Announcement

**LDiCesare** · October 29, 2001, 03:39

Gary, you forgot your mantra here:
A square is a polygon

**Gary Thomas** · October 29, 2001, 11:52

No, no, you are out of date. My new mantra is:

Polygons never split!

Cheers

**Lord God Jinnai** · October 29, 2001, 19:31

Well i'm wondering if you can code up a model using polygons using current model data structure...i think that would prob be the only way to convince a lot of people.

**Simon Loverix** · October 30, 2001, 08:12

If polygons (that can be squares, which are polygons) never change and you have only one layer, you have a very static map. And thus a very static game, which is not what Clash is meant to be, I suppose. Now that's the same with squares, but squares have 1 size and 1 shape. The player will accept it as to uniform base, as one of the rules with which to play. If this isn't the case, the player will want to shape those polygons the way he or she desires, to reap the benefits of their irregularity.
FE if there are four polygons (picture a square

with diagonals). The player has three opposite ones. The fourth is barbarian land, and they raid the three other ones. Now the player wants to give a little piece (the corner near the center) to one of the two polygons adjacent to the barbarian polygon, so he has only to defend two polygons. Even better, he could make a border polygon around the barbarian polygon and defend only one. He will ask himself: it's possible to have irregular shapes, why am I not allowed to make them if I can benefit from it?

If polygons don't change, what is going to happen when FE climate changes and the southern part of a large tundra polygon becomes suitable for farming? What if the French and the Germans start a trench war a certain distance from their border (using a tileless movement system)? What if only the southern part of a polygon is hit by an earthquake (with a range of effect)? What if..
I think it is necessary to move by going from one smallest map unit to another. That's the minimum move. And consequently, you can make changes to the smallest map units. So they have to be available to change, to be different from others.

This is even more problematic when you only have one layer of polygons (and I think you mentioned 'several layers' in the previous map generator thread, Gary). Then at all times the map will have polygons with only one kind of terrain, only one kind of vegetation, only one kind of population density,... which makes it necessary to have such small polygons that they will have to be square-sized (be it not square-shaped)
If polygons don't change, you have to make them small from the start, to allow such things as in the examples above, which no one can predict (that's the fun of it). Goodbye calculation benefits.

To decide when to apply an effect to a whole square or not , you can set a priori criteria, because you know the size and shape of the affected area. Fe you can say: a flood affects all adjacent squares. But you can't say: a flood affects all adjacent polygons, because they can vary in size (imagine the above huge tundra polygon being flooded) and shape (a long polygon could be flooded all the way to the center of the continent, while the others (more compact ones) would only be on the coast.

It is easier to have small-sized polygons, since the unlogical effects of irregular shapes are less if the size is less.

Risk and civilization that are a) boardgames and consequently very simple and very abstract and b) historically mapped games, i.e. they play on a world map and use the same areas that have played an important role in history. Imperialism was not historically mapped, but Very abstract: all polygons had the same (small) size and they consisted out of a group of regular polygons. Clash is also simplified, but less and much less abstract than a boardgame. It seems to me that you can have the complexity either in the units you work with, or in the game rules, but not both. Clash is also not historically mapped, since it is meant to have its own dynamics rather than mimicking a certain period of history exactly.

Microterrain:

MoM also had the same terrain for battles on the same squares. Sadly, it didn't impact the outcome much. The effect of the battle terrain was a little higher in Heroes of might and magic, but not much. The squares weren't recognizable enough to choose the battlefield anyway. And if you did, the enemy could just run away.

**Gary Thomas** · October 30, 2001, 14:04

If polygons don't change, what is going to happen when FE climate changes and the southern part of a large tundra polygon becomes suitable for farming?

You redesign the scenario so there are no large tundra polygons.

There should be no large polygons in useable areas at all. The large polygons are ocean, impassable mountains and deep desert. Even jungle needs to be smaller polygons, since the jungle might be cleared.

Normal polygons will be from around one to ten squares, with ten being pretty rare, or possibly non-existent. If we make nine the maximum, we are talking of 300 km x 300 km. Micromanagement of climate within that kind of area, when the universe of discourse is the whole world, seems to be carrying detail to an extreme degree. Most polygons will be 1, 2 or 3 squares, but with the shape designed as the situation requires.

Where polygons come into their own is when you have a medium sized river valley (with mountains on both sides) running NE to SW, 200 km long and 100 km wide. With squares you have two squares, touching diagonally, giving the impression of two basins, and looking extremely ugly and artificial. A single polygon will depict the real situation accurately and gives an enormously more pleasing looking map. Again, subdividing such an area for climate purposes is getting too detailed.

A large proportion of your objections seem to be based on the idea that players are unable to adjust to anything but squares, or at least uncomfortable with anything but squares. As a counter argument I present all the role playing games which are map based, and if they have squares the fact is hidden. Tabletop wargames are another counter example.

This is even more problematic when you only have one layer of polygons (and I think you mentioned 'several layers' in the previous map generator thread, Gary).

I originally thought in terms of physical and political polygons as being separate overlaid systems. That would work, but I concluded that it wouldn't be worth the effort.

The old board games rules used to say that the map had an overlaid grid (usually hexagons, but occasionally squares) "to regulate movement". This was the only reason for the grid. In a board game it is not practical to measure distance with a ruler as they do in tabletop games. However, with the calculating power of a computer this limitation is removed. When the first military boardgame came out (Blitzkrieg from Avalon Hill) many wargamers refused to move to board games because of the square grid that it used, on the grounds that it was unrealistic. I know, my son was one of them.

Cheers

**KrikkitOne** · October 30, 2001, 15:32

From what I have seen in the Debate

Polygons: Assuming the Polygons are Generated at the beginning of the game and Do Not ever change shape (all effects are to the entire polygon)

Advantage, probably less computer space (because of larger areas): No "square" effects

Disadvantage, No way to affect Part of something (Greening of the Sahara is either All or None)... it follows that populations would probably stay in "small units"/squares of roughly equal area

Squares:

Advantage: set distances, area, etc. An "effect" always has the same minimum area

Disadvantage: corner effects, excessive data storage (assuming polygons are on average larger than squares)

I'd have to weigh in on the side of "squares" (ie fiixed shape polygons) with one caveat, make them hexes. That way the "corner effect" of squares is eliminated.

[PS note this comes from a position of far more gameplay than coding experience]

**Gary Thomas** · October 30, 2001, 16:56

Polygons: Assuming the Polygons are Generated at the beginning of the game and Do Not ever change shape (all effects are to the entire polygon)

Advantage, probably less computer space (because of larger areas): No "square" effects

I doubt very much whether polygons will use less space in the computer. It doesn't matter anyway.

Disadvantage, No way to affect Part of something (Greening of the Sahara is either All or None)... it follows that populations would probably stay in "small units"/squares of roughly equal area

There is no way to affect part of something in squares either.

Squares:

Advantage: set distances, area, etc. An "effect" always has the same minimum area

Set distances is a disadvantage. Moving from one square to another is always 100 km, even if you started from right at the edge of the square. The whole system is artificial.

Disadvantage: corner effects, excessive data storage (assuming polygons are on average larger than squares)

Storage space is not an issue one way or the other.

I'd have to weigh in on the side of "squares" (ie fiixed shape polygons) with one caveat, make them hexes. That way the "corner effect" of squares is eliminated.

Hexagons have only 6 movement directions, squares have 8. Hexagons have all the disadvantages of squares and none of the advantages of polygons. I really don't like my units to be forced to move in multiples of 45 degrees (squares) or 60 degrees (hexagons).

[PS note this comes from a position of far more gameplay than coding experience]

Has the game play included tabletop gaming? Or computer role playing games?

The advantages of polygons are:

The map is enormously more attractive.
Movement is realistic, based on coordinates, not an artificial grid.
Some things which cannot be properly implemented in squares become simple in polygons - roads, river, clusters of cities, for example.
Players do not have to make esoteric mental calculations about movement.

The advantages of squares are:

Some people are used to them.

As usual the real point has been completely missed. The map is actually used for two different things.

There are area aspects of the map - whether an area, such as a square, is hill, or mountain or whatever. In this context I prefer contour lines, as happen in atlases, to a grid system that chops everything up in an artificial fashion. Imagine an atlas where all the maps are shown as a square grid in which the squares are internally uniform. However it may that the proponents of squares may wish to point out to the atlas makers (and RPG game makers and tabletop gamers) the error of their ways.

The other, and actually far more important aspect is that, for movement or placement purposes every square exists only at its centre. This imposes restrictions which I find insupportable. The position of a unit should be given by its coordinates, not as the centre of the square it is in. Similarly, roads and rivers should be able to wander anywhere, not be restricted to going from the centre of one square to the centre of the next square, or along the edge of a square. Notice that nothing in this paragraph has anything to do with polygons versus squares. The real coordinate system could just as easily be implemented over a square grid as over polygons, once the concept that a square has one coordinate is discarded. However you would still be left with an ugly map.

Cheers

**Simon Loverix** · October 31, 2001, 05:29

The smaller the polygons get, the more I like them. A problem could be that you don't really know beforehand which polygons will be intensively used.

There is no way to affect part of something in squares either.

Indeed, but because you know that squares are uniform, you know that effects with a distance of x squares will influence exactly the area you want them to influence. You can't predict the influenced area with irregular polygons, unless a range of effect is used, and you will probably need to split polygons to apply the effect properly.

Set distances is a disadvantage. Moving from one square to another is always 100 km, even if you started from right at the edge of the square. The whole system is artificial.

See above. Since minimum movement step is from one square to another, you can't start from the edges. But this applies to the movement system rather than squares or polygons.

The map is enormously more attractive.

I wonder if that will still be so when the computer draws the maps? But I agree with the theory.

Movement is realistic, based on coordinates, not an artificial grid.
Players do not have to make esoteric mental calculations about movement.

This is about the movement system, regardless of which kind of polygons we use.

Some things which cannot be properly implemented in squares become simple in polygons - roads, river, clusters of cities, for example.

They can be properly implemented, but their effects become more difficult to estimate. Again, regardless of which kind of polygons. And roads linked to transport infra can be another problem, since transport infra is contained by a polygon.

However it may that the proponents of squares may wish to point out to the atlas makers (and RPG game makers and tabletop gamers) the error of their ways.

Atlases are static. If something happens, they draw a new map, with new polygons. RPG and tabletop game maps are usually nothing more than a backdrop for combat.

**Lord God Jinnai** · October 31, 2001, 22:59

Hmm...i've been thinking...

In Gary's version there would be an underlying skeletal structure of the world made up of hudreds-thousands of triangles vs. current system of hunders-thousands of squares in skeletal form.

No matter how big, the polygon will still be made up of these 100km X 100km X 100 km triangles. I then though about how you were saying that once a polygon was created, it can't be changed by peices, and then i thought to myself, "Self, this reminds me to some extent of Richard's old 'ecological provinces' idea, except its more rigid." And then i thought, "Well in Rich's system he had it so squares linked together for visual and whatnot purposes could be devided if need be, then why couldn't this, if it were used carefully and not below that 100km x 100km x 100km triangle?' Then I thought, 'But wait, wouldn't that be like the comment about polygons being devided?' and then i thought, 'Sure, it could be, but so was rich's and no one objected then. The only differance would be the base of a triangle instead of a square and the fact the wouldn't ness. haveto have a maximum size?' Then i stopped and pondered and thought and pondered and thought, and went to the bathroom and the pondered and thought.....(skipping ahead).....Well then i finally asked myself, if we used my idea, how would we impliemt it?' and then i thought, 'Well for the most part terrain at that magnitude is static. So whenever something important happens that would change atleast 1 triangle, it would send a message saying that coordiante XY is changed to Swamp...it would see if a swamp polygon next to it was there, if so, it would welcome it into its community, if not then there would be a new triangluar polygon.' Then i smiled but then frowned as i realized, 'Well wouldn't this haveto be checked alot then?" Then i scratched my head and shook it, 'Nope,' i thought, 'This is sent when it needs to be changed and something drastic or long term (such as multi-generational overuse of forest) would haveto cause a terrain to change drastically enough for the scale we're on...we can check rich's old idea to see details...' And then i thought 'Yea, i should post this' and i have

And now i'm done thinking, for now.

**KrikkitOne** · November 1, 2001, 04:06

The advantage of Hexes over squares is that for a Hexagon all 6 directions are equivalent in every way.

In a square system the 4 cross directions are different from the 4 diagonal directions.

However, the idea of storing the Position of units exactly might be interesting, but essentially that is just giving a smaller "square" for measuring one type of thing...and one is left with the problem of a sudden transition [we are on the edge of the forest, are the effects different than the center of the forest?]

Perhaps if Units and "roads" including canals, rivers, etc. were given coordinate based paths/positions and then the background terrain was represented in hexes then I could agree, but I don't want to have to cut down the Entire Pacific Northwest before I can build a single farm (or something that I need to clear the forest to build)

The socioenvironmental factors (changeable but largely non mobile) encompass areas, the areas should be all of the same size because One area might change wheras a neighboring one might not Even if it was identical in one or more characteristics. That way the "polygon" for an area represents the minimum "change region" which should be constant rather than the minimum "characteristic region".

As for "Mobile factors" (units and the roads/rivers they move on) those might be usefully modelled with a coordinate system... since rather than every coordinate storing characteristics, each unit stores its coordinates. (for a "road/river segment" it is two coordinates)

Political regions would probably be polygons, but polygons composed of certain subunits (hexes, squares, etc.)

It Might be doable for ecological regions as well, but then the same principle applies, they would simply be composed of subunits that every so often would have to be split off of the group.

I would Never do it for Social variables because those will have lots of necessary variation dealing with internal distances, etc...
They should "break up" too often

(in any case, I prefer hexes to anything else [squares/triangles] for the subunit type because they avoid the corner effects.

**Lord God Jinnai** · November 1, 2001, 19:54

Well a hex is composed of 6 triangles so you wouldn't haeto have a corner effect with triangles either and could get much better shapes too.

Plus hexes do have the disadvantage of being able to move up/down faster than left/right (or vise versa, depending on which way they are).

**KrikkitOne** · November 1, 2001, 21:51

Triangles still have corner effects [they are worse than squares], (by corner effect I am referring to the "adjacent" question are two areas actually adacent if they only share a corner, should you be allowed o move through corners because that moves you faster?.. for hexes that never comes up because they all share sides.)

Only a Coordinate System allows truly equal "rates of movement" in all directions [a square moves you much faster diagonally than vertically/horizontally]
The only disadvantage of a hex system is that either directly vertical or horizontal movement must be a "non allowed movement"..
But it has 6 directions of continuous equivalent movement wheras a square has 4 and a
Triangle has no movement that is truly continuous. (you move from a side-side transition to a corner-corner transition on every movement)

**Lord God Jinnai** · November 1, 2001, 23:21

Well the squares were using right now have a modified version of 8...ask mark for details.

But anyway if we're using rendered objects and coordiantes, we can make a direct line path using this...as far as i know the trianglular structure would only be used for visual purposes to make the terrain look more natural.

And no hexes do still have a corner look to them for rivers...triagles allow for more curvers and whatnot and the corner look can be modified so that it looks like a natural curve.

**KrikkitOne** · November 2, 2001, 15:04

I wasn't thinking as much about the graphics (since those can be modified to take into account what the surrounding hexes are ala Civ 2 Forests/mountans/hills)

What I was thinking about was the fact that a distance of "1 step" involves a different distance depending on whether it is diagonal or vertical/horizontal in squares, but in hexes it is the same

Announcement

Polygons and microterrain

Polygons and microterrain

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment