When you are starting out you will probably have very few people, simply because you are rebuilding from scratch.  However your world will remain post apocalyptic only until you and the other channellers rebuild it (or cause another apocalypse).

Hey, magic skyscrapers... although don't sue when your children grow three heads that turn you to stone if you gaze on them.  Actually, come to think of it that is a problem that solves itself.

Also, remember that you need population to make units. If you make an army of 1000 men (supposedly, that's a very real possibility), you'll loose 1000 men from that city.

Also, big cities will take a huge amount of space, seeing how every time your city goes up one step, you'll physically expand to an additional square.

Personally I don't enjoy huge maps. It takes to long to play, it takes to much micromanagement. I prefer smaller maps. I wonder if most people enjoy big lar maps or smaller medium ones.

I love huge maps, although there is a limit. For example in galciv2 the biggest map size was too big for me. But there was a lot of needless micromanagement in galciv that I'm hoping will be cut out in Elemental. For me, the ideal map size is inversely proportional to the amount of micromanagement: the less tedious micro I have to deal with, the larger the map I want to play.

I think in Elemental I'll enjoy tremendously huge maps also because of the huge, unbalancing spells of the late-game. Those should be fun to use on really huge maps

If resources are done right, I'd think there shouldn't be a need for such artificial limits. To take your example, as the outposts grows it requires more food, thus there would be less for the larger city. If you divert a major portion of your resources to a single city, it limits the resources available to the rest, creating a natural limitation.

Personally I think that a city with a million people should only be possible on a huge map after a loooong time and a lot of effort and planning. Based on the goal of being able to field multiple armies each consisting of ten thousand or so soldiers, I'd like to see normal large cities reaching 100,000 or so. Assuming a soldier:civilian ratio of ~1:10, this would allow each large city to field its own 10,000 strong army. A population of 1 million would be double the widely accepted max population of ancient rome (~500,000)...

I definitely agree, perhaps to the point of thinking that megacities have no place in the game anyway. A lot of my response to this sort of stuff will depend on how the game treats time--and if even a fairly long game lasts only a few game years, breeding a million people (or even gathering most of them from the wilderness) shouldn't be possible. But I haven't seen anything from dev-land about how much internal time a game on a Most Ludicrously Large map might take.

Re the hard limit, I mostly talked myself out of it by finishing the first blurb and more than half-expect that resource logistics will provide satisfactory limits to sprawl. I just threw the hard-limit idea out there to see if it got kicked or patted on the head.

I personally support the notion of having population growth be a variable set at the begining of the game.  Just at game creation you can choose a speed at which population increases.  That way if you want a marathon game, you can, but if you want a fast game that doesn't have much time wasted while you wait for cities to grow to a managable size it can go pretty fast.

I just reread this and I can't believe how huge a mistake I made...  I blame it on being late.

With a more realistic population growth rate of 3% per year (natural + magic + refugees), population will double every ~24 game years, and you get the following figures (assuming an initial population of 1000):

1 turn = 1 week: 1248 turns  (24 years) to double to 2000 population. 8112 turns (156 years) to reach 100,000 population...

1 turn = 1 month: 288 turns to double. 1872 turns to reach 100,000.

1 turn = 6 months: 48 turns to double. 312 turns to reach 100,000.

Quite frankly, even the 6 months thing doesn't seem that reasonable to me (keep in mind that these populations are your total empire, not an individual city). This is obviously a simplified population growth model, though. But nonetheless, making population growth by turn realistic would make other aspects of the game equally unrealistic (or just plain unfun).

For one, it would take 156 game years (with a growth rate of 3%) to reach a total population of 100,000 - and if you care about realism then your army would constantly be dying of old age, and unless you make a magical excuse for heroes and channelers, they'd all be dead.

Secondly, if the time scale were, say 1 turn = 6 months, my channeler would have to be able to do a whole lot each turn, and my units would have to be able to move real far. If my channeler could only cast a handful of spells in 6 months he'd feel mentally challenged to me, and my armies and units being able to move a realistic distance per turn would probably be WAY too much for the sake of gameplay. And building something like a castle would still take a huge amount of time, if you want to be realistic. It took an average of ~10 years to build a castle (that's 20 turns if 1 turn = 6 months), and usually that was just until the castle was habitable - work would usually continue for years afterwards strengthening defenses.

My point is that there is no way for all the different timescales of all the different processes that will be within the scope of this game (and any other 4X game) to be both realistic and be fun. So in my opinion, screw realism. Choose a timeframe for each turn, but ignore it completely when it comes to deciding actual gameplay.

i think the population growth rate should be made semi-realistic. maybe a compromise between absolutey insane banging, and realistic, quaint banging.

Actually, human population growth has been higher in the past 3 centuries than it ever was any time in the history of mankind. This is due to a combination of relatively abundant availability of food and housing, better food quality, improved personal hygiene, sanitation, and modern medicine. None of these increased the rate at which people gave birth, but they decreased the mortality rate - people lived longer. Less deaths each year mean a higher population growth for the same birth rate.

So the only way a post apocalyptic earth would have a much higher growth rate than 3% (and 3% is pretty much the max - it's usually half that) is if all of those mortality rate-reducing factors were present. That is pretty much nixed due to the 'post apocalyptic' modifier . Now in the case of Elemental you could say that magic can take the place of all/most of those factors, I guess, but to go for a sustained growth rate over the low single digits is not realistic. Unless, there is also population growth due to immigration.

That would require a pop growth of 5%. So if you include refugees joining you from the hills, fine, that's close enough to realism so that only a jerk would complain. But that brings us back to the problem of how many turns will that take? Well let's see:

1 turn = 1 day: 10950 turns for population to quadruple
1 turn = 1 week: 1564 turns for pop. to quadruple
1 turn = 1 month: 360 turns for pop. to quadruple

So as you see, we have still not solved the problem of how to make pop growth realistic and still be fun. Even at a growth rate of 5%, it would take ~95 years for a population of 1000 to reach 100,000. That's either 34675, 4950, 1140 turns for your total population to reach 100,000. And that's if you start out with 1000, if you start out with less the time goes up dramatically. I dunno about you but that's unacceptable to me - unless the first several hundred turns play really fast.

On the other hand:

1 turn = 6 months: 60 turns for pop. to quadruple (190 turns to reach 100,000) - even too fast in my opinion
1 turn = 2 months: 180 turns for pop. to quadruple (570 turns to reach 100,000) - actually not bad imo

But unfortunately, 2 months/turn still forces a whole slew of other things to be unrealistic. So if a large majority agree that population growth should be realistic, and don't care about any other time-scale realism, then I'd vote for a 2 month turn system. On the other hand, if it's not important to many people, or if there are lots of disagreements over which time scales should be realistic, then I'd vote to screw it all, let SD determine an arbitrary time-scale and give people the option to make it generic "turns."

That's completely untrue. The only possible way for a race to die out is if population growth is negative. If growth is always positive then it will grow and grow until it reaches the carrying capacity of the environment. So any race that has ever died out has either had a consistently negative growth rate, or died due to some major catastrophe. To have a 3% growth rate in a hazardous world would be a miracle - we have less than a 3% growth rate and our mortality rate is tiny.

Yes - it's the weakness that balances the superior strength and abilities of individual fallen. 

common plebe ftw!

Pardon me for having started doing 'online' nonsense with a modem that required a Bell standard handset, but what the frak do you kids around here mean by FTW? I thought it meant "fuck the world" for a while, but I've finally bothered to try looking it up and I'm utterly confused.