Hey everyone! It's Jorden.

I've been unable to blog for a bit since I was working on projects that we're not quite ready to talk about publicly just yet. An unfortunate aspect of game design is the number of things that get worked on just enough to become clearly unviable and then have to be scrapped. Nothing big has been tossed aside yet but naturally we can't talk about anything that might not even make it into the final game. This previous week I put together a document on some proposed personality traits for our future Warmaster AIs (the ones that control AI corporations and the ones that can control player corporations if the player chooses to take on a role with lower authority such as Admiral, Mayor, or Pilot). I'm sure Arthur has discussed the IAUS system that we're implementing for our AI decision making, or he will sometime soon, but to put it very briefly: you create a series of weighted curves representing variables. Many of these variables are environmental such as "Do I have the necessary resources?" or "How far am I from a source of resources?" or "How close is the nearest enemy I can see?" When these environmental variables are checked they return a value and when you multiply the values together it informs the AI of what decision it should make. Personality traits are a set of non-environmental variables that don't change and they basically skew the result. This creates the illusion of an individual who actually prefers to do certain things rather than a program multiplying numbers to pick an action. The fun part is that you don't have to be a coder, or even good at math, to design personality traits and assign values to them. All you need is to understand the different strategies available to the game in question. Without further ado here are (some) of the AI personality traits I've designed so far.

Assertive - Passive (How often they create encounters/Desire for interaction) Assertive AI try to always act first and will initiate encounters with other corporations, Passive AI prefer to react and be left alone

Expansionist - Developmental (How rapidly they grow their territory) Expansionist AI grab territory as quickly as possible, Developmental AI try to build up their infrastructure in the territory they already have before taking more

Conservative - Wasteful (How they manage their economy) Conservative AI value every single Refined Material and point of Power and don't build more than they need, Wasteful AI try to gain resources as quickly as possible at the expense of efficiency and will build an oversized military and power grid to get things done with more certainty

Specialized - Generalized (How they prioritize research) Specialized AI choose a few tech trees and try to reach the end of them as quickly as possible, Generalized AI try to broaden their skill set by researching many different technologies

These traits are designed in pairs and represent different approaches to playing our game. The more an AI leans towards one side of a given scale, the more it will chose those kinds of actions even if they are inappropriate to the situation. If an AI's values place it close to the middle then it will try to take whichever approach is most appropriate and become more flexible. For the layman: Being heavily one trait is more interesting but less strategically sound. Being more on the fence leads to more strategic desicions (because the weights are less skewed) but is also a little less interesting. A good mix will lead to AI that are interesting and challenging to play against.

Naturally these traits are subject to change and more may be added if the game takes on greater complexity.