On the research front, rather than just pumping money/resources into one of four fields, I''d be inclined to design it so that research comes in two classes: short term or applied research towards a well defined technological goal - eg hydrogen fuel cells, reactive camoflage - where the player has a fair idea what sort of technology will come out at the end, and what sort of cost the research will have; and long term, pure or "blue sky" research into something with no obvious application - eg string theory - but which will unlock a lot of applied reasearch topics. Depending on how "pure" the pure research is, there may be more or less of a hint as to what it''ll be useful for, the costs and duration of the research may be more or less accurate, and the pay-off may be more or less assured. Depending on how micromanaged you want the research to be (and how much detail you want to put into the research tree) you can either let the player determine which blue-sky projects get funding, and just play around with the time taken to produce results, or just let the player allocate a certain amount to "blue-sky" research and then give him reports on what''s been discovered (or at least what applied research has been unlocked).

Once an applied research project has been unlocked, then the player can pick it by name, allocate funding specifically, and know pretty much what he''s getting out of it (and how close he is to getting it). Possibly have some applied projects unlocked prematurely - for example, controlled fusion research has been going for a while, without getting any reproducible results (last I heard) and probably needs at least one more breakthrough on the pure theory side to really get going.

Another thing worth considering is that, in real life, universities tend to use their human resources for both research and education simultaneously - while commercial research, though it arguably makes more efficient use of its scientists'' research skills, pretty much assumes that they''re done with education. If you''re looking at education infrastructure as well as research, you probably want to consider the balance between dedicated research sites and universities - particularly since universities tend to be full of grad students who make a reasonable number of the actual breakthroughs, and who tend to explore the space of research projects with high efficiency (probably breadth first) - though in a manner relatively free from direct influence from on high.

I was in the shower this morning thinking about the whole educational system. There are many ways to handle an educational system and a political system. I think about this in terms of MOO (1, 2, 3... all the same) than Civilization, since you actually micromanage in MOO than in Civ.

I think of the different political systems, their implications on business (in Communist systems there really are no businesses) and education. In the US, the government manages the education system at the state level, with a national oversight, but the corporate world actually funds a large amount of education. There are also options of having corporations handle more and more of the educational system.

Then I started thinking about how this would affect the game world/universe. There really is a sweet spot in education management, where corporations invest some and the government invests the remaining amount (basically lower levels of education). The more that the corporate structure handles the education system, the more the corporate structure will take control of the people''s lives. The more items will cost and the more likely it is that other governments will gain either the technology (by buying it from the corporation) or the use of the technology (buying a computer doesn''t mean you can make the CPUs) by buying the product. With greater control over the technology and education you end up with less money actually invested though, so your technology advances will come much slower.

I could literally go on for weeks and weeks on a design for a strategy game at the interstellar level.

Something to consider on unit refits though:

I wasn''t actually thinking of upgrading smaller units. Most militaries have a similar system to ours, where you simply move older equipment ''down the line'' instead of upgrading it. Naval ships are somewhat of an exception and I believe that space vessels would be also. Upgrading fighters and tanks individually is relatively unfeasable. The unit is too small and the time required to replace the equipment is too long for the overall benefit.

You can create first line units, 2nd line units and 3rd line units. This is what I like. You have 1/3rd of your force as 1st line units. These units are the "First to go, last to know" type units. They get the best equipment, go on nearly constant operations, but everyone loves them and the esprit de corps is among the best. They are the ''crack'' troops that are the workhorses of any war effort. Another 1/3rd of your force are 2nd line units. These may be highly trained reserve units or basically the bottom line active units. They usually are given the duty of either following the main force into any situation or simply garrison duty. If a situation is beyond the capabilities of the top 1/3rd, they come in to reinforce the top units. Lastly you have the bottom 1/3rd. They are the National Guard of whatever your nation is. They don''t generally leave wherever they''re stationed. They''re as much an elite police force as a lower quality military force. They have the oldest equipment that''s still in service (and servicable). They are as much in place to maintain the peace among the populace as they are as a backup, reserve fighting force.

Back at Tom.

I wasn''t really talking about experience of individuals, I was talking about experience of the force overall. The better trained your military, the more effective they are. Not just pilot Jim Johnson, I mean the entire force.

Dauntless.

Yeah, the education thing is completely the reason why Toyota is still building cars here instead of going to Korea (save MAJOR money on construction) or Mexico. A friend of mine works on Robots and he told me about a production line that he helped put into place and that was exactly what happened... manufacturers were so happy with NAFTA that they started moving production only to find out that a year later they were doing everything back in their US factories again.

The problem in the US isn''t the education system (though my children are being homeschooled), it''s the lack of caring by parents and teachers. The system itself is just fine. The fact that parents are under the impression that their kids don''t have to start learning to read until they''re actually 5 years old and in school. Much more responsibility is on the parents than is on the educational system, though the infrastructure must be there to take children who''s parents are irresponsible much further than they would otherwise go.

If you don't have the ability to upgrade units (or if the upgrade doesn't pay for itself), then it is implied that you need added resources to build new units as needed. There are two constituents to this: (1) resource gathering equipment and refinement facilities must be upgradeable, and (2) the map must have way more resources than is typical in strategy games. Alternately, there must be an extremely efficient and cost-effective way to recycle obsolete units.

Here's the thing: in real life, yes, it is very difficult to upgrade existing units. But in a game world, in which units are specifically designed to be customized and upgraded -- well, all bets are off. I'm already pushing the envelope by invoking a compatibility cost into upgrading a unit (10 percent of the chassis cost per level of difference, where 'level' is an arbitrary value that measures the technological difference between components). Technically, in a game like mine, there should be zero cost to upgrade.

solinear: By experience of the force overall, are you talking about general unit upgrades, e.g., Age of Empires, in which each stage of research raises a unit's abilities? (E.g., Soldier -> Veteran -> Champion.) I don't understand how you would implement this in a game, other than by starting at the educational level. Even in that case, it still comes down to the individual unit, because units already in the field did not receive the same level of training as new units. Unless you have required training programs in which every soldier must participate regardless of active status (which means they'd be pulled out of a fight just to go back to school), you need to keep track of an individual unit's ability.

I agree, it should be possible to start units off with better training. Lots of strategy games implement this. Still, even the most highly-trained green soldiers will not compare to a veteran of two or more wars. Your survivors will be your most valuable units. They'll probably go in the 2nd line.

rmsgrey: That's not a bad idea if you know what you're looking for, but in general you can't specify something that is leagues above what is already known -- e.g., Einstein could not have researched anything leading up to quantum computers because the word "quantum" did not exist back then, let alone computers.

However, I have considered this, and in my particular case the only plausible solution is to allow research in sub-fields. For example, a subfield of infrastructure would be transit; research in transit would yield more efficient means of getting around, although you wouldn't specifically know what sort of upgrades you'd be getting. Military would have a subfield of weapons. And so on.

Anyway, I logged back on to share the pattern I've developed for my tech tree. (There's only one because research is a holistic process, and fields are synergistic to other fields.)

Technology
v
Products
v
Upgrades

(Technology is where that 'level' attribute comes from.) You can assume Products represents columns and Upgrades represents rows. When you discover a technology, you don't necessarily have to discover all the products and upgrades to access the next technology. . . you can access it at once and skip right to the next state of products.

It's important to note that future products are not designed to replace past products. If you want a better version of a product, that's what upgrades are for. The arms race ideology is based on the fact that units have programmed obsolescence, and that future units will take their place. In most games, existing units receive the upgrades automatically. For games in which they don't, you have to bring the old fleet out of service and replace it with a new one. (Or you can keep the old units as backup, as we've already discussed -- but the upkeep costs could end up being quite horrendous.)

[edited by - Tom on April 4, 2003 12:43:10 PM]

Something else I''d like to mention. The concept of military hand-me-downs is a good one (as far as gaming is concerned). When I was at BMT in the Air Force, our flight spent an entire Saturday afternoon clearing furniture and crap out of a dorm so it could be refitted for training. I''m not sure exactly what happened to all that furniture -- some of it was pretty jacked after we got done hauling it -- but I do know there was a huge stack of battered desks that was being sent to the Marines.

Anyway, let''s analyze this upgrade thing from a slightly different perspective. Why would you want to upgrade a unit? So your forces are better armed and armored. Is it practical to do so? This is based on two factors: the cost to upgrade (in both time and resources) and how often that unit is likely to engage in combat. For garrison units, it is entirely sensible to leave them at home without bothering to upgrade them. (Incidentally, this leaves you with more resources to put in other departments.) I did this all the time in Stars!.

A case in which you would definitely want to upgrade is when your entire front line is getting trashed, and you need to send in reinforcements. Refitting your old units will probably be far cheaper than building all new ones (although there is the logical possibility that your old chasses may be incompatible with newer technology -- say, technology that is more than three rows away on the tech tree). You need a quick way to soup up these units before you send them in, because you know that in their current state they would certainly be wiped out.

While I''m on the subject of chasses and such, I''d like to mention that I''ve made engines a component again, but transport modules (either for personnel or cargo) are built into chasses, so there are indeed specific transport-type chasses that cannot be reasonably converted into combat units.

Another good use for "second-line" units is to send them on missions of low priority. In these cases, the mere presence of military power may be enough to ward off a planned attack. I''m thinking of such things as cargo transport, non-VIP personnel transport, scouting unexplored terrain, defending a non-critical installation, etc.

Tom-
If you have the computer sometimes decide what upgrades to do, you might want to consider the "keeping up with the Jones''s" factor. If your spies or field commanders observe a new unit with a device which your side does not have the capability to build, that''s usually a signal to the government that they want that kind of thing too.

Actually this brings up a whole slew of related information gathering aspects that need to be considered. In traditional RTS games, players already know what all the units can do by reading the rulebook. However, in a game where you can design your own units, no one''s going to know exactly what they can do...however, some base guidelines should be known simply for identification purposes. For example, if you see a huge armored vehicle with a turret and a long barrel sticking out of that turret, obviously it will be identified as a tank. But something like a Bradley may be a bit more sticky...it could be a light tank, but the small gun size kind of gives it away. However, until it is observed to deploy infantry, you won''t know it has that capability. And if you do upgrades, it''s possible that you may fool your opponent into thinking the unit does something else. For example, you may decide to upgrade your Bradley chassis so that it''s basically a light reconnaissance tank by dropping the infantry cargo and stocking up on more ammo or maybe a reloadable TOW launcher (actually the US Army does something similar to this with the M3 version of the Bradley as opposed to the M2). The enemy player, seeing the Bradley chassis goes, "aha, he''s going to dismount his troops and use them in concert to attack me", and yet the capabilities of the new Bradley will be different.

Getting back to the specialization of research, it sounds like a good idea. I wouldn''t want to specialize the scientists too much as that could get into micromanagement hell. Having the research fields is a good idea too, and if you combine that with what rmsgrey suggested I think it could be very potent.

quote:

---

Original post by Tom
rmsgrey: That''s not a bad idea if you know what you''re looking for, but in general you can''t specify something that is leagues above what is already known -- e.g., Einstein could not have researched anything leading up to quantum computers because the word "quantum" did not exist back then, let alone computers.

---

Sorry, I may have been babbling a little too much for my ideas to get through clearly. I was thinking of a system where Einstein isn''t in a position to research Quantum Computers because that particular applied research topic hadn''t been unlocked yet. But people contemporary with Einstein could do research into the applied topic "Atomic Bombs" because the "pure" research needed in order to unlock it had been done.

Player sets some amount of budget aside for "pure" research. Some time later, a report comes back saying that a breakthrough has been made in, say, atomic theory, and half a dozen eminent scientists believe it''s now possible to build an atomic bomb. "Atomic bomb" appears on the list of applied projects and the player can decide to divert some funds into developing one.

Essentially, the tech tree is largely hidden from the player, and he has little or no control over the order in which applied projects become available (maybe allow him to split his "pure" research funds into general areas, but keep a random element involved) but, in addition to updates on the general progress of scientific research, he gets regular proposals for specific projects with immediate applications and a good chance of success (possibly also projects that aren''t ready yet - like modern day cold fusion or VR, the one needing at least one more breakthrough, the other needing more incremental improvements in the underlying technology to become economically and ergonomically feasible). So the player has complete control over what sort of hardware becomes available next, provided he keeps up the funding to the pure science running in the background making the new discoveries possible.

rmsgrey: Yes, that''s exactly what I had in mind. Looks like we''re in the same boat.

I started playing Startopia again yesterday, and I''d forgotten that it uses a research system very similar to this. But in Startopia, research starts at the most basic level: with resources themselves. Here''s basically how it works:

You can buy any goods from trade vessels as long as you have a stardock (or from Arona, the default trader, if you want to pay horrendous prices -- but he usually has stuff the other guys don''t). Once you''ve purchased an item, you can either (a) build it, or (b) donate it to your labs for research. When the scientists have analyzed the item, they add the hardplans to your database, and from that point on you can build the item yourself (provided you have a factory).

What''s cool is that you can also work from the bottom by analyzing goods. Each time the scientists make a breakthrough, they unlock the first item(s) associated with those goods. (For example, researching garbage unlocks "litter bin.") When you research the next item in the chain, you open up more items. It keeps going until you reach the end of the chain.

This opens up a whole new level of strategy. Your combat objective to this point has been single-minded annihilation of enemy forces. Now there''s a new objective: capture their technology so you can study it. (You might also be able to study wreckage, but the information garnered from it would be incomplete.)

Of course, studying enemy technology doesn''t necessarily mean you know how it works. (In the words of Arthur C. Clarke, "Any sufficiently advanced form of technology is indistinguishable from magic.") But you can probably get a very good idea of what sort of research you need to do to unlock it. So, you''d get a list of fields you need to complete before you can build the item. Because of the invisible nature of the technology chart, you won''t know exactly when those fields will pop up, but when you start seeing them, you''ll know immediately that that''s where your funds should be going.

There are probably better ways to implement the "study captured technology" idea. I''d like to hear them.

Something else I should mention about Startopia: even when they weren''t specifically analyzing something, your scientists were constantly searching for breakthroughs. You''d get regular messages that they upgraded an existing item, which I think makes the item cheaper or more efficient, something like that. Labs that don''t specifically have research orders could randomly (or logically) pick a technology and try to improve it using their basal energy allowance. Upgrades in this manner would be slow, but definite.