Why squares?

[RobN]

Can anyone tell me why the CIV-type games use squares rather than hexagons? I think hexes are better for the many reasons which have been outlined by numerous people over a considerable time.

[Lefty Scaevola]

To conform to the 8 direction keys on the keyboard number pad, perhaps, and probably a preference for 90 and 45 degree angles over 60 degree angles. You cannot have both north-south and east-west axis on a hex grid.
I prefer Hex grid myself, being an inveterate wargamer.

[SerapisIV]

Didn't Gettysburg use hexs? I didn't play it, but I thought it did.

[Deimos]

Another possible reason is that you can have more places to put workers, etc.

A two map-unit radius around a base yeilds 24 units of territory with a square map-unit layout versus 18 units for a hex map-unit layout.

[Sikander]

I prefer hexes as well, in part due to my wargaming past and in part because the distance between discreet areas (hexes) is the same whereas the diagonals on a square grid are longer (6/5 IIRC). I imagine that the reasons have to do with either the graphics being simpler or the AI having an easier time with grids (or both), but perhaps there is really no good reason.

[Lefty Scaevola]

I believe the distance discrepancy is larger, 1.414 to 1, or precisely the sq.root of 2 to 1. Approximately 2 diagonally = 3 orthogonaly. This is a significant skewing of distance, and is, of course, the reason for the notch in the corners of the 2 range base grid.

[johndmuller]

point counterpoint

Well 18 is a kind of wasteful number from a binary point of view and we do tend to think better in orthagonal fashion,...but

Hexes definitely have that equidistant between centers in all 6 directions thing going for them which is very good. Those bases of 19 might pack together without any overlap or empty space (I scribbled a bunch of hexes on a scrap paper, but I ran out of hexes before I was sure it worked right). You can still use the numeric keypad on a hex grid, just two of the updownrightleft arrows would be ignored, which two depending on the orientation of the grid, but it would be intuitive in any case. I don't know about the rest of you, but I rarely use the keypad because I generally leave the 3D terrain on and sometimes its hard to tell which key would be the right one. Strangely, its those very same kind of spots where I do use the keypad (those times when I'm sure that I know which way is which and its hard to get the mouse in the right spot).

Those right angles do conform better to our bipolar thinking, ....but

As far as whether or not the programmers can handle hexes better or worse than squarish shapes, its the kind of routines in the program where they are a PIA to get working absolutely correctly, but once you do, from then on the program just calls those routines and gets back the distances or whatever it needs without having to think about it again (please note that I said absolutely correctly). Which shape it was might make a difference in how many headaches you got before nailing it down, but would hopefully not figure in much from then on. Besides, that is what computers and programmers are for, to handle this kind of thing, right?

What I'm waiting for, among other things, is the Earth-is-Round view. I think I've heard that Civ3 will give us more of the Earth-is-a-Cylinder view. For some reason, I recall something to the effect that the north and south polar regions will take you somewhere if you move off the top or bottom, which will be interesting if true; what will you get, a secret passage to the other side of the world a la Clue? ... a duplicate of the tiles you just came from - what if you built a base there? ... there must be something I'm forgetting.

I've given some thought to a workable representation of a sphere. It would have squarish tiles. Starting from the equator, each row would have one fewer tile; thus, a typical tile would have 6 or or 7 neighbors, 1 on each side and 2 or 3 above and below. The polar areas might be made to work out just right if there were exactly 4 tiles in the penultimate ring. The required deviation from a square would be a matter of subjective asthetics.

I'm looking forward to things that might be in the next generation of representation - perhaps a move away from tiles altogether, just dealing in points and radii around them representing the centers of influence of cities, terrain, units, what have you. In between would be fractional degrees of whatever; for example, an area occupied by an army unit might be 40% forested and 20% urban and 30% water; it might be 85% ruled by me and 15% by you. Everything would be so much more natural, assuming that's what we want. Units would move some unique distance determined by the units capability and the terrain and other influences, not large discrete granulations as at present. After all, if you have a large fleet or army, it doesn't all arrive in the city together, some of it may still be outside and the city doesn't just go abruptly to forest, there are gradations. Lets use all that processing power for something else besides graphics.

[RobN]

I agree. There is no specific need for discrete locations. With the help of the processing power available these days units could be moved like they are in tabletop wargames - that is by simply measuring distances and angles on a realistic map. The shape and density of units could be controlled by the player, within limits. It would be more complicated certainly, but also more satisfying. I suppose it illustrates the age-old question - just how realistic do we want CIV-type games to be?