This is something that I think the AI should try to do, and I don't think it would be overly hard to code some simple logic rules (not a coder like you or Brad, so if I'm wrong let me know)

a) When the AI builds an improvement, have the AI try to build the influence towards an unclaimed special resource tile if possible.  Unsure if there should be any exceptions to this.  Second priority should be goodie huts.  Maybe the third priority should be towards chokepoints.  Fourth priority towards other zones of influence.

when the player builds somethings, have the same AI logic used to place the player's building if unspecified by the player (I'm sure you do this already)

Maybe at lower difficulties it would be more random.

It already does matter, just not as much and not for typical buildings. You can build a line of buildings out towards a resource and then build a missionary/town hall.

Hexes can be done as a two demisional array (Age of Wonders: Shadow Magic seems to) if the shape of the map is roughly rectangular/square. It just complicates accessing the 'adjacent' tiles, but the overhead is offset by only having to access 6 rather than 8. The main issue is the logic is harder to work out as the way to do it is not obvious.

Edit:

It would have two cases for accessing the adjacent hexes. In AoWSM, the hexes line up into colums, and the X value offset for a hex in a given colum is the same as a square grid. Accessing the hexes above and below it are the same as a square grid (+/- 1 in the Y offset). Depending on if the colum is even or odd (which one it is depends on how the hexes are placed) will determine how to get the adjacent left/right above/below the target hex. In one case the ones above will have a Y offset that is one less that the current and those below will use the same Y offset. The other case will have those above have the same Y offset and those below have an offset that is 1 greater. Switching between these two cases would only add the overhead of an if statement that checks to see if the colum is even or odd (X % 2). If the hexes are oriented diffrently (like in HoMM 1 through 3 for the tactical combat) then row is interchangable for colum.

I like the changes. Nice work!

carielf, in several of my games I have had a situation where my influence completely surrounded another faction's city and had actually forced the other factions influence to within it's city, at that point I was more or less expecting the other faction's city to switch to my faction, is this sort of situation going to be added in in future?

harpo

It's not currently planned, but I personally would like to see that implemented since it was very useful in GalCiv2.

carielf, if you could add it to wanted/desired features list it would be appreciated.

thank you

harpo

Your language of "marking" everything in the bounding rect imply to me that you've bypassed the Euclidean calculations. I'd suggest that for the sake of adjacent borders, you're still going to need a proper "distance" calculation to get things to work right.

Personally I'd rather have a game use Euclidean distance for idealistic reasons, but Elemental already isn't a Euclidean game for units (see: treatment of diagonals) so you should be using the Chebyshev distance. It's still computationally cheaper (no squares or root, just absolute values and a comparison) and, I'd say, will give more intuitive results when borders are contested than some rectangle system.

Edit: And you should be able to use the old Euclidean-based code and simply replace the calculation formula[e]. Alternately, you can use the Euclidean distance only for determining tiebreakers when area is contested (otherwise I see awkward looking tiebreaker scenarios), but allowing cities to extend via Chebyshev distance to empty land.

All that said, how is the new, proper handling of influence going to affect the limits regarding building within 2 tiles of an existing city? It seems that once influence becomes more firm as a mechanic, you could keep track of exactly which map tiles "belong" to which city, avoiding the awkwardness that sometimes arises when attempting to claim a resource between two cities.

I feel like influence should bubble a little from sources of prestige when compared to corners that are just workshops or something; if you're going to make it the growth conform to the shape of the city at all.

Here, I'd suggest making the influence revolts be based on the housing perhaps?

You get a majority of the housing, or the town center if there is no housing- you get the town.

my thoughts are based upon the town 'voting' which faction is best for the town, ie stay with the faction they are in and have a tiny influence area OR switch to the surrounding faction and GAIN access to the resources of the influence zone, possibly also factoring in what improvements are in the town, eg pubs,ins,townhalls and their equivelents contributing to the stay, and if the  surrounding faction HAS MORE of the improvements in it's surrounding towns, then switch.

harpo

It is already using the Chebyshev distance formula just by the way it's coded.  I've also seen this algorithm called the max(dx, dy) algorithm, where dx = change in x and dy = change in y.  

For a city hub at location 0,0 (for convenience, we don't actually use negative coordinates) and radius 2, these are all the points that the code loops through (using 2 for loops, for those of you familiar with programming):

(-2,-2) (-1, -2) (0, -2) (1, -2) (2, -2)

(-2, -1) (-1, -1) (0, -1) (1, -1) (2,-1)

(-2, 0) (-1, 0) (0, 0) (1, 0) (2, 0)

(-2, 1) (-1, 1) (0, 1) (1, 1) (2, 1)

(-2, 2) (-1, 2) (0, 2) (1, 2) (2, 2)

Take the first point, which would fail under the Euclidean method because it would have a distance of 4:

d = max( abs(-2 - 0), abs(-2 - 0) )

d = max (2, 2)

d = 2

Since all points are relative to the center and the radius is 2, the max distance of any point will be less than or equal to 2 already, just because of the way that the loops are setup. 

This does not affect the limitation of how close you can build to another city. While improvements exude influence, the influence is not specific to that city, and with this code, it will be possible to have your own zones of influence overlap, so it wouldn't really help anyway. However, now at least you may be able to get the resource within influence without needing to build another city next to it, and then you could fortify it.  It's not as convenient as having it absorbed into a city, but it will be somewhat improved.

Well what I meant (re: gameplay and building placement) was that if influence/territory is a well-defined quality, perhaps the map tiles could be assigned ownership not only to a player but to a specific city. I am assuming that the "at least two tiles away" rule is meant to simplify the management of improvements-to-cities situation. If each city has its own territory influence that competes against other cities (even in the same nation), then ownership is simple. Granted, I don't know how you're representing these things internally but if it's doable, it would make for more natural feeling gameplay. The two tile limit feels very artificial, and the fact that it can in certain (and by no means exceptional) circumstances lock the player out of a resource tile is annoying; I'd rather not metagame my city placement.

In short, I don't like the two-tile limitation, and it seems to me that the influence boundary calculations might allow it to be removed.

But apart from that, I am curious about [player] boundaries: How will you break ties in influence conflicts?

Okay I like the underlying concept but don't really like the "square" look in the bottom pic and am wondering if the "corners" in the border area can be softened more, if only for an aesthetic effect.  The spread of influence needs to look like an organic flow, IMHO, which is betrayed by hugging the lines of the grid too closely.

This is really looking good! I appreciate all of Stardock's hard work!

Although this all looks and sounds pretty nice, why is this being focused on?

I'd much rather you talk to us about the spells and the core changes that are coming to the game. I didn't even know that there was an issue with ZOC's.

But that's just me

Cheers,

V.