Spartan, thanks for the Karma! What do you meant by different models in your Reply340?

There are so many ideas floating around in here, its really great.  Whatever choice they make, I hope it accomplishes a couple of important things well; giving deep and rewarding economic gameplay, and strategic economic choices for the player both in their own empire as well as in their dealings with other empires.

What I think is important:

1. There should be 2 types of resources, raw/natural (wood, metals, grain, etc) and manufactured (weapons, potions, etc). 

2. Only natural resources should be created automaticly.  All manufactured resources should be created as demanded by the player.

3. Once a natural resource has been processed by a town near that resource with the proper infrastructure, the resoucre should then be available to all towns.

4. Stockpiling and organizing resources should be straight forward.

5. Even though all the components of an item are considered resources, in general the player only worries about the finished product he wants produced. Whether that is an elite soldier with a +2 magic sword, or 20 magical potions.

As long as only natural resources are created automaticly, you can choose to create and stockpile 100 shortswords for a rainny day, and don't have to worry about the AI wasting your iron to do so.  If done right, this really keeps the entire process simple and transparent to the player, but if you'd like to you can really get into the micromanagement and reap the benefits of doing so.

For instance: Lets say it takes 3 turns to make a basic infantry unit, with one turn of that time being used to make their spears. You can go to a town, and tell it to make the unit and 3 turns later you have one ready for action. If you had pre planned a bit for the need, you could have made the spears ahead of time, and when you needed them moved the stockpile to the town, and 2 turns later you'd have your unit.  The time it'd take is basicly the same, but at the time you need them you get that infantry unit out quicker.  The trade off being, you could waste the natural resource without ever making use of the item, and if the stockpile is in a captured city it falls into enemy hands.

This also lets you choose between trading, say, your iron or your shortswords. On one hand you're giving away weapons, on the other you don't know exactly what they'll do with the iron.

I just hope the economic system gives me the ability to be creative and flexible without the obvious headaches. I'm interested to see what they decide to do.

Here is another system. While I prefer to have a micromanagement system, it's probably not going to happen. So I'm trying to expand off of Camp 2.

1.     There are only natural resources (wood, iron, crystal, horses, etc.)

2.     When a natural resource is controlled, the player assigns that resource to a specific town. The player can change their mind later and assign the resource to another town. (Limit 1 resources assigned to a town, assuming max number natural resources in the game are 20)

3.     The assigned town is the primary user of the resource. (Assigned iron town with a blacksmith can build swordsmen at a fairly quick rate)

4.     Natural resources can be stockpiled in assigned towns. (Assigned iron town stores iron, Export on/off button, storage limit, but have upgrade options)

5.     Natural resources are automatically sent to other towns that have unit build orders by caravans, the drawback is it takes time and small quantities are shipped (Non-assigned iron town with a blacksmith can build swordsmen, but they have to wait for the iron to arrive. Caravan and road upgrades lower the time and increase the quantities that can be shipped)

Why wouldn't you want to automatically stockpile weapons and armor?

If I have to manually create everything to have an efficient economy, I'll go batty.  If there isn't any inefficiency in not creating them, the system is retardedly unrealistic and might as well be abstracted so they at least don't get laughed at.

If you're making too many swords, switch out production.  Such insanity would be a nightmare.

Training a new unit should take the longest time.  Something like 10 turns.  Once a unit is trained, upgrading his armor/weapon/helm (the only three slots we've heard about) should be instantaneous, provided the MR is available in that town.  No new training required for upgrade.

Caravans should go the most quickly.  Let's say that a caravan on a road should take 5 turns to cross a major continent, one side to the other (perhaps a quarter of the map under normal map sizes).  If the largest map is 224 tiles across, let's make an assumption (cause I can't find any data) that a normal size map is 100 tiles across.  A quarter of that is 25 tiles, and five turns means 5 tiles/turn.  We want roads to matter, so let's make roads triple the movement.  So, a caravan normally has a move of 2, with roads making that a 6.  Close enough. 

Builds times at factories should be between these numbers, with an average build time for sword, armor and helmet being, I dunno, let's call it 6ish turns.  Training to max training should take 10 turns, with each 2-3 turns spent increasing the effectiveness of that training by 1 point, to a maximum of 5, let's say.  Once you've made a fully trained swordsman, it shouldn't matter to him if the sword is wooden or adamant; he'll figure out how to make the most use of it, because he's...you know, trained.

Training mounts should take a similar amount of time.  Training spellcasters and priests should be longer than that, with perhaps some increase in time as you go up in spell levels.  Fireball wielders can come in 10 turns (maybe a level 2 spell), but firestorm wielders (level 5 spell) might take 50 or more turns to train.

What does this matter to the economy?  It goes to the question of how fast things should go to provide a balance.  If training is too quick, you'll always just build fresh, and scrap your old units instead of upgrading them (in CivIV, for example, upgrades are instantaneous, if you have the cash.  If it only takes 2 turns to build a fresh, the only reason to keep old units around is the promotions, which I think we're abstracting to "training" here, unless there's another level of unit complexity that we haven't seen yet).  If caravan movement is too quick, it (in the limit) essentially becomes a CivIV model again, where a resource somewhere in your empire means that resource is everywhere.  The increase in sophistication here is just that we're limiting the total amount of resource that can be used per turn by making a limited resource production rate, not transport rate).  If conversion is too fast, then you have the problem where caravans become less useful, since you'll just ship all your stuff to the frontline cities and convert it there, rather than try to build a network.

Thoughts for a Monday off...I just wanted to get them out there before the next week hit, and I lost it all out of my head.

No. Let me make this plainer for you.

Where it says "the iron ore is then directed to go to the city of Torgeto" in camp #1, this is a "push" model. This is made more apparent by the mention of the sliders and ratios directing how much of each resource gets sent where. We are directing the resources from their point of origin. To do this without automation is potentially a huge pain; instead of clicking a central point and directing things to filter into it, you have to go round every exterior point that needs to be redirected and readjust it manually. Readjusting things on a "pull" model, like camp #3 - even without explicit automation - can be done from that one central point, and this is true even if you forced the user to create each route from an outlying area individually. In one case you click on maybe a half-dozen different points and direct them each to one location; in the other, you click on the one location, and make maybe a half-dozen routes. Camp #1 has the potential to be relevant if sources of resources are few and their users are many, but with "everything" being treated as a resource, it doesn't appear this will be the case. Camp #3 is more useful if sources of resources are many and heavy users are relatively few - a much more common 4X case. There's still the issue of balancing the resource output of a rare resource being directed to several locations without having to click round all the cities it's being used in more than once (which you'll presumably do anyway to start construction of the units using the resource), but I suspect some kind of assignable priority metric would do a half-decent job on this.

There are a number of issues with AI governors, which are a different thing again. Unless what they do is immediately apparent to the player, there's a strong risk of them making decisions that the player wouldn't in their place and doesn't know they're going to make. This doesn't necessarily make them bad decisions, but the player either has to accept a niggling doubt as to what the AI governor is doing, or spend time watching it to see what it's up to, both of which lead to unsatisfying gameplay. (If you're just willing to assume it's okay and hope and make units with whatever resources it decides to grant you, you have no business telling us that you want a complex economy; you're not planning to deal with the economy directly, so you'd really be just as well served by a lobotomised model.) In the case where some invading army has severed your supply lines, knowing exactly what the AI governors plan to do about it is important, as it's quite possible that their plans and your emergency solutions do not line up. If the AI governors decide to all send the resources to a particular city that can efficiently make units to stem the invasion, they are deciding your strategy for you, which is miserable; if they sit on their behinds and cut all supply chains, this is liable to be sadly suboptimal; if they send them to the closest appropriate manufacturing point, this also may require some tinkering to get things functioning to your taste. Again, the "pull" model matters: it is more likely that you will want to redirect five or six cities to one or two locations than that you will want to direct these cities to a large number of locations, and so clicking on the destination and working from there is likely to afford the more compact UI.

In short, "automation" is not a magic wand that will solve all user interface issues, and thought needs to be given as to exactly what form such automation ought to take and what the interface should involve. Some of the details of that interface are already being discussed (sliders/ratios in camp #1, "pulling" resources in camp #3), and deciding on these details will constrain what automation possibilities are available.

I hope this helps.

There's also an intuitiveness-of-UI issue with camp #3 in that, once you've been presented with a new resource, you can't just direct where it's meant to go from the point it's produced. This feels clumsy.

So, I guess I have issues with most of the camps.

Camp #1: easy at first, intuitive with new resources, unintuitive with building new manufacturing centres, quickly nightmarish when disrupted.

Camp #2: perhaps too simple (though I suspect that increased clarity of which towns are actually important will aid strategy, and the strategy games in which the most sophisticated play exists generally have a very high level of abstraction e.g. chess and go)

Camp #3: unintuitive with new resources, intuitive with new manufacturing centres, still requires work when lines disrupted.

Camp #4: seems to add a lot of complexity without adding much of value

Some UI which incorporates both push and pull options: this may be clumsy in operation, owing to its redundancies, and increases the risks of doing things without quite understanding their effects elsewhere, but it has the possibility of combining the best of camps #1 and #3. Still doesn't entirely resolve the need to change up the supply lines when an invading army marches in.

I'm actually leaning towards camp #2 now, the more I think about it...

I don't doubt that Stardock has smart people.

What I am saying is: Everbody is different. And so people will have different ideas about what the right decision is given a specific situation in the game. With little control over what the AI does in such a situation, its decision can only please some people and possibly annoy the rest.

I feel very stressed from reading all the posts asking for camp 1. If camp 1 is put into the real game I am sure I will play Elemental for a while and then get bored that I am playing a spreadsheet and not a game.

The success of Sins or Demigod have been because at the core, the basic concept and design is relatively simple. Simple is good it works for the most people. Their are those who love spread sheet games and their those who like super simple games. I think that the masses are typically some where in the middle and I say make a game for them.

Camp three sounds pretty simple, but I think that it could be tweaked a little to make a little more indepth on how much micro is needed in order to pump out an elite group of +10 ass kickers. It sounded to me with in camp three there was already some wonder how complex the system should be based on the multiple item requirements to the simple local methood.

Camp Three's idea is simple, takes the least amount of effort to maximize it's effect, and is also able to get more complex if needed. I say that camp three is the best bet for a fun and interesting game.

I am sure Elemental will be nothing like civ4, but the greatest compliment I could give is that it is like civ4. Civ4 is simple and yet complex. I have spent hundreds upon hundreds of hours in civ, and like civ I am sure the player will be more converned about some big bady knocking at his door than wondering why he still hasn't got that shipment of 5 ore so he can make 1 sword and send it half way around his kingdom and 30 turns later just to pop out one unit to go kill that bady. Maybe you guys want to play a game like that, but I don't because it would suck.

Going back to my Civ4 point, I think you will see that most people put the workers on auto-pilot. Sure they get stuck and spend a gizzilan turns building a farm and then a mill while the city is constantly switching in and out of stagnation , but it is done because micro'ing all your workers in civ4 is a pain in general and the workers can usually do it better. The Cities in Civ4 also love to select the wrong plots to use. If you are a expert civ4 player then every once in a while you check in all your cities and make sure the stupid AI hasn't sacraficed 10 units of production tile for a tile that has 1 extra food and 7 less production. Now Civ4 is super far away from camp 1 and the AI in not bad 90% of the time or at least 50% of the time late game, but I think it goes a long way to show that to make a even more complex system work then stardock must seriously code the heck out of a minor part of it.

I beg you guys at stardock DO NOT DO CAMP ONE!!!! I DONT WANT TO SEE YOU MAKE A BAD GAME!

Seconded. 

I remain in favor of Camp #3.  Camps #1 & #4 are both needlessly complicated IMHO, while #2 is a bit simplistic (although still preferable to 1 or 4 if it were to come down to that).  There's no way most of us are going to truly have the patience to manage 100 (or more) resources unless the system in place for it is heavily automated.

I don't think it will be easy to calculate it for a game either. Say, what if you ordered some other unit in another city that uses one of the resources you need for a production order you're trying to calculate? What if there are many units like that, requiring different resources? I doubt that a simple formula will be enough. So to predict it in a complex case, you need to run a simulation of the entire process for several turns in advance.

But even that will not be enough. What if you'll order just another unit later? Or, say, you'll make another resource-producing/converting building. It will change the whole thing once again and your prediction will be wrong. In normal games, you can predict consequences of such change, but here even a computer will have a hard time calculating all of that so you'll not have a chance either.

So, you get a complex "black box" simulator, results of which can't be predicted even by a computer itself. How will you explain to a player how it works in a simple terms? There is no way to do it.

But there is more. What's the point of all that? You don't have a 10-turn advance warning or scouting in TBS so you can't plan ahead your production anyway. When you defend, it takes too much time to produce units on warning. When you attack, you just account that feature as some variable delay in your plans, but you don't need to know an exact number of turns, and it doesn't matter if it's just simple delay due to a production speed or it's some complex simulator running in background. So, that complex simulator is useless and counter-productive. You may consider checking the link: KISS principle.

For what it's worth I'm voting for option 1