The objective here is to solve several problems we currently have with the alpha:
- the game focus shifting from generic AI design to map specific team composition
- the rock/paper/scissors phenomenon regarding the bot classes balancing
- too many AIs being created to implement map specific strategies (real sub-trees will play their part later)
Each class costs a certain amount of points depending on their stats.
Special classes like shotgun, machine gun and sniper cost more than the default/assault class but "low cost" classes are to come.
You have a limited "budget" for your team.
Consequence: teams can still vary in term of composition but also in term of size (from 1 to 6), offering more strategic options while limiting usage of special classes.
Every player will (re)start with a team made of 4 assault bots (less than max budget) and points will be earned by playing ranked multiplayer matches. Every victory adding the equivalent of the score bonus to your point balance. Balance is not impacted by defeats and draws.
Once you have enough points, you can "buy" a new bot of your choice (not sure to keep current league based unlock system).
Each map has at least 8 slots by team. Default bots positions can be changed in training mode.
Ranked matches won't display the (randomly) selected map anymore, you will just have to click on the "PLAY" button to initiate matchmaking and submit your team.
This will force players to find generic solutions.
PS: I still have to find a way to make the handling of several teams interesting, apart from testing different strategies...
Any input is welcome