World Creation Paradox
For months we saw ourselves come to the same major question whenever we wanted to plan for the level creation tasks. The question was basically, “Where should we start from?”.
There are always at least three main poles which need proper attention in order for a well balanced output, these are 1- Visuals (Graphics) 2- Game Play (Fun factor) 3 – Technical Feasibility (coding). If you want to create a level which looks great, is really fun to play and can be created considering the technical constraints or project constraints, then the three elements seem like they form a closed loop and it would be hard to decied where exactly the process should start from. Should we worry about the visuals first? Should we see what we can achieve technically and constrain the rest or should we consider the fun gameplay elements and then add the other layers?
There seems to be no strict answer to the above issue and after lots of discussions, we came up with a specific methodology to perform which we are following currently. Seeing a presentation from this years GDC from the Bioware team which was used for Mass Effect 2 strengthened our selected methodology.
What we are doing currently is to come up with the general visual concepts first, prepare the first phase for the level map which is 2D and then a 3D spatial map using boxes, fill it with the first stage of 3D models, play test the crude (un-textured) level in the game and implement the necessary code features and then re-iterate the loop by polishing the concepts, the 3D elements and gameplay testing and level design tweaking. We have planned for four iterations. So far it seems to be a good choice although we need more time to really be able to tune the methodology and level development pipeline.
This task is very critical since it requires very close collaboration between members of different departments. Its quite fun also.
Hmmm…what normally happens is that the art direction starts creating standards for the concept artists who will then create the concept art. Parallel to that, you research and pick a rendering/network technology which cover all your needs (no need to update to the latest and best every couple of months), and start producing a low-detail gameplay concept of the game. You then keep iterating over the 3d art and textures (and the design, combat, etc. of course) until you feel the gameplay is fun and that the small environment you have created suits the game settings/story. That’s when you go into full-blown production of more environments, models, etc.
The important thing is to have a run-able code base at all times. If developers don’t get to see their work in the game as they create it, you will run into huge problems prototyping stuff, i.e. prototyping will be so cumbersome that people will just prefer to produce blindly!
My two cents, And don’t forget,
Whatever you do, you should have fun.
Dear B.Kaveh,
Thanks for the comments. Well we are already in the production phase and our concern was more related to the details of every level and the main ideas and game mechanics have been tested and implemented. We are trying to find the best work pipeline to enable the design of a fun, nice and feasible level during production.
BTW, I guess you were the designer for Asre Pahlevanan, what are you up to these days? Your post-mortem on the game was valuable for us.