Mirjam Palosaari Eladhari, Gotland University

Problem statement

How can a sandbox virtual world system for story construction with semiautonomous avatars be enhanced to cater for dramatic choices?

(Problem statement derived from MP Eladhari’s research)

A brief statement of a solution to this problem

The approach MPE is experimenting with to explore story construction in virtual worlds include the use of semiautonomous avatars. (The Mind Module)  The prototype game play draws upon the application of psychology in the Mind Module, where characters have personality, emotions, moods and emotional attachments to entities in the environment.  One of the intended drivers for game play is our constant struggle for self knowledge where we mirror ourselves in each other. The current prototype is sandboxy.

In order to enhance the possible experience in narrative terms, to cater for dramatic choices the following approaches might be interesting to explore:

• Life Path Systems
• Information revelation and withholding by autonomous agents

Expanded problem statement

1.   Why do it in the first place?

Six years ago I “took a break” from the game industry where I had been working as a lead game programmer for a short time period. I had, in the game we made, experienced a clash between two approaches; an object oriented approach to story logics where the casual dependencies where tied to objects and geographical locations, and a linear hypertextual tree structure that was put as an overlay. This led to false causal relationships, which resulted in solutions that were ugly from both a narrative perspective and from a systemic perspective. I spent almost a year tackling this, which led to a masters thesis called “Object Oriented Story Construction” and an AI paper called “Causal normalization”.

During this period I started to suspect that a possible approach for story construction in large persistent systems with a multitude of players could be to further utilize the intelligence of the player, i.e. to use the artificial stupidity of current state of the art to, instead of making NPCs that MMO players don’t care about so much in the first place, enhance the player character. I call these entities semi-autonomous avatars.

It might be possible to harness the power of human relationships and plotting in virtual environments if these are tied into the game mechanics to achieve drama where no characters are supporting parts, but instead main protagonists and antagonists from their own position in a relational network. Exploring this would take time. It came to my attention that there is a loophole for time and thinking in the form of doing a PhD. So I pursued that. 

2. What is there right now?

I started out mapping out my theoretical approach to story construction and narrative in games so I would be able to put it to rest.  (“Story construction and expressive agents in virtual game worlds”, 2004) At the time there was a vivid discussion in the game studies field about the tension between narrative and game play. To me this appeared to be dead end, and I wanted to build instead, and see where it would take me.

In order to see how to potentially use the intelligence of the human mind in junction to story construction I turned to the areas of psychology and neurology. I implemented an agent-based architecture for player characters in multiplayer games, the Mind Module. I built it as a spreading activation network, and it gives player characters personality, emotions, mood, and emotional attachments to other entities in the world.  Potentially the emotional attachments can be combined in relationships to create drivers for player action, while the personality would make sure that the experiences are highly individual. I thought a lot about how these emotional values would be expressed to the player: i.e. the individual emotional state, and built, together with a composer, a prototype that expressed the values in the form of music. This, after all, is the language of emotion. (“The Soundtrack of Your Mind: Mind Music - Adaptive Audio for Game Characters,” 2006)

In order to see whether the approach used in the Mind Module would be useful there needed to be a world for the modules to live in. During some years I incorporated it in three different projects with unsatisfactory results. The main problem areas where of two categories: Firstly, technological research projects are always under funded regarding implementation, a strong reason for that being that the success criteria is not measured by functioning shippable products but in publications. In two cases the development was cut just when we had platforms and frameworks in place so that we would be able to start doing interesting things.  Secondly, there was a design problem. I had been rushing, and not been thinking deep enough.

So I set about doing a full design, instead of just “plugging it in”.  It has the tentative crappy title “World of Minds”, and takes place at the institute of Pataphysics (IP). The staff at IP have locked themselves into so disparate and alienated world views that membrane between reality as we know it and projections of mind is ripped apart. The player characters enter to investigate and deal with the situation. Together they can use affective actions and mind magic to defeat negative manifestations. What they can do depend on their personality and their current mood. Player’s can affect each others moods through interaction in order to modify the action potential.
(“Semi-Autonomous Avatars in World of Minds - A Case Study of AI-based Game Design”, 2008)

What I have been struggling with design wise is how to create potential for dramatic choices. i.e. How can I make it into something more than emergent story construction in a sandbox system?

As far as I have gotten to this point is a system where player characters need to make choices of how they manage their time and effort. On one hand they would be able to sign social contract between each other. For example “romantic exclusivity”: This would give the benefit of a powerful mood boost in proximity to each other, and a permanent slighter boost, but would also require restriction in the use of affective actions toward other player characters.

Combinations of social contracts and number of social contracts would be a management issue with dramatic potential. I had thought to combine this with the pursuit of principles, where a similar time management challenge would meet the players. The pursuit of principles would be measured by what types of actions the player performs in relation to the world, rather than in relation to other player characters.

…But it doesn’t feel right.

So this was the problem I needed help with, and all the above is what I explained in the rogue group.

In short: How can I use the system of the mind module and redesign the WoM prototype to further support dramatic choices?

Expanded solution description

We were looking for approaches that can, combined with a sandbox virtual game world with semi-autonomous avatars, hold the potential for individually meaningful experiences for players. This includes being able feel that choices of the player character are meaningful and dramatic.

(Semi autonomous avatars are player characters that have a level of autonomy apart from the player, where the level of autonomy depends on the game design. See section diagram below.)

Approaches to explore:

1. Application of Life Path systems.
   What a life path is defined as depend on context. In paper RP it usually facts provided by players at the character creation stage that covers stages life leading up to the point in time where the playing is done. Game Masters often find it a useful tool for bringing in new players and help them in their early RP experiences.  In an MMO setting it can be argued, depending on perspective, that LPS do not exist at all, or that they are the whole game up till the level cap where the elders game begins. In the context of WoM an LPS that connect characters in the world to each other might be interesting to model, prototype and test.
   (Reference: BattleTech Solaris, MechWarrior 3rd edition.)

2. Balancing progression
   In the current design character progression occurs at the following axes:

• Social contracts – amount, nature and strength. (i.e. how many, what type, and how it has grown)
• Self knowledge – for each character trait (30) that is mapped out through choice of action abilities and max-energy and resistance is added.
• Titles tied to activities in the world are earned, and adds to max-energy and resistance. (compare to WoW Achievement system

Note: In the current design achievements in the world were tied to principles, which were intended to create dramatic choices, but might end up becoming more of a time management issue, emphasizing the already existing sandbox style.

3. Revelation and withholding of information
   An important tool in plotting and drama is the withholding of information. Consider for example the deaths of Romeo and Juliet. Also in sandbox systems such as Sims2 this is highly functional. For WoM a possible experiment could be to combine the system of social contracts with various levels of knowledge that can be retrieved from NPCs. If a social contract is broken a character might be able to find that out, depending on what relationship she in turn has with the NPC.
   

4. Story significance and celebrity (side note)
   Allen described a system where each character has a score from 1 to 10 in significance for the story. For example, Luke Skywalker would have 10, a storm trooper would have 1. In a scenario of conflict Luke can never be killed by a storm trooper.

In some RP oriented MUDs only very accomplished players are allowed to play characters that have a high significance in the back story of the virtual world. In a Final Fantasy MUD for example, the important part of Vivi would be entrusted to a player that can do justice to the role.

Given that virtual game worlds are our new “third places” (ref Oldenburg) we no longer gossip and keep track of the people in our village. Neither do we watch TV that much anymore, so celebrities from that realm are not that interesting. For MMO players it could be the case that it is the players on the same server that sparks the interest in the way the blacksmith's wife or Brad Pitt did before.

A system that could measure story significance on a server in terms of attention from other characters and drama could be interesting. I am picturing a subsystem that can be used by those players who enjoy role playing where social relationships and perspectives on interpersonal conflicts are tracked by consenting players (actors) and propagated to interested players (spectators). This would constitute a sub drama played out by those who want to in participate and express themselves while other players, who might be more interested in questing/raiding/PvP/whateverActivity could see it as an extra flavor to their world. However, I will not be able to try this out past modeling, since my current prototype will not be played by a full server population.

In summary, the approaches I intend to try for story construction in virtual worlds include the use of semiautonomous avatars with emotional attachments with game play derived from the application of psychology in the Mind Module, combined with life path systems, information revelation and withholding by autonomous agents where the main driver for the game play is our constant struggle for self knowledge where we mirror ourselves in each other.

All these approaches may not be possible to try in the same system. An inherent problem researching this is that existing general systems do not necessarily give context that in turn can give interesting results. Reversely, specialized systems may not provide us with solutions or answers that are general enough to be useful. Nevertheless doing this might be better than not doing it at all, and hopefully the voices coming from this project may have a pitch to add to the chorus of game development, game playing and game research. 

Items from the brainstorming lists that the group thought were worth reporting

Quote from Richard Garfield (Magic the gathering): "all multiplayer games basically the same - a player who is good at one is often good at games of the entire genre. "
Bartle's book, the sad general tiny mud derivatives.

Paper RP games that feature social interaction rules:

• Castle Falkenstein - 19th century alternate Europe romantic melodrama, fiendish spies, playing cards: combat, negotiate or persuade.
• The Dying Earth - based on series of novels of Jack Vance. Roguish characters, social persuasion system, various skills and approaches.
• Pendragon - Arthurian legend. Saxon knight, Celtic, etc. different cultural values: opposed pairs for each culture. To be good in 5 of 20 profiles.
• GURPs (copy of champions) - choose impairment (stuttering) to gain skill.
• Call of Cthulhu (sanity system)
• Vampire: nature and demeanor rules for character creation: similar to inner and outer mood. RP aid more than mechanics. Wraith: The Oblivion.

Betrayal of a relationship. Rating of how severe the relationship is. A rating of how prone an NPC would be to tell another player character about that betrayal.
Kingdom Hearts (inner demons materializing)
***
Categories of goals:

• elimination of certain manifestation
• certain spells
• reaching energy levels
• relationships of different kinds with other players
• can I induce certain moods in other players

-> Titles for the achievements.
Meeting someone from the same company in the same profession that could have influence over you, maybe less so if not in same company.
City of heroes, the exemplar and sidekick system - a way for players of different levels to team up effectively.
Dramatic choices.
Time management problem whether to pursue principles or careers, or relationships.
Potential driver of drama:
Lack of information (ref Romeo and Juliet)
How?
Haunting hidden relationships.
Social overlay, facebook friendslisty type.

Other reference material

section 9

next section

select a section:
1. Introduction  2. Speakers  3. Executive Summary  
4. Concept Sketches for Game Design: Best Practices for Prototyping within the Game Industry
5. Multiplayer Game Atoms
6. Game Designer as Artist
7. Please, Let Me Play the Game I Bought
8. The Gaming Date
9. Dramatic Choices
10. Game Designer's Bookshelf
11. Project Unity
12. Schedule & Sponsors