GDC 2001

Creating Emotional Involvement in Interactive Entertainment

Moderated by Mark Barrett

Attendance
Again this year attendance was good, representing a solid mix of development disciplines. More importantly, I failed to detect any disagreements running along the party lines of particular job descriptions – for example between artists and programmers, who are frequently cited as being at odds in the development process. I hope this is an indication that we are avoiding any stratification in our overarching design goals that might otherwise be driven by individual interests.

Key Points
I began each session of the roundtable (which met at 4 p.m. each of the three days of the main conference) by noting that we were not going to:

Support arguments with sales data.
Equate the phrase ‘emotional involvement’ with the word ‘storytelling’.
Monopolize the discussion or topics of discussion.

The prohibition against sales-based arguments came from the previous year’s roundtable, in which people tried to validate (or invalidate) theoretical arguments based on the success or failure of particular titles. The problem with this kind of argument is that it does not take into account the fact that a valid theoretical approach may have been botched in implementation.

The issue that I thought might cause more problems for the participants was the issue of separating emotional involvement as a goal from storytelling as a process, but simply pointing out that it was a trap to be avoided seemed to do the trick. While there were still discussions about the efficacy of user-defined stories versus author-defined stories, there seemed to be a much deeper awareness than last year that neither position has the upper hand.

Finally, my admonition about not monopolizing the conversation or the topics discussed was not so much a prohibition for those (like me) who have a desperately important opinion on every subject under the sun, but was instead intended to welcome any and all commentary from more reticent attendees. I believe this effort was successful, as the discussions involved a good number of participants across a variety of subjects.

As to what we were going to discuss, I proposed specific questions at the beginning of each day’s debate, which are noted below under the individual reports for each session. In general, the propositions put forth echoed last year’s roundtable, and centered on the idea that emotional involvement is not being treated as a issue central to the long-term health of the interactive industry.

[Note: for tracking demographics this year I proposed five categories and asked people to pick the one that best described their principal role in the industry. Because I did this at the beginning of the session, before stragglers slipped in the door, the total number of people who were polled is less than the total number who attended each day.]

Day One
Total attendance: 45+

Artists (graphics, sound and writing): 6
Programmers: 5
Designers: 16
Producers: 2
Other: 3

We kicked off the discussions on day one with the following questions:

* What is emotional involvement, and where does it come from?
* What techniques can we use to promote emotional involvement?
* What experiences have attendees had that they themselves found moving?

Proving the dearth of reliable techniques that produce emotional involvement directly from interactivity, most of the examples cited and points made on the first day referenced either passive mediums (particularly film), or passive techniques which had been adapted to interactive works. Although many of the points made were good ones, it was clear that creating reliable and sustainable emotional involvement from interactivity itself is still a long way off.

In short order the question of how to create emotional involvement from interactivity devolved (as it often does) into a period of confusion, as people came to realize that the very definitions we use to define our arguments are themselves unformed and incomplete. Words like ‘narrative’ and ‘story’, as well as ‘interactivity’, turned out to have disparate and conflicting meanings for different participants, making a consensus all the more difficult to reach.

During several failed attempts to push through these lexical obstacles, mention was made of the idea of ‘changing’ or ‘influencing’ the game world during play. Pointing out that this idea had come up more than once, I made the observation that interactivity as a process involves choices that determine outcomes, as opposed to choices which simply reveal pre-designed outcomes. I also observed that interactivity is the one thing which differentiates our medium from all others, which suggests that finding ways to create emotional involvement through interactivity should be our goal.

While there was no debate about these observations, and the definition of interactivity which I proposed did help us to move on, the conversation again quickly came back to effects that were achievable primarily from passive techniques. An interesting subtext began to develop, however, when participants began to cite different emotional experiences as indicators of what they themselves we’re striving for, regardless of the theoretical foundations of those experiences.

The question then proposed was, are we all looking for the same kind of emotional involvement? Should we be? Or are there going to be different audience (market) demands within the framework of what the medium can do? It was quickly agreed that we not only don’t want to limit ourselves to one market or genre or approach, but that there also seems to be no evidence that we need to adopt such a narrow focus in order to be successful.

A word which surfaced several times during the session was ‘immersion’, which was at times likened to an emotional depth of experience, and at other times to a kind of imaginative involvement. Again, after the same point came up several times, I made the observation that in passive mediums the equivalent of immersion is suspension of disbelief, which describes how an audience becomes emotionally and imaginatively involved in a work to the exclusion of the mechanism of delivery.

While it was generally agreed that emotional involvement in interactive works also requires suspension of disbelief, there was recognition that the medium of interactivity is having a hard time creating and sustaining the effect. The question was put as to whether this was a problem inherent in the medium, or in our current approach to the medium, but no strong conclusion one way or the other was reached. One thing that was generally agreed upon was that the industry’s current emphasis of generating emotional involvement from non-interactive techniques was not going to help us solve the problem.

Because the majority of the session involved defining the problems we face in trying to create emotional involvement, I said at the end of the session that we would simply pick up where we were leaving off if anyone wanted to return the next day. I also said that the focus would move from analyzing what was working in current games (where the emphasis is on techniques borrowed from passive mediums), to how we can improve the effects we’re getting from truly interactive experiences.

Day Two
Total attendance: 45+

Artists (graphics, sound and writing): 6
Programmers: 5
Designers: 15
Producers: 5
Other: 4

We kicked off day two by really emphasizing that emotional involvement is not a synonym for storytelling. Specifically, emotional involvement is an end, where storytelling is simply one process that is used to achieve that end. The question of the day, was: How can we use interactivity as a process to reach the same goal of emotional involvement?

As the discussion got underway, several people made reference to games like The Sims, and to the idea of ‘investing’ in a game as an agent of emotional involvement. Looking at the issue more closely, it was generally agreed that interactivity demands a rules set in order to support the kind of investment which can promote emotional involvement through interactivity. Although not stated specifically, it seemed understood that these rules sets needed to have sufficient complexity in order to sustain interest.

One way in which this unstated belief was manifest was revealed when the related idea of discovery was mentioned. Whether as the result of emergent behavior, investment, or authorial control, the revelation of new events, levels of complexity, or simple outcomes was deemed to be capable of considerable emotional power. At this point I made the additional observation that discovery as a technique is important in passive mediums as well, which might mean that there are other basic passive techniques for creating and sustaining emotional involvement that we can use in interactive works.

The discussion took an interesting turn when a personal anecdote broached the idea of deriving emotional involvement from the relationship between save games and player-character death. Specifically, there was a question posed as to the difference between the emotional tension in an RPG in which you could save at any time, and the tension in an RPG in which you had to get your avatar safely home before you could save your progress. While it was agreed that the latter case would certainly increase the player’s emotional involvement, there was considerable concern that the tension might not be so much from the play experience, as from the fear that the game’s design itself was going to negatively impact the player.

No conclusion was reached on this issue, but it did start an interesting secondary thread regarding the issue of death, and whether player death needed to be or even should be the way in which we motivate emotional involvement. Is it necessary to ‘kill’ the player, or other characters, in order to drive emotional involvement? If it is, is that a good thing, or should we be concerned about it? If it isn’t necessary, what else can we do? Along these lines there was the unspoken point that although death seems to play a large part in the design of interactive works, we don’t actually seem to be making death more ‘real’ in our supposedly more ‘real’ medium.

Toward the end of the session, as a summation of the first two days of discussion, and in anticipation of the third day, I drew a simple diagram and chart as follows:

I first made the observation that interactivity needs to be seen as a separate process from game or story, and that we need to keep looking at ways for interactivity to provide emotional involvement before we tack on victory conditions (games) or characters (stories).

I then made the observation that the relationship between games and stories in interactive works is very unstable, and often seems to erode many of the things we’re trying to do. As an example I said that I had played two very good games that seemed to blend story and game quite well for the most part, then at the end the gameplay became so difficult that it became impossible to suspend disbelief in the story. When I asked people to identify the games, Half-life was identified immediately, but I had to name Baldulr’s Gate.

After the roundtable broke for the day an interesting observation was made to me by one of the attendees, who said that he had been very surprised that adventure games as a genre were never mentioned, and that the word ‘story’ was only mentioned once.

Day Three
Total attendance: 40+

(As the result of a raging head cold, I forgot to ask for the demographic information on the final day. Recognition of a number of the people in attendance leads me to believe that the attendees roughly mirrored the ratios from previous days.)

I began the final session by redrawing the diagram and chart from the previous day (see above), and posing the following question: How do we keep gameplay and storytelling from competing for the user’s attention? The examples I cited – Half-life and Baldur’s Gate – were examples where I felt the game designers ramped up the difficulty of the game to the detriment of my established engagement with the story.

An early point, which hadn’t been raised in the previous two days, was that music could serve not only the aims of both story and game, but that it could serve as a bridge between the two as well. Moreover, because music itself can be interactive, responding to the choices the player makes, it stands a very good chance of deepening the user’s emotional involvement as choices are made.

The next point we focused on involved point of view, and the question of whether first- or third-person perspectives inherently offer more or less opportunity for emotional involvement. Those who preferred the first-person point of view repeatedly mentioned a disconnect that they felt whenever an on-screen, third-person character responded to an event in a way that was at odds with what they themselves were feeling. Those who preferred third-person said they liked the fact that they could witness more engaging and emotionally compelling dynamics between characters. After some discussion it was agreed that these differences may have more to do with audience preference than any specific inherent superiority, and that this may mean there are two distinct markets which can be defined by a preferred point of view for the main (player) character.

During the point of view discussion, several examples were mentioned of games that failed to hold the player’s emotional interest. Upon further analysis, however, it became an open question as to whether the games had failed because they had tried to do something impossible, of because they were simply badly made. Looking back on the other two days of discussion I noted that this is a common problem that further reduces our ability to hone in on effective techniques for deepening emotional involvement: we often don’t know if something is inherently unworkable, or if we’re just not doing it right.

Investment again reared its head as something which bridges the gap between story and game. In a game, the player invests in their choices, where in a story the audience invests (through sympathy and empathy) in the characters. The Sims was again cited as a game in which there seemed to be a powerful effect created by the intersection of these two kinds of investment. Time was also noted as a key element of both, indicating that deep emotional involvement is something that must be developed and nurtured as the user interacts with the product.

Toward the end of the session one participant said that they thought the designers of Half-life were right to have ramped up the difficulty at the end of the game. It was that person’s feeling that this was the right thing to do for the gameplay, even if it hurt the story. My response was that I couldn’t say for sure which approach was right, but the mere fact that there was a disconnect between the story and game made me wonder if something else shouldn’t have been done. It was the fact that the two diverged which concerned me, not whether one should have been forced to follow the other.

Other Points of Note
One notable discussion missing was any question whether 3D was going to be the dominant form of graphics. I cannot recall a single instance in which cell-based animation was mentioned, and only one instance in which FMV was discussed, and then only in the negative.

There were also no us-against-them debates about solo play vs. multi-player play. There seemed to be a recognition that solo and multiplayer gaming are not competitors, and that each may provide avenues of growth for the other.

One other interesting omission was the ongoing debate about women and games. Despite the fact that there were a number of sessions at the GDC concerning ‘girl games’, I can only recall one or two instances during the three days of the roundtable in which there was any gender-specific commentary, despite the common assumption that what ‘girls’ want is a more social and emotionally-driven experience.

Conclusion
The broad aim I had for the roundtable was simply to help people come to terms with the various essential elements that make up our medium, and with how those elements might be used more effectively. The fact that we were all able to talk about interactivity, story and game as disparate but not necessarily antagonistic parts of an ideal whole was a very comforting sign for me personally.

If any one thing was generally agreed upon it is that we’re getting to these discussions far too late in the development process, and that’s hurting our ability to create emotionally engaging products. The only way to increase emotional involvement is to design for it, but right now we’re not really thinking that way as an industry.