TEAM NAME:    The Born Losers
TEAM MEMBERS: Scott Miller <scgmille@freenetproject.org>
              Ian Clarke   <ian@freenetproject.org>
ROBOT NAME:   Amherst

Amherst is primarily based around a modified version of the A* 
pathfinding algorithm (see 
http://www.cs.csustan.edu/~rsc/CS4960S02/Messer.pdf) which determines
the steps required to achieve a goal (such as "get to a base"), and
finds the best course of action to achieve that goal, within a horizon
determined by the time available.  A* is commomly used in games like 
"WarCraft" and "Civilisation" to determine which path a unit should take 
to reach its destination.

Rather than simply base routing on the most efficient way to achieve a 
goal in terms of spacial location (the typical usage of the A* 
algorithm), we also introduce fuzzy biases or strategies, such as "avoid 
crowds" and "avoid water", which will modify the directions taken by the 
bot to reach its goals.  The strength of these biases was hand-tweaked 
but given more time, could have been optomized with a Genetic Algorithm.
This should result in behavior such as chosing a circular path around
other bots or water which we want to avoid, When the bot's "state"
changes, such as when a package is picked-up or dropped, some of the
strategies will change accordingly.

The bot is written in Scheme, and was developed on SISC 
(http://sisc.sf.net/), a Java-based Scheme interpreter by Scott Miller.
