For an overview of the war between 431 to 404 BC, I
STRONGLY urge everybody of have a read of this one page on this site:-
http://www.warhorsesim.com/epw_hist.html
....it is jam packed with information.
One of the main problems of trying to recreate this war as a Civ3 scenario is going to be the proximity of Athens & Sparta. If we are not careful, then the AI is just going to march against the enemy capital right from the start, and will just keep slugging away until it is taken.
How the actual war was played-out was something like this:-
Every year, Sparta invaded Attica, burnt the fields right up to the walls of Athens, but were unable to take the city, so marched home again for the winter (the Spartans hated city walls: walls do not fear the famous Spartan reputation, and they will not run away in fright

). Meanwhile, the Athenians were unable to capture Sparta, even though she had no walls (Spartan Hoplites behind a wooden barricade were just as good).
So it was a war upon each other allies, in an attempt to strip away each others support.
One way of simulating this is to make each captical almost impossible to capture, because of defensive bonuses (even at the wars end in 404 BC Athens was not occupied....she surrendered on terms, had her 'Long Walls' to the port of Piraeus torn down, and a pro-Spartan government installed....but was NOT captured).
As 'Flood Plains' will not be needed on the map, perhaps these could be used as the site of both Athens & Sparta, with a massive defensive bonus of say 400 or 500!
This would make the capture of the enemy captial an almost impossible task, forcing it to be a 'war on the allies' to gain points before the scenario ends through sheer exhaustion after some 27 years of conflict (if one side or the other left their capital weakly defended however, then there would still be a chance of it falling).
Plus, as Louis XXIV pointed out, this would in addition cause the odd 'plauge' and population drop to break out....as it did in reality.
Just a thought.
