how dialogue as gameplay works.

Christopher Aaby
profile image
I have to disagree. The Ace Attourney games are completely linear, and are largely based on matching the game's expectations 1-to-1 to your actions. One interesting thing that could be said about it is that it really is all about listening - you spend 95% of the game reading, and very little time making decisions... but the reply-mechanic is not very transparent (like how a bottle of nail polish could mean "this character did it" - a thought-up example, but relevant), and often confusing.

Alfe Clemencio
profile image
I "dialogue as gameplay". I did not say anything about being non-linear.

Nick Harris
profile image
I've thought that the Mass Effect approach was all wrong as you could exhaustively traverse every path in the dialogue "tree" without any intrusive time pressure. What was needed was a way to zoom in on someone far off in a crowd and having targeted them initiate a conversation by hailing them with a press of the (A) button.

This gets around the old chestnut of running after an NPC to step in front of them in order to trigger them stopping your free movement as they start questioning you.

Once you've said "Hey!" with (A), you can terminate the conversation at any time with (B) for "Bye". During the conversation your options are limited to (Y) for "Yes" and (X) for "No" the rapidity of which conveys your degree of enthusiasm to the most recently suggested course of action. Your character could appear weak if let them appear to be too keen to accept the first proposal from the NPC without a modicum of persuasion.

Only the game knows what is possible, so it is pointless to get you to properly "talk" to it. Even you do not fully appreciate the situation your character is in and every NPC's secret affiliations and yet to be revealed hidden agenda. Quests and mission (sub-)objectives should only be added to your character's "to do list" by speaking to the right person at the right time and saying "No" to the first few things they try to fob you off with, as an unexpected negative response may lead the NPC to question your motives, perhaps offering you something harder (if they previously offered you an easy sounding mission), or conversely, something much easier if you rejected a challenging quest - hence, dynamic difficulty levels balance procedurally-generated thematically-coherent mission content.

Because this need not stop the action (your character can freely move around, change stance, go through doors, board vehicles ...only to be followed by the NPC you are talking to, unless they choose to terminate the conversation... throw grenades at oncoming enemies, even fire a weapon during a fire-fight), as these face buttons: (A), (B), (X), (Y) need to be used with the Right Bumper button (RB) temporarily held down to enter the combat mode, e.g:

LS - Left Stick (pushed all the way forward) Sprints

LT - Left Trigger throws a Grenade (rather than Zooming / Aiming when a weapon is equipped)

RS - Right Stick targets the body part of an enemy to Melee with the aid of an optional floating reticle

RT - Right Trigger is a context-sensitive Melee / Assassination / Human Shield when used with delicate grip

(A) - Ability (e.g. "slow-mo") / hold for Action (e.g. open door)

(B) - Use Health pack / hold to Give / or Revive incapacitated person

(X) - Reload / hold to cycle weapon Modes

(Y) - Swap weapons / hold to turn towards (hidden) objective marker

LB - Left Bumper is Jump

If this all seems a bit passive, this could be remedied by putting your character in command of a squad of largely self-sufficient, non-suicidal, AI team-mates. Their behaviour could be optionally directed through the use of the D-Pad - e.g. Look at some objective / enemy / intervening point of hard cover and push Up on the D-Pad to send your squad there directly, or Left / Right to flank it. They would only go if they felt it was safe, but would only be able to determine the threats they had seen (more could be marked by clicking the Left Stick in whilst looking at it). Bad leadership would lead to deteriorating morale and here the dialogue system would come into play requiring you to assert your command, but also to recognise tactical initiatives they may spontaneously come up with - better to have a dumb AI voice a proposal and be stopped, rather than watch them pay with their lives for their mistakes. Morale could be restored by asking them to regroup on your position (D-Pad Down) and if all else fails a misconceived assault could be turned into a retreat into the closest safe cover with a push of the Back button.

It wouldn't take much to take a multiplayer FPS and replace one team entirely with a mix of different AI and boost the ranks on your side with AI under the command of each cooperative player on your side.

Drama is nothing without jeopardy, but it is clumsy to repeatedly kill off the main protagonist every 30 sec. Ok, so they are not cannon fodder, but they do create a space of play with which you can experiment that lies between your character and the bleeding edge of the frontline. It is behind the lines that narratives can be allowed to develop dynamically, as the "band of brothers" reveals contextually-triggered pre-scripted parts of their histories under an emotional system that synthetically models their hopes and fears. After all, it is only away from the cacophony of the frontline that characters can hear themselves speak.

War creates stories of heroism, pathos and betrayal.

So why not simulations of war?

EDIT: Looking at this again, I now realise that it makes much better sense to require the Right Bumper button (RB) be held down to 'shift' into an alternate gamepad layout where the face buttons are used for passive dialogue and the D-Pad for active squad orders. This makes it a lot more conventional, with (X) for Reload, etc. when (RB) isn't held. I should have realised that the default mode was combat, not conversation - so apologies for that.

Justin Keverne
profile image
The first thing that comes to mind when reading the description of her talk is of a tilting table on which sits a marble. The marble represents the current topic of the speaker lecture, and the table define all the possible pre-scripted topics that exist for that lecture. As players tilt the table along each axis, "bored vs interested" or "confused vs comprehending" the topic of the lecture changes appropriately. A rudimentary AI could be implemented for the speaker which would attempt to steer the topic back on course, metaphorically distorting the shape of the table. Both players and the NPC are attempting to push the conversation in most desirable direction.

Rodolfo Rosini
profile image
Proper conversation is an AI-complete problem, however you can approximate a lot in a game by constraining the space of dialogue you are going to cover. I'm fairly excited about the shape of things to come.

Christopher Aaby
profile image
I second that. Just think of examples like the Sims - it has a pretty good and flexible dialogue, considering that not a single meaningful word is actually uttered. It doesn't have to be a problem of perfect natural language processing.

Bertrand Augereau
profile image
First, let's admit I really like Emily Short's interactive fiction (last one I played is Savoir-Faire, appeals to a french guy of course).

Lots of good innovation emerges from the if scene, but sadly the huge majority of game designers (and even writers) in the industry is too young (and/or too illiterate) and not interested in the text interface, therefore dismisses the sometimes brilliant content.

Yet there was a really well-done and solid dialog system in a mainstream game recently, and I'm thinking Alpha Protocol (hint: a good writing on top of good mechanics help, cheers MCA)

Patrick Dugan
profile image
I just think it's cool that now there are a lot of people, both on academic and commercial sides, interested in banging their heads against the same walls. Maybe if we get enough heads banging, the walls will finally come down.