Distrubuted Production?

[Gyrfalcon]

I was looking at the Developer Diary interview where it mentions emphasizing the tactical over strategic and had an idea about how ships are produced. I'm not sure how good an idea it is, so I wanted to see what people think.

What if the industry of more than one colony could be combined to make ships at one planet? What if production, instead of being discrete at each colony was more abstracted across the empire?

The immediate problem I can see is that you could start a new colony and have a pile of dreadnaughts built there almost immediately. However, I'm not sure how much different that is than having the kind of late-game big fleet that tends to get built. There are still budget limitations and it would be more expensive to scrap ships in one location to build new ones somewhere else when you could just move the big fleet you already have around. Also, you wouldn't be able to build big ships until you get a starbase in, although, if you focus all of your production on one colony, that would happen pretty quickly, too. It would also kill off the fiction of great ships surviving the war through many desperate battles if you scrap and replace ships frequently. The advantage would be that new ships would be up-to-date and already in the location you plan on using them. The main advantage of making ship production more abstract would be less micromanagement when putting a fleet together.

Which raises a completely different question: how will assembling a large fleet be managed? Right now, it seems that I'd have to build each ship individually on each individual colony and move them individually to wherever I wanted to assemble the fleet. Is there going to be a planetary or imperial production queue? The former is certainly something I could use and the latter would be interesting if it could be made to work.

And when I say imperial production queue, I'm thinking about something like this:
- Colonies have a flag which can be set to indicate they're available for imperial production. Otherwise, they produce whatever default item they are set for, which seems to be Research right now.
- When I want a fleet of ships, I specify where I want it assembled and it's composition. Then the game pools all of the colonies flagged as available for imperial production to assign each of them a portion of the ships I've ordered. When they're ready, they are piloted or delivered to the target I've picked.

[luciderous]

I don't know. While I definitely understand the mechanics of what you are trying to achieve, I still ask myself one question - why? What problem with 'classic' design you are trying to solve? Does it have something to do with management? Game pacing? Can this problem be solved in any other way?

Good game design is a very complicated business and changing something that works to something else without clear reason is simply asking for trouble in the long run. Personally, I am a proponent of simple design patterns. Any complexity just for the sake of complexity/realism etc. is something that should be avoided at all costs.

[sven]

Yeah, there's a lot of different ideas floating around here; and these are topics where I have a lot more I want to say... The trouble is figuring out where to start, as much as anything else

I think there's a lot to like about the concept of pooled production. Just from a pure world building perspective, it makes sense that you're empire's greatest ships aren't necessarily going to the product of a single world -- the ability to gather resources from across a vast empire to assemble a handful of incredibly powerful ships 1) has a certain thematic appeal and 2) could really simplify the late game micromanagement load.

That said -- I also agree with luciderous, before diving into a a complicated distributed production system, I think one needs to think carefully about just what you're trying to achieve, and whether or not there are simpler ways of getting similar result.

If the idea we're trying to capture is "multiple planets can contribute to building the same ship", one reasonable option is to create an SOTS-like* model, where both a local resource (industry) and a global resource (money) are required to build ships. There are things I don't like about the way the strategic balance ended up playing out in SOTS, but, as an elegant way of capturing the idea of distributed production, I think think a "local+global" ship production cost is worth looking at.

Arioch and I have been kicking around the idea of trying something roughly similar in SiS -- making a global resource like "metal" along with a local resource "wrenches" necessary components in ship building. And as I think I've mentioned in other places, we're actually thinking of broadening the resource model even further -- food and rare trade goods are also on the list of things to experiment with.

One of the important consequences of a broader resource model, I think, is that there will be more variety in the kinds of roles that a particular planet may play in your empire. SOTS, and, to a lesser extent, MOO2 both have a bit of a "planet cloning" problem -- where all your worlds tend to become effectively identical and interchangeable, as the best possible planet, in all cases, is just a fully developed, fully terraformed world. And so your empire ends up feeling very homogeneous. We'd like late game empires in SiS to have more variety to them -- and creating a resource model where there's enough complexity that you want to have some, but not all, of your planets dedicated to things like mining or farming feels like a potentially elegant way of getting there. Obviously, we have a few pressing issues to sort out with tactics/encounters/ai behaviors before we can get to those experiments. But we are planning on some such experiments before release -- and it's worth keeping in mind that the resource model you're seeing right now is slated to change.

* edit: Heroes of Might and Magic is another example of a "local+global" unit production system that allows you to in some sense "distribute" the costs of producing your armies across your empire. I think it's another example worth keeping in mind, because HOMM, even more than SOTS, is very much a tactics-focused game. The strategic game may involve an unusually large number of different resource types, but while the model is broad, it's certainly not deep

[Gyrfalcon]

I don't know. While I definitely understand the mechanics of what you are trying to achieve, I still ask myself one question - why?

Currently, if i want to build even a small fleet, I have to drill down to the production queue on a specific planet for each individual ship and then, when they're done, send them to wherever I want to form the fleet. The problem I see is that, at the point in the game where I'm putting a dozen ships in a fleet and making multiple fleets, this becomes a very tedious task. I definitely would like to see individual production queues for planets so I can set build orders for several things and get a notification when they're all done, but I'm trying to think about ways to require less management of individual planet production queues.

However, I am assuming that, later in the game, there will be huge fleets. Perhaps even a mighty empire will only have a dozen ships at the height of its glory.

[Gyrfalcon]

SOTS, and, to a lesser extent, MOO2 both have a bit of a "planet cloning" problem

I think Kerberos quite deliberately abstracted planetary development and resource production in SotS to allow focus on the ships which, it sounds like, is your goal also. Although it seems you are also deliberately avoiding a game with hundreds of planets. If I never have more than a few dozen colonies, giving each of them a different character makes sense, where it doesn't if I have a hundred.

[sven]

SOTS, and, to a lesser extent, MOO2 both have a bit of a "planet cloning" problem

I think Kerberos quite deliberately abstracted planetary development and resource production in SotS to allow focus on the ships which, it sounds like, is your goal also. Although it seems you are also deliberately avoiding a game with hundreds of planets. If I never have more than a few dozen colonies, giving each of them a different character makes sense, where it doesn't if I have a hundred.

Yeah, as I think I said on the SpaceSector forums -- while I am aiming for a simplified strategic game, what I should perhaps really say is "simplified relative to MOO2 or Civ". The gameplay experience I'm aiming for has a relatively heavy strategic management side, if compared to truly stripped-down games like MOO1 or SOTS. So while the planet improvement system is far simpler than MOO2s, the multi-species population management system is more complex than SOTSs -- and the resource model is headed towards some sort of HOMM/MOO2/SOTS mashup -- which, again, I imagine will end up needing less micro than MOO2, but, certainly more than SOTS.

[sven]

Good game design is a very complicated business and changing something that works to something else without clear reason is simply asking for trouble in the long run. Personally, I am a proponent of simple design patterns. Any complexity just for the sake of complexity/realism etc. is something that should be avoided at all costs.

Well, yes, but I wouldn't quite say "at all costs". I think you can get in trouble with either extreme -- I tend to quote Niklas Jansson on this one. "If I have the choice between simplified, simplified and somewhat realistic, realistic, I choose the middle one." This is a game about building a vast interstellar empire, and one wants to give appropriate weight to that idea, while not being afraid to abstract for the purposes of better gameplay. That means we end up with a lot of messy compromises and awkward middle grounds. For example, limiting each star to just a single planet feels fairly ridiculous, from a world building perspective. But allowing unlimited planets in any system would create a lot of problems for both the UI and the game balance. So we allow 0-4 planets per system. Is it as abstract as it could be? Nope. Is it realistic? Nope. It's a compromise. But a fairly good one, I think. One that lets us do things like capture the flavor of the distinction between a blue-giant and red-dwarf system, without actually going so far as to try and model those differences correctly.

[luciderous]

Well, yes, but I wouldn't quite say "at all costs". I think you can get in trouble with either extreme -- I tend to quote Niklas Jansson on this one.

I agree that going into extremes is not a good idea, and yet it seems that it is always better to 'cut' rather than to 'add', design-wise. Many would-be good games fell prey to the dreaded feature creep and trying to improve things that weren't broken in the first place. One such example that I often point at is FreeOrion project - a game that started like a dream to become the ultimate MoO successor, yet turned into a haywire of over-engineered design concepts that resulted in an almost unplayable and dull experience (IMHO, of course). Sometimes, less is more, and the principle of Occam's Razor remains true to this day as well

Going back to the case with SiS, I'm not really opposed to new ideas, and getting the middle ground between simplicity and realism. These are definitely fine and dandy. The problem is that not everything that looks good on paper will turn out to be quite playable afterwards. And going forward with new ideas is a bold and risky endeavor - something that should be given a lot of thought and debate prior to implementation.

[Gyrfalcon]

Yeah, as I think I said on the SpaceSector forums -- while I am aiming for a simplified strategic game, what I should perhaps really say is "simplified relative to MOO2 or Civ". The gameplay experience I'm aiming for has a relatively heavy strategic management side, if compared to truly stripped-down games like MOO1 or SOTS.

In which case, I will submit my request for some tools to manage late-game micro-management bloat and leave it at that.

Many would-be good games fell prey to the dreaded feature creep and trying to improve things that weren't broken in the first place.

I'd add SotS II to that list as well.

[Arioch]

We definitely need a production queue of some sort.

[wminsing]

Coming at this from the other direction, based on the 'homogeneous empire' comments. One thing that MOOII (and nearly every other space 4x I've played) doesn't really cover is colonies set up purely as resource extraction operations. The default model is you settle a world, and then want to build it up to the maximum population/building size/terraformed state, etc, since that maximizes the planet's efficiency, and since each planet is fairly 'atomic' within the production system that's the most desirable outcome.

However, if humans were to settle Titan, for example, it wouldn't be to turn it into a 'second Earth', it would be slurp up all those lovely hydrocarbons and ship them elsewhere. Ditto for using Saturn itself as a He3 extraction operation. Real-life colonies were often established primarily to gather some useful resource and ship it home, and a broad-based civilian settlement happens later, if ever. So a more 'realistic' model might be that colonies are set up, and then can ship their production (via Transports?) to another planet that is actually building something. Civilization V dabbles in this idea with allowing Caravans to move production and food internally within your empire. So you'd have a network of the homeworld (and larger self-sufficient colonies) being 'fed' by smaller colonies.*

This has a couple of advantages; it makes colonies on 'marginal' worlds more valuable, since even though they are never going to cost-effective to terraform that's not what you're really interested in anyway. It simplifies the management of small colonies (turn on 'export production' and leave them running) and saves any micro-management for the larger and more important worlds. It makes blockading smaller worlds to starve the main production centers of resources a viable strategy. It might (depending on where you take your resource model) give Gas Giants a valuable purpose.

Anyway, just a suggestion.

*The Only 4X game that does this currently that I'm aware of was Starships! , where major colonies were *really* rare (2 or 3 per player including the homeworld was typical) and everything else was assumed to be some sort of mining operation. The freighters appeared on the map and could be attacked, meaning that most early wars were primarily battles of commerce destruction/protection. Might be an extreme model, but when I first played it was a welcome change from the endless planetary assaults of MOOII.

-Will

[Arioch]

Another way to do it is to have additional resource types, such as food and metals, and associated infrastructure (farms and mines), so that you can have specialized agricultural or mining colonies. You can also have special resources, similar to those in Civ III-V, that could make otherwise marginal worlds world putting a small colony on. Food and metal resources could be transportable, but I don't think factory production points are really something that should be transferable. Internal distribution of resources could probably be handled automatically by the trade pool freighters (in a manner similar to MOO2), and foreign trade routes could be something that are manually set up (perhaps being built around the special resources).

These are changes that we plan to make to the strategic model in the future.

[Dinkelsen]

About the fleet composition thing, maybe a "pull" mechanic would be easier to use. At the moment there are lots of planets individually ordered to build ships which are then individually "pushed" into a fleet. But a fleet could know what ships it needs and could be ordered to automatically "pull" them from the industry (like placing an order at a planet that is set to "build ships") or a pool of ships.

The ship would then join the fleet if finished and would move to join the fleet physically. I believe there are more planets than fleets so this could cut down clicking cost.

[Gyrfalcon]

Thinking about this, and having played several more games, I'd put a production queue at the top of my requested features list. Having to go back every turn for thing like refitting destroyers gets quite tedious and it'd be lovely to set up a new colony with its first few infrastructure projects and be able to mostly forget it until it's ready to make stuff.

More than that, I'd really like to be able to queue up a fleet and then get some kind of notification when it is ready. Although that loops back around to wanting to be able to create a fleet with multiple planets building its ships and having them all land at one place when they're done. But being able to stick a flag at the bottom of a stack of ships so I get a ping when they're ready would still be useful even if I do have to request the whole thing at just one planet.

[Arioch]

I agree that a planetary production queue is definitely a must-have, and is included in our plans. Being able to set rally points is also a pretty standard feature that should be relatively straightforward to implement.

We would eventually like to have higher-level automation options (such as planetary governors), but that's a bit farther down the road.