Before concentrating on the design strategy of the Worst Case Scenario it is necessary to have a look at the goals for the game. The goals that where (and indeed still are) important are these:
I have selected a tile-based strategy game because it incorporates a bit of everything. Actually it will be a replica of a boardgame that has been designed and playtested thouroughly by my friends and me. The game has its roots in Risk, but has many, many extensions that make it a complete new game that is exceptionally complex, due to the number of options. Nevertheless its easy to learn the rules (there are none, well ... almost).
I plan a layered combat system containing a dynamic number of layers. As
the computer or player builds tanks (or other stuff) a hierarchy of command
will automatically be build. Every unit can be in charge of max. 2 other
units. Every unit has a genetic code. The performance of every unit will be
measured and may lead to a unit being promoted (and his boss will be demoted
(is this english?)). Also, better performance will lead to a higher change
of reproducing the genetic code into newly build units.
There is also a fallback scenario: if a unit is disconnected from the line of command then it will defend or attack according to its own "intuition".
Another AI part that I plan to incorporate is a smart timing engine: My units can be grouped (also possible in Red Alert). Groups can be directed to a destination to attack it (also RA). Waypoints can be put along the way to adjust the path to the destination. Synchronization points can be put on this way that trigger events in other groups.
A scenario may look like this:
The player (or computer) builds 10 airplanes, 30 tanks and 5 battleships. The battleship group is send to destination X with 2 sync points. When the first syncpoint is reached the tanks start moving toward their destination (which may be the same as for the battle ships if it is a land-based target). When the battleships reach the second sync point the airplanes take off. If the waypoints are placed effectively the target will be attacked from 3 sides at once. With this system of sync points and waypoints extremely complex attack plans can be realized. These scenarios have been play-tested already in a real game (this means: Not a computer game).
For maximum reusability and upgragability the graphical interface will be clearly separated from the game engine. This means that better quality video and sound can be made without changes to the game engine and ai part.