I ran into this problem recently. I just played a game where I city spammed to push monster spawns back and in the process I spammed libraries. Well, that quickly maxed out my pop. I never ran into it with a small number of cities as the prestige gain, combined with housing tech improvements, outpaced anything I could spend the population on.

There have been many times i hit the population cap and i have had to destroy a few buildings to make some troops. If you can't find the food resources on the map and your behind the comp in production, you have to focus more on the warfare tech tree or you will get run over by the ai.

However, with the tweaks in version c and d, i find this less of a problem for a HUMAN player.

Now for a COMPUTER player, they run into a different problem. Since they can't predict when they will need an army, they will take advantage of the short term benefits of building more resource production - the computer can't make a decision on the cost benefit of saving 30 population or so to pump military if it needs it. Since they have all the bonus gold, they can keep paying for more buildings and end up burning through their available population. That's why frogboy was saying they did not build groups much before. Then they would build a hut, free up some population and pump out more single units and spam some buildings.

Ai play style is different from a person's play style and frogboy is slowly bridging the gap.

Incidentally, did you play frogger much as a wee tadpole Brad ?

My standard strategy is to spam infinite cities for their trade route bonuses and spam non-stop libraries in every satellite.  I've STILL never hit the pop cap.  You're doin' it wrong.

I don't depend on those trade  routes for gold and i find it not worth the resources to defend those little cities which don't provide me very much benefit when i am swimming in food. Besides, caravan routes die easily to things and i hate micro managing the trade routes.

I just hate that all these strategies sound so "gamey". Like in the real world, "Spamming" cities wouldn't work because there wouldn't be enough wealth to distribute on the "infinite trade route bonuses" and with spamming libraries you would run out of efficient researchers- making it pointless. Housing would be based on the wealth of the individuals living there, not based on some channelers random idea to upgrade all huts into houses. It's just a shame that more realistic mathematics haven't been put forward as ideas. I'm not a mathematician to say the least but I imagine equations similar to these could work

[individual citizens skill]x[need or lack of that skill in area]=[individual citizens wealth] if building pre-requisites (smith for a blacksmith) exist.

So naturally, high skilled citizens could afford the houses or manses while less skill citizens would get paid less, and have huts and thus having less to contribute to the coffers, or caravans. All this would be done in the background. Random citizens would have random skill, no need for micromanaging anything. So When tax time come around...

A caravan arrives in City A. Takes [% of all citizens income based on the tax rate you set]+[small % of prestige?] (higher living). Caravan would have to travel from City A all the way back to the capital then with the money it received from City A before that money would go into the coffers to be used by you. 

I imagine the devs could come up with better equations but I've never understood why games have such gamey equations instead of looking at how reality works and trying to make equations based off of those observations. This is supposed to be a living breathing world. If this could be a game where a natural, world like economy could flourish and cities grow and shrink based on wealth, and resources are treated as valuable as gold, you would have a game that would break barriers in the cause for immersion. No game has ever been able to make me feel like I was actually the leader of anything that mattered. Every game has me thinking up "gamey" strategies just to win. I want to have a game that actually makes you feel you are part of a real world and making a different experience every time you play. Is that really too much to ask for?

I recommend playing some Dwarf Fortress.  Not that I don't think the Elemental mechanics need some serious revision (they're basically all fantasy/strategy tropes mixed together with very little feeling of coherent direction), but there are some serious levels of abstraction that most players want in a strategy game.

what about TONIGHT'S AI WORK?

Happy new year all!

Couldn't agree more. I was sold on this game because it was supposed to make you feel like the leader of a nation in a fantasy world, with living breathing cities and people. But what we have is a game where you have to rush towards a victory goal by building 'cities' filled with libraries or arcane laboratories, so you can be the first to start cranking out uber-units. The only thing I kind of care for are my heroes and children, because even though they are all pretty much the same, at least I invested some time in improving them...

okay - finished reading the whole post.

I keep wondering if the AI could think of the future in terms of available present "virtual" units.  "I really need this other building or unit, so I'll start building it" where "building it" means doing the necessary research/buildings/saving for townspeople/etc.  It means more work to calculate how much a virtual unit costs.

And the crux of that problem is determining which virtual units (out of an infinite space) to choose from.  It's a fuzzy problem, balancing between: 1. economy, 2. intelligence (recon) and 3. security (military).  I'd want to use a genetic algorithm to approach this, where techs and units are given weights to determine build order.

Hope to god your underlying architecture is a clean design that lends itself to a lot of flexibility.    As the architecture, so the game.

Research balance continues to be a thorny problem. On the one hand, flexibility is always nice. It's nice when a design doesn't put limitations on players like "you can only build 1 archivist per city." On the other hand, the other options don't seem nearly as fun, useful or attractive compared to maximizing your research. It also under cuts the value of resource nodes for research.

Couple suggestions I've made in the past that I think still work now:

-DIMINISHING RETURNS, INCREASING COSTS.

Either or, take your pick. There is no trade off in buildings right now. 1 archivist will always be worth the same amount of research, and cost the same amount of resources.

In reality, setting up another 5 people in an archives somewhere shouldn't be as effective at the highest levels of research. It's not right now, actually, because the research level values get pretty high....but when you stack, you counter act the increasing scale of cost within the research trees. Since that's not really costing you much extra in the long run as a player (the resource costs, population and otherwise, are trivial. The space requirements are negligible), the player doesn't end up having to invest that much more in cheap research to get it to where they want it to go.

So every additional research building after the first in a city (each DUPLICATE that is), the cost at least should go up. At least the player is investing more heavily in something to get the benefits out of it, instead of mildly investing and getting the whole farm. Often research is the only thing left to build once the other essentials are covered, in any event. That really needs to change.

-RESOURCE NODES GO BACK TO % INCREASES TO RESEARCH

The current state of them isn't fun, balanced or interesting. To be fair, it wasn't any of those things under the old system either. % increases to research made for horribly overpowered research rates. But at least they were important. Now, LL and temples are really just saving you time and resources building that stuff yourself. (Which we know are both trivial things after the early game.)

I always find it amusing when people say games are 'too gamey' and not real life enough, then load it up and blast a couple of dragons with a wand of freezing.

In every game there will always be strategies designed to exploit game mechanics. It's impossible to avoid. If you want realism, play the game with your own house rules. Otherwise go kill some dragons in real life and don't bitch.

Just chipping in to say that I'd rather play an Artificial Intelligence (AI) then an Artificial Cheat (AC).

I haven't been playing tons of elemental lately, since I'm waiting for a few more AI tweaks, so someone correct me if I'm wrong but in regards to archers in tactical battles.  The few times I have seen the computer with archers its been fairly easy to rush and take them out.  So once the computer gets in the habit of making them, maybe work on making it keep a guy or two back to defend them also.  Obviously your sovereign can just nuke them with spells, and your own archers cant take them out, but thats less time they are focusing on the foot soldiers baring down on them.  Also I don't know if its in the game, but maybe have all enemy archers focus their attacks on the weakest enemy to try to take them out quickly or some kind of focused attack for the computer.

One way to do this is to introduce nation-wise corruption.  The total research generated by city buildings (not map-generated libraries/sanctums for thematic reasons) is charged a logarithmic rate of corruption, so that more research=higher rate of corruption.  This takes care of spamming libraries either in one city or via city spam.

My family is very excited for when massive super sized maps will be possible.  I should have waited on buying my 6GB laptop, but it's been useful for other stuff.