First Test of Prototype Game Level - Paper Version

As part of their midterm, students in the intermediate conversational German course tested what they had developed so far today in class. Over the past few weeks, students in this course have been researching assigned identities for the NPCs in order to develop more realistic depth of character. I am hoping that this research will result in more carefully articulated characters instead of typical German stereotypes and that students' work in small groups will help them to formulate connections between their NPCs that will, in turn, create more authentic gameplay. Research consisted of using available library resources, databases, and online materials to flesh out the sociocultural backgrounds of the NPCs more fully and the dialogue they developed based on this research was to be augmented by our class discussions of our weekly reading topics (e.g., East-West German relations 20 years after the fall of the Wall).

Students came to class today armed with their dialogue on 3 x 5 index cards and were provided prototype evaluations sheets to evaluate how the gameplay functioned. It seemed to me that this testing and evaluation was a critical part of the entire development process as students finally understood how their individual components fit into the big picture. A paper-based walk-though is standard operating procedure in many game designs studios early in the development process. I found this procedure difficult to fit into the academic calendar, however, thinking that it would perhaps be akin to midterm testing. I will have to think more carefully of how to apply external design assessment models in an academic setting in the future.

Below are the comments (verbatim) provided by the student prototype testers and their respective questions:

1. What seemed to work particularly well with the prototype test?

• Using ideas in class with my character's conversation, mostly of unemployment and discrimination.
• General game will work well when finished.
• Getting food and money.
• The individual conversations.
• The character stories that we worked on.
• Some cards led the player to his objective well / were appropriately complex.
• It flowed within each person's cards, but not as a whole.
• When we all sat here together it seemed to come together more easily, when we met out of class is was all unclear, also since we finished our cards we realized the missing connections.

2. What did not work so well with the prototype test?

• We all didn't realize how crucial the connectors were so it was more individualistic I think.
• Everyone did not really know what was going on. It was unclear.
• No one was sure which objectives to do or how to lead the players to certain objectives; we didn't know who the player was; some cards were not so conversational; some information about the objectives was contradictory between different NSCs.
• Connecting goals in a realistic way.
• Relationships / interactions between people.
• Finding some objectives.
• It would have been nice to know the background of the player beforehand. Should coordinate our cards better.
• Weaving other characters into my conversation.

3. Were there any moments when gameplay seemed to lag? What will you do to correct this in the next design iteration?

• Throughout later note cards, it was hard to find conclusions to the conversation.
• When there were dead ends. Not all goals were addressed. Mine was too involved.
• Just from uncertainty on how to play.
• The interaction and relationships between the characters.
• It was hard to find a person who had a branch to a specific objective. Sometimes you can talk to a person for 5 minutes only to get the same objectives that you just got. We have to have more interconnections.
• Some cards were too conversational and not enough about the objectives -> include more references to objectives or other NSCs who can help with the objectives.
• Yes, it was because it was the first run through I think if we all had see what others had done before it would be more clear.
• Only when there was a dead end and no connection.

4. Were there any challenges that were particularly easy to solve? Or too difficult? What will you do to correct this in the next design iteration?

• Some answers went directly to the goals too easily.
• Getting food, almost everyone could lead there.
• Some cards made the answers to the objectives too easy -> hide the answers under a few more layers of conversation.
• Geldautomat and Döner were easy. Telefonkarte was hard. We just have to keep working to connect all 6 objectives.
• Geldautomat was too easy (lead player along more).
• Too difficult: Telefonekarte, Untergang. We will correct by having people actually lead to these objectives.
• Mine were too long and involved. Others seemed to be fine.
• Figuring out order and connections between each card.

5. What are you going to do in the next design iteration in terms of character development?

• Research more the other characters, integrate them into my conversation.
• Go more into depth, maybe throw in some "Easter eggs," More clues.
• Incorporate other players more.
• Do more than just give the answer right away. Create more relationships with other characters. Help lead to most / all of the objectives.
• I am going to connect more other players and have a way to reach all 6 goals from my character.
• Write the part of a player for an American journalist; reference/lead the player to others who can help with each objective; more information about my character in the dialogue and more the objectives a little harder to achieve; reference class topics / current events when possible.
• Mention other objectives, add in "is there something I can help you with"
• More connections with other characters, not leading to a dead end right off the bat. For me: Zeitungen.

So, to summarize: It seems that students will need to work more carefully on how to weave their dialogues into a cohesive whole that can lead the player along specific challenge paths, but without being blatantly obvious so that gameplay becomes predictable and boring. Also: the students will have to figure out how to incorporate game objectives into their dialogue in a manner that seems to occur naturally and conversationally. As for many of the students the game seemed to "come together" for the first time only during the prototype testing phase, this is something that needs to happen sooner in the semester, rather than later. Here are some images from today's prototype test:

Left to right: Elizabeth Leman and Matt Trucksess figuring out points of connection between their characters. Elizabeth's character is an apprentice auto mechanic by Volkswagen, trying to help the player, a rookie reporter from the USA, navigate the intricacies of the German pedestrian zone.

Left to right: Christine Mader, John Lesko, and Matt Trucksess hammer out the details of their game components. Christine's character wants to have a smoking ban in Germany.

On other fronts, I am still working on my submission to L2 Journal and I will probably hold off all 3D mesh and texturing development until this article is behind me. Once I have that done, I will be working almost exclusively on the game and getting up-to-speed with Blender (and Python).

Been Really Busy. Sorry.

Sorry that I have not had the chance to put any new posts up. As I had anticipated, teaching responsibilities this semester have put a slight damper on my development activities. In addition, I have also been investing more effort into the research side of the endeavor - and reaping the benefits! I have two articles coming out in the next few months: one in The Language Educator and the other in Computer Assisted Language Learning. I will post more here on these articles later.

I'm also working on a piece right now for L2 Journal, a new online journal based at the University of California. The first effort at publishing here was rejected (which I would have to agree with - it was a half-baked idea), so I'm reworking what I got into an examination of narrative structures in 3D computer games as a potential means of supoprting second language acquisition. Hopefully, the second time around will be better. And then, well, I'm thinking of focusing more on Blender to get one game level designed and use what I have learned in this push forward with a book project. Gotta feed the tenure beast, right?

Toward More Perfect Textures

I have been playing with GIMP to create the textures for the fountain. Although not as "intuitive" as, say, Adobe Photoshop, I have found GIMP to be adequate for my needs. It is also free, which it great for an indie game producer/university professor with a tight budget yet big ideas.

Using Blender to export UV maps on which to create the textures was hardly easy: It seemed to take me forever to get a handle on the idea of the texture channels, not to mention that key parts of the interface were spread over different parts of the application. Perhaps the upcoming Blender 2.50 release will address some of these issues. Anyway, here is what I got so far:

I was particularly please with the water and rust stains I was able to replicate using the GIMP airbrush tool and grabbing samples from stock photographs of rust. Nevertheless, there are still some areas I would like to address in my second attempt at creating a more perfect model texture:

1. Remove seams: Blender unwraps a 3D mesh along predetermined and computer-generated seams. Here, the rust stain actually crosses a seam, making it visible to the viewer. Although the seam here is not so noticable (compared to the other side of the fountain and the lip of the basin), seams can be reapplied in a different pattern to an existing texture map, thereby allowing edges to be graphically altered and minimized.


2. Smooth stretching: The curved surface of the basin lip introduced a lot of image distortion when the texture was applied to the mesh. After unwrapping the UV map, I will probably have to go in an manually tweak the map in order to minimze stretching.


3. Create bump map: Although the stone texture looks fine if you look head-on at the fountain, the way that the light plays on the fountain at an angle creates the sense that the fountain is "unreal," are carved too perfectly. This is not the feeling I want to create with a fountain that has been exposed to the elements for years. A bump map will break up the surface reflection a bit, thereby creating the sense that the stone has texture. I will also have to cut down on reflection a bit, too.


4. Resize textures: The main pillar in the center of the fountain has, I believe, the right sized texture applied to its mesh. The grain of the stone appears to be the right size for the mass of the structure and, accordingly, just feels right. The basin, however, is another matter and the stone grain just looks too large.

In any case, a second (or even third) stab at it will hopefully create a more realistic sense of actually being there. Once that is done, I should probably start looking at some Python programming to structure the player interactivity with the object...

Old Rusty Pipes

I spent Saturday afternoon playing with the procedural textures that ship with Blender to create the water pipes on my recent project, the fountain that will go in the center of the pedestrian zone. Thinking about how these pipes should look like, I thought that they would have been exposed to the elements for some time and would exhibit natural pitting and wear. As they are in a damp environment, they would be slightly mosit and reflect light, but not too much as to be distracting. The black iron would have slighter areas of grey highlights. After tweaking the settings and numerous test renders to get the texture just right, I came up with the following:

I have already grabbed a sandstone texture off the Internet that in the coming days I will apply both to the pillar in the center of the fountain, the basin itself, as well as to produce a bump and specular map. When I'm all finished, I will create a turntable render and post it here. I'm particularly pleased with the reflection off the water and the slight surface distortion. We'll just have to see how the remainder of the project develops.

And here is a screen capture of the complicated Blender interface, which is gradually becoming more manageable:

The large image size (1920 x 1200) is the native resolution of my computer screen. Next time: Applying the stone textures!

More Student Work on NPCs and the Bologna Process

We spent a fair amount of time in our last class period fleshing out the NPCs for the DigiBahn game. Here are some images of the work that the students produced on the whiteboard:

The students will be working in small groups over the course of the next week, working with library resources and databases to expand the backgrounds of these NPCs, articulating connections between their NPCs, and reflecting on how this context and background influences the use of language and aids in identity formation.

In class next week we will be moving away from the experience of playing the text-based interactive fiction version of the game, which has hopefully given us some ideas on how to structure the 3D version, and will begin a week-long discussion of Germany's role and responsibilities in the European Union. I thought that an examination of the Bologna Process would be interesting for them and highly relevant as students. Should also be useful in formulating the NPCs, as well.

Finally, one last image of a render I recently did:

Hast du Durst?

Since every Fußgängerzone needs a fountain, I've been busy these last couple of days modeling one in Blender in my spare time. Creating the basic shape was fairly easy, as was making an animation for the water in the fountain:

And here is the original image on which the design of my 3D model is based:

I was impressed with how easy this was and how good it looks! In the future I will work on getting the mesh textured, and then will make a final turnaround render. Maybe I can even get water to come out of the spouts? Of course, the game will not have the same level of visual quality that the render has, but I still think that we can get some basic funcionality established. I'll have to look at the Yo Frankie! open Blender game, which has water in it, to see how they did it.

DigiBahn Comes to the Intermediate German Conversation Course

Over the last few days I have been attempting to create a fusion of my research and teaching agendas by getting students in my intermediate German conversation class involved in the DigiBahn Project. In class we have discussed - auf Deutsch, sogar! - language as a sociocultural tool that we use to structure reality and, in turn, which ultimately shapes and informs our identities. We also broached the topic of language as a sign system that can be manipulated and experienced in a playful manner in order to construct meaning. In a way, approaching class in this manner is a huge gamble for me as it represents a dramatic break from the way I have taught similar courses in the past and I am not exactly sure how the students will react to the project. I am excited, however, to see where this fusion of new technology and classroom instruction will finally take me (and the class) and what type of 21st-century pedagogies will ultimately be developed. Initial feedback from the students seems positive and we spent the end of last class beginning to flesh ideas out on the whiteboard:

Students (left to right): Alexander Howard, Christine Mader, Elizabeth Leman, Melanie Reyer, Alexander Hudson, Christian Monson, Matthew Trucksess, and John Lesko.

John Lesko (l.) and Alexander Hudson (r.) wonder how (and if) it will all fit together.

Arndt has a problem with the bottle, Jana is a widow living in Stuttgart, and Florian grew up in the former East Germany. What will they have to say to each other and how will their respective backgrounds influence this interaction?

Unfortunately, we spent too long in class discussing other things, so I will have to revisit the design phase in the next course. The primary ideas, that I hope to visualize on the whiteboard, are that language can be seen as a sociocultural space of playful possibility and conversational opportunity, which unfolds based on a person's interaction with this space and the people who occupy its communities of practice. Over the course of the week we will also be playing an earlier text-based version of the game, Ausflug am Wochenende nach München, to get ideas on how to develop the NPC dialog for the 3D version.