Architecture Complexities
Ironically, flashy real-time 3D games require that you keep visual details to a minimum. Curved surfaces, complicated architectural details and massive textures will slow down the game’s framerate and potentially ruin the level. Picking a structure that essentially consists of nothing but curved surfaces (the Roman Colosseum) may have been a bad move, especially for a beginner. I don’t regret tackling the structure, but I imagine the result would have been better (and the time spent decreased) had I cut my teeth on a building that was a bit more traditional.

Tedious Model Creation
Although modeling a low-polygon character is an arduous task in its own right, the model conventions for the Jedi Knight II engine are particularly annoying. Since the game engine allows a player to dismember enemies with swords, this requires models to be constructed and named in a very specific way. Among other things, this involves the added task of lining up textures along these multiple seams, capping the separate limbs and naming them appropriately, setting up a complicated hierarchy structure and praying that it doesn’t all fall apart when exported to the required XSI format. Add to that the fact that models are restricted to a specific skeleton that ships with JKII, and you’ve got yourself a modeling job that would test anyone’s patience.

Underestimated Texture Work
Due to my lack of experience, I simply did not realize the sheer amount of time I would spend in Photoshop creating the custom illustrations that would be pasted all over my architecture. Even though I understood that every single surface in the game would require its own map, I assumed I would only spend an hour or so on each detail. The problem I discovered is that textures are sometimes the “first impression” that either draws a player in or leaves them cold. Even if your architecture and lighting is exquisite, there is no way to cover up a crummy, unconvincing texture. The result was that I spent hundreds of hours creating and tweaking in Photoshop and Max. The one consolation is that I am more or less happy with the final result.

Not Enough Time on Gameplay
The majority of time on this project was spent on learning new tools, and then using those tools to create art assets. In order to make the game even slightly interesting, I needed to have at least one compelling location (the Colosseum) a bunch of assets to make it real (textures, sounds, etc) and at least three or four custom characters to populate the world (the models). Once all of this was complete, I found there were only a few months to really make the level and tie all these assets together through scripting and mapping. I am happy with the result and I will undoubtedly add new levels and features for my own entertainment, but I am looking forward to the next project where I will try to achieve more of a balance between art asset creation and level design/scripting.

Inflexible Engine
If you talk to many mappers/modders, they will tell you that the Jedi Outcast engine is largely useless because it only provides the ability to swap assets and make maps. Other engines such as the original Quake III allow access to the game source code which allows greater flexibility in development. In my case, this meant that if I wanted the gladiator to fight with and use a shield, I was out of luck. If I wanted a soldier to have a sword in each hand and swing them in a particular way, I was out of luck. If I wanted to completely alter the way a character interacts with the game world – you get the idea. Given my time constraints, this lack of flexibility did not affect me as much as it could have. In addition I felt that being able to leverage all of the existing swordplay and animations was worth the trade-off.

No Academic Support
When forming my thesis committee, I sought the advice of a couple of RIT faculty members that had experience in the game industry. Unfortunately, neither were interested in participating in my project, so the project quickly became a very non-collaborative effort. My thesis committee members were extremely helpful in providing feedback and critique and used every opportunity to contribute and provide resources for the tasks at hand. My chief advisor was also able to arrange a meeting with a professional 3D modeler/animator which was invaluable.

However, specific problems or technical difficulties were essentially obstacles I needed to resolve on my own. In the section “What Went Right”, I mentioned the extremely helpful nature of other mappers and modders on the World Wide Web, and I will mention again that I would not have been able to complete this project without their help.

Messy Scripting
I mentioned that the scripting implementation, ICARUS, was very easy to learn. What I did not mention is that the number of scripts I had to write ballooned in size as I got further into the development of the game. As a result, some of the scripts got a bit messy and it was sometimes difficult to track down a particular piece of functionality if it had to be changed or tweaked. The nature of level building with ICARUS makes it difficult to keep things clean and tidy anyway, but I could have done a number of things to make my life a bit easier. Some of these things could have included strict naming conventions for the script and variable names and grouping together of scripts by functionality.

Tired Subject Matter
I mentioned that the setting of ancient Rome made for exciting gameplay and a rich level of historical detail. Focusing on gladiators and the games did have certain drawbacks as well. The biggest problem with the setting is that it has truly been done to death in the entertainment industry. Before “Gladiator” ever hit the screen, there was “Spartacus”, “Ben Hur” and many others. The video game industry has been returning to the Roman history well countless times for Real Time Strategy (RTS) games, and Lucasarts is launching a gladiator-themed Role Playing Game (RPG) within a few short months. Through the course of my research, I even found other amateur mappers and modders who were busy fashioning their own creations related to ancient Rome and the gladiatorial games.

Apart from the blunt educational aspects of the game, “Colosseum” is by no means an original game concept. The attempted historical accuracy obviously precludes any truly original or creative idea, but most players will also notice countless references and tributes to movies and books alike. The upshot of all this was that, since it was such a rich source of subject matter, I never became bored and I never ran out of new ideas. The game, as it stands as of this writing, only includes about 30% of the features I wanted to implement. As a result it has become its own little hobby I can return to and tweak and edit without ever really reaching a logical end to the story or adventure.

<< previous section (What Went Right) | main | next section (Conclusions) >>