Here we are at day two of the Serious Games Summit DC 2005. I'll be live blogging to the best of my ability today. For my coverage from yesterday, click here.

There were many concurrent sessions both days, and I had a talk both today and yesterday, du ring which I could neither blog my own session nor cover others, so the content here represents a segment of the total sessions covered at the event.

Keynote: Riding the Cutting Edge of Distributed Intelligence (Dr. Dave Warner)
Roundtable: Beyond Q&A: Assessment Methods for the Next Generation of Serious Games (David Michael, Sande Chen)
Lecture: Waging Peace Through Chaos (Nigel Papworth, Håkan Wallin)
Lecture: Serious Games for Religion and Religious Institutions (Les Perdew)

Riding the Cutting Edge of Distributed Intelligence
Dr. Dave Warner, Director, Institute for Interventional Informatics

The body is part of the interface to the mind. Information flow from the outside world to the inside world . All realities are virtual, a subject Warner called Neuro-Cosmology... the consciously experienced perceptual state space. Pharmacology isn't required!

Warner named a concept called Grok-It science: in 198"something or other" he started working on computer models for fractals. The computer models accelerate our capacity to understand things. Warner related chaos to neurology and neuroscience in medical school, looking at brain activity used to "take squiggly lines and make spatial-temporal flows." Warner showed topographic brain maps that show a clinician how the brain processes work though computer models and graphics. Different brain responses had different patterns, which leads to new diagnostic models. Warner articulated the idea of an "admission accelerant" that would allow hospitals to make better hospital admission decisions.

In the late 80s, the use of interfaces like data gloves provided new data input procedures, showing that the glove was more accurate than the doctors. Motor activity in surgical procedures was now possible in real time using these devices as feedback. In rehab, the glove was used for real-time graphical interaction after traumatic injuries. Similar devices also helped increase the expressive capacities of patients coming out of comas, in paralysis, and so forth. Quadroplegics could use electrodes to interact with the outside world. In one case, Warner plugged a child's electrodes directly into an NES, encouraging him to do his therapeutic exercises. However, Warner's team had to figure out how to map computer controls onto the unique and individual needs of these patients.

Then, the Marines came (accompanied by Darth Vader noise). Working with interfaces to military problems, specifically mobile robots for use in the field. Using the same techniques of asking about the specific needs of the users, they developed controls in the thumb and fingers turning any graspable object into an interface. Moving to mine detection, Warner worked on spatially locating mines. A system that tracked the head of the mine device and gave it a graphical interface that built more comprehensible visual interfaces than the previous audio interfaces -- the users could see the pattern of mines after learning those patterns.

The humanitarian arm of the military doesn't practice for their operations as much as they do in combat. The problem had to do with managing critical time decision and communicating across complex bureaucratic boundaries. Sharing information between refugee camps and operations management was the first step. Head mounted displays allowed geographically remote managers to help individuals on the ground across long distances. Taking the NASA telemedicine system, they were able to teach combat medics to use systems they'd never used before. Remote experts could distribute their intelligence through these communications systems. This also allowed the team to offload activities to geographically remote teams in different time zones when they got tired. Abstracting relationships for communication through visualization and graphics allowed multiple stakeholders with different concrete conceptions of situations on the ground.

Then Warner got sent to refugee camps in Africa. Being on the ground in such situations massively changes one's perspective on the realities. Dealing with hard problems like this requires getting serious about the truth of the problem.

After 9/11, Warner got pulled into text analysis systems used to find indicators for things like disease. Looking at problems like major disease outbreak, we need new methods to facilitate decision making across complex groups with conflicting interests. Warner again showed the use of abstract computer visualizations. After 9/11, the amount of data to scan increased exponentially. The system was overwhelmed with information. Taking the same method of geospatial location and grographic mapping (again using abstract red and green rings... red is bad, green is good, big is more, small is less). Warner showed an example of using data to understand the content of prayers at mosques. In another case, sensors around the freeways and physical area around the Superbowl. This method was repeated in the humanitarian operations between Kuwait and Baghdad.

At Sandia Labs, looking at the physiology of online multiplayer games, a group developed a "quad pod" -- a four-person synchronized computer system with physiological monitors, providing real-time data streams. They were looking for leadership performance qualities in real time. They also conducted practice experiment at the Burning Man festival. This sort of practice is useful, for example, for thinking about evacuating a city of 30,000 people. Warner used these same techniques to set up a medical operation center in Afghanistan. In a world of instantaneous text messages, why did people die three miles away during the tsunami in the south Pacific. The goal was to get imagery from helicopters to NGOs working relief in the area.

Things happen in space and time in complex ways. Going from the real world to data plotted on a computer to see the ways that a situation is managed. Warner called this "sharable situational awareness."

Overall message: large networks of heterogeneous sensors can help provide new pattern matching opportunities in human subjects across otherwise inaccessible computer and social networks.

Beyond Q&A: Assessment Methods for the Next Generation of Serious Games
Sande Chen, writer and desginer
David Michael, independent developer and author.

(note: I was standing at the back of the room transcribing on my Treo, and I had to leave for a phone call, thus the paucity of coverage for this session, which was more interesting than my meager notes might suggest)

This roundtable discussed gauging the success of games proof of learning, integration with classrooms. What does the future hold for testing and assessment. One wing is the future of serious games as a rtecher monitoring the performance of students. Another is an automated computer program that scales the performance of layers.

The human and the techology side aren't necessarily mixing. The hum&n currently needs to do most assessment, but some will be offset with technology. Pre-assesment iin educational contexts requirr leveling students into the roper levekl of challenge. Even letting the games into the classroom is difficult.

The future is probably somewhat between the two. Prescriptive learning has been servicing specific answers... The future will look at the fuzziness of answers... The grey areas in games is not explored. This might be opposed to good techng, which deals ith situated knowledge.

Waging Peace Through Chaos
Nigel Papworth, 258 Productions
Håkan Wallin, University of Luleå

Nigel and Håkan presented Foreign Ground, a research project for the Swedish military. The product was developed to address the "fourth generation warfare" where indecision and globalization rules. Normal military tasks have changed, for example through civilian contact rather than military contact. Different cultural values exacerbate the problem. The challenge facing the military was developing a game based application to illustrate cultural and environmental sensitivities.

The test scenario was Monrovia (in Liberia), in which Sweden is involved. The game had to be a mod of an existing platform, it had to include a believable dialogue system, and the AI has to take the dialogue interchange into consideration. Moreover, the result had to be indecisive, "fuzzy."

They began in April 2005 and delivered a prototype in October 2005. A production crew of 10 worked on the game. The team was distributed throughout Sweden. They used video-conferencing and web-based project databases to collaborate.

The purpose of the game is to teach culture-related social skills. A dialogue system needed to affect the behavior both of the soldiers and the civilians. They did not want to use Skinner-style operant conditioning. Many military games work to train specific responses to controlled stimuli (you see a soldier, you shoot). Instead, the team wanted to hone the sensibility of the soldiers to the environment. They wanted to stimulate their awareness of cultural idiosyncrasies, and increase their ability to negotiate and control situations by non-violent means. They wanted to illustrate that there are no set answers to difficult situations. The same eight steps would not necessarily produce the same result every time, although it would always be logically consistent.

The development team didn't have this specialist knowledge. The first thought about acquiring a conventional information expert to collaborate, then iterate. This would take two much time. Instead, they build a general platform that the instructors could fill with content. This is more flexible and allows more general adaptability.

Games are interested in people's reactions to systems. Non-realistic techniques that appear credible to the player are sufficient. The player ascribes their own perspective onto the models. Playability was a great concern: a game needs to feel fresh and flexible every time. They also wanted to allow realistic restrictions that would allow freedom of behavior within the restrictions of the environment.

They used Unreal for the game, but set the level of graphics quality to scale with the production time available. Programming and graphics took the most time. They also did a great deal of role play with physical toy soldiers on desks. The idea of chaos came out of this physical play: moving the objects is easy, but an AI that does it smartly is hard. The AI had to control at least 50 agents in a "fuzzy" way, and it had to be credible. And the dialogue had to be built into the game integrally. All this had to be done in 6 weeks design, and 4.5 months development.

AI had to deal with group behavior, positioning, pathfinding, and the motivation of agents in subgroups... for example, is the agent capable of stealing? killing? All of this had to be done with limited numbers of parameters. Nigel was influenced by the book The Newtonian Casino, about students who tried to break the roulette system using microcomputers, the fictional heros of which were the original proponents of chaos theory (small divergences in start states can produce radically larger end states). The team used a very small number of parameters capable of producing many different end results. Finally, they settled on two parameters: Aggression and Threat. Aggression services needs, threats stop agents from getting them. They built a model:

A - T = G, where G = the governing value for controlling a game state choice. They also added strength:
A (x S) - T = G, such that numbers could affect aggression.

The behavior system was then linked to the dialogue system. Nigel then loaded the game and demonstrated the process. The scenario was a patrol through an area of Monrovia. They pulled up a character interface for dialogue. Pointing a rifle at a character, for example, causes him to run away. The dialogue system allows a choice of numerous dialogue segments in a tree, which is directly affecting the AI. They then demonstrated the ability to spawn a new character and define his attributes using simple menus, including the ability to increase the aggression, strength, and threat. This system also allows the user to author new dialogue for a dialogue tree. This new dialogue can be controlled by rules, for example, a new statement is only said if aggression is at a certain level. A state machine editor allows the user to govern what the NPC does and when it does it, for example, move from walk to attack based on a change in governing values.

The interpretation of the AI is what's most important to create realistic behaviors with simple components.

Serious Games for Religion and Religious Institutions
Les Perdew, Alpine Studios

Perdew started making games in 1987 on Commodore 64, and worked since on many titles in a variety of roles. He described wanting to work on games with more meaning, and has chosen to focus on games with a religions theme. They first created Do You Know the Bible on PC, which has sold around 60,000 units, and just released The Bible Game on GBA. The Bible Game is a GBA game in which the player combats heretics on a quest to gain the armour of God. The player captures "decievers" who then prompt questions about Bible knowledge. The player gets keys for answering questions correctly, and eventually gets to fight the "master deceiver." The game is intended for kids 8 - 12 years old. Do You Know the Bible is another quiz-based game played on the PC as a multiplayer game. each player has a key to to respond, much like the old quiz game You Don't Know Jack (but without the wry humor).

Bruce Thompson was also on the panel, from a background of marketing in Fortune 100. He founded Agile Interactive who create games without violence. They are m more focused on publishing games in the "family values space." He compared the opportunities in games to that of the books of C.S. Lewis. He claimed that a $25 billion market exists for Christian games.

Perdew argued that the Christian market is part of a broader trend in "value added games" -- games that are more than just entertainment. They want to create games that "were good" and didn't include "blood and gore" -- this was a real challenge because so much of the market expects those games. He described the change as a "moment of conscious."

Sales of Christian products through the specialty retail channel is around $5 billion, representing around 2,500 - 3,000 stores. Wal-Mart sold over $1 billion in specialty Christian products. Games are starting to gain traction in the market. Moreover, the Christian sects are starting to gain more traction in media -- on radio, for example. The Serious Games space might also create products to help ministries -- missionary simulations, third-world infrastructure support, for example.

Many questions in The Bible Game are based on facts in the Bible, rather than doctrine, to appeal to a larger group of players.