How is this estimated?
Stores not drawn: assume the place was provisioned for a siege of this long (only when no granary or storehouse is marked).
Uses of the buildings
How is this estimated?
An illustration of the route found, not a siege simulation. The breach exists only in the replay.
How is this scored?
A simplified game model, not a siege simulation. The map is cut into cells of 2 m (4 m on the big maps) and the cheapest way from the map edge to the target is found. Each obstacle costs points: a wall by its height and thickness (height counts most against medieval attackers, thickness against guns), a rampart by its fill and facing (worth more against artillery), a gate or drawbridge less than the wall it pierces, ditches by depth and width (water costs more), rivers, climbing (per metre, much more on slopes over 35° and cliffs), and every metre within flanking range of towers, gates and bastions. Ruined works count for less; buildings block the way. Added: the keep’s height and wall thickness (estimated) and a raised entrance; a well inside. Taken away: up to 80 points for the share of wall faces open to fire and unflanked (a quicker version of the Weak spots check). Only the works standing in the timeline’s year count. Ratings: under 130 weak, under 250 defensible, under 400 strong, under 600 formidable, above that near impregnable.
Green: well covered; red: open to fire and not flanked.
3D view not available on this device. The plan still works, and you can save your castle to a file.