March 31, 2010

Aurora Music

I did a lot of experimenting with the Aurora music system, hoping to essentially build a soundtrack out of sound effects. I could make it context-sensitive and everything! But, alas, I don't have the musical knowledge to do it right. There's a lot of potential there, but nothing sounded quite good enough to work. I also had several technical issues that put everything painfully off-sync.

Still, I didn't want to abandon the idea entirely. I found an album of free music called Rooftop 120 that was filled with great ambient tracks, all at 120 beats per minute. I can set the rhythm of events in the game to the same speed, keep my musical sound effects, and use this excellent music as background. It sounds pretty awesome, and I don't think I can do better by building it piecemeal.

I'm now working on smoothing out the AI and adding maps, and then it's time to start building menus and completing the outer structure of the game.

March 29, 2010

Values at Play

This term I'm taking my second game design course from Professor Mary Flanagan. This one is subtitled "Values at Play", with a focus on encoding values in gameplay.

Last year's game design class required one game per week. I'm guessing that this year will be similar. I'll post descriptions of my games here.

March 2, 2010

Aurora Thoughts


I'm making progress on Aurora. Recently I've been mostly focusing on experimenting with the UI and making the music system work (more on that later), but I'm also putting thought into the core gameplay structure. To remind you, the basic idea is that there are buildings and troops. Buildings produce troops at a constant rate. Troops can annihilate themselves against enemy troops, enemy buildings, or friendly buildings (to upgrade). Buildings can be upgraded once, doubling their production rate. It costs 100 troops to build, upgrade, or destroy a building.

In trying to create a pared-down RTS, I may have broken the genre. By removing most elements of luck and putting players on an even field, the game tends towards a stalemate. Most of the strategy involved right now involves trying to slowly-but-surely gain numerical advantage over the enemy. There are a few ways of doing this:

  • Upgrade your stars before they do and have a higher production rate until they catch up.
  • Let the enemy annihilate troops against your buildings, but not so many that your buildings die (so they are essentially wasted).
  • Focus fire on an enemy building and gain a production advantage until they send reinforcements and recolonize it.
  • Destroy an upgraded enemy building (you spend 100 destroying something that took 200 to fully build).
  • Let the two enemies waste their forces against each other and build up strength from afar.
In the current setup in which each of three players has an outer building and an inner building (all three of which meet in the middle) in a symmetric formation, it's almost impossible to gain enough of an advantage to win. Usually you have to use the last strategy I listed: pull back, lose your outer building, and gather enough forces at the inner building (while your enemies engage in a stalemate battle) to eventually sweep over the map.

That takes too long, and it's too passive to be consistently fun. I'm led to believe that RTSs are mostly defined by the imbalances, asymmetry, and chaos they introduce. Rock-Paper-Scissors combat mechanics, asymmetric factions, and interesting maps are examples of this. I was trying to distill the RTS down to something simple, symmetric, and direct, like Chess, but the "essential" rules to which I am committed don't offer the complexity that's required to make that scenario interesting.

This leaves me with three options.

First, I can add larger, more complex, and maybe more asymmetric maps. Just the possibility of having multiple fronts of battle might be enough to solve my conundrum. I plan to try this first.

Second, I can cave in and change my basic mechanics, thereby admitting defeat in my design experiment, and make it into something different. Multiple building types, multiple troop types, flanking effects, etc. I'll add complexity until I force strategy.

Third, I can report the results of my failed experiment like a good scientist. If it's still unfun (but also still an honest attempt at an essentialized RTS), I can release it as-is and show it off in order to teach others the lessons I learned during its design. Principled and uncompromising, but maybe unsatisfying as well.

For now, I'm hoping that the first option will be enough.