Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[praguepride]

Ah, now you're just splitting hairs. That's like saying that "every game" is a management game because you have to "manage" your fingers inputting commands.

Obviously, to you, this is a management game. To me, this is more of a chaotic simulation. Sure I can try to tell the dwarves what to do, but in the end they do what they do. They wall themselves in, they "flee" right into the path of enemies, carp kill everyone... I wouldn't consider this a management game like civilizations or railroad tycoon because half the time the stupid dwarves don't do what I tell them to, AND I LOVE IT!

This game is COMPLETELY different from Civilizations, but there are similar traits. Just like this game is also similar to the Sims (keeping people happy), XXX Tycoon (there's money involved, you buy and sell things to buy and sell better things), Command & Conquer (Pausable RTS), Exploration Games (I can't think of any off the top of my head but this would involve the joys of digging to find what is beneath you), Little Big World (or any freeform "design your own environment" game), Incredible Machines (whether it be just linking pumps and power to huge elaborate death traps), Dungeon Keeper (see above about huge elaborate death traps), Oregon Trail (picking your starting equipment can drastically alter the game), Final Fantasy (grind XP to become more powerful), Fallout (no "level balancing" garbage. When you first embark you could end up getting completely wiped out by a random monster), Total War (building and raising armies), The "I forgot what it's called but the game where you basically build a castle to defend against enemy seiges game, it had a couple of expansions including a Crusades one. You had to manage your food supply, build weapons, and build walls to defend against the enemy. Each stage would set you up at your starting point and you'd have to scramble to build a castle in X time before the enemy waves start coming"....or Lords of the Realm series have similar concepts (Defending seiges).

You say that the whole point of the game is micromanaging your minions, while I could be just as justified in saying that the point of this game is to build up your defenses to repel enemy invasions. Which would mean that this is an RTS, not a "management" game.

And if you respond with "but RTS' are management games..." then you're just going to sound whiney . The focus is on ECONOMY, not just generic management.

[Squirrelloid]

First of all, I can think of games that involve no management, or little enough management that they aren't defining aspects of gameplay.  A generic FPS is such a game.  Fallout 2 isn't really a management game, as the few management options (ally AI settings) aren't a major gameplay component.  Heck, I can even think of a digging game that isn't really a management game.  Ok, i can't remember its name - its a flash game about digging on Mars for precious resources.

You say that the whole point of the game is micromanaging your minions, while I could be just as justified in saying that the point of this game is to build up your defenses to repel enemy invasions. Which would mean that this is an RTS, not a "management" game.

And if you respond with "but RTS' are management games..." then you're just going to sound whiney . The focus is on ECONOMY, not just generic management.

Why do people like poisoning the well so much?

RTS games are very obviously both management games and *economic* games.  Anyone who is any good at, say, Starcraft, understands that.  In fact, you can't be good at the game without good resource management.  How fast should you expand?  How many resources should you devote to units vs. research/new buildings?  These are all economic questions.  Resources are most definitely limited, and most games between serious players are won by who controls the most resources.  I know there are sites where you watch replays by 'professional' Starcraft players - go watch some.  Its very educational.

[praguepride]

 

And he goes there... Sure "some" RTS' like Starcraft are "management" oriented, but others aren't. Dawn of War II comes to mind in that there is no base building and no economy.

ANYWAY, back to the topic at hand. We've gotten rather far off tangent and I feel I should do my part as the inflammatory troll to bring us back to the topic at hand:

Does DF need to accurately model a global S/D economy:

You say yes (obviously)

I say "no", you can fake it by just making it random.

Here's my reason:

Given that the computer has complete control of everything outside of the player's realm, everything is still going to be randomized. If you do the global S/D model then either it will be predictably static (boring!) or there has to be some element of randomness.

Now, outside of the player's realm of vision, some randomness = total randomness. Whether a price goes up because of a RNG function or it goes up because a drought followed a war followed a horrible leader RNG function...the end is the same: It's random!

So again, instead of wasting months (yes! I again assertain that it would take months of tweaking code and testing enviornments repeatedly) to model things that the player won't be able to see anyway, why not just cut out the middle men.

So instead of

RANDOM -> EVENT -> EVENT -> EVENT -> Price Change -> Player Impact

Where each chain has to be modeled, controlled, and analyzed (and consume limited computer resources).

Why not shorten the chain: RANDOM -> Price Change -> Player Impact

You can even reverse the chain a bit so that you can have your "player impacting world". Again as soon as it goes off screen it disappears from system view, but basically you can do your market manipulations (dumping mass products, restricting rare products) and it'll have an Impact of X*RNG. So maybe you'll alter the market in a huge way, maybe nobody will even blink an eye.

[Granite26]

What are you all even bickering about?

I think we've gotten 5 steps from anything even remotely important:

First : Predatory pricing was a bad example.
Second : S/D does have weirdnesses over short periods of time.  That's fallout and imperfect information, not a failure of people to pay at most what something is worth to them, or sell it for at least what they could get for it someplace else.  These failures are covered by individuals, personalities, and skills (I.E. Haggle skill and 'makes rash decisions), not by the process itself.
Third : It doesn't matter what 'type' of game DF is.  'Types' are artificial circles drawn around games products so that advertisers can more effectively target markets, and social cliques can more effectively be whiny little bitches (That's not REAL industrial, it's just techno or some bull like that). 
Fourth : DF does have a layer of economic simulation, making it a valid target for suggestions.  You can disagree with the content or complexity, however
Fifth : S/D is real, it is factual.  The arguments that have been brought up serve to justify government controls around it, not deny it's exisitance, much like handrails protect from gravity, not prove gravity doesn't exist.

Spoiler (click to show/hide)

Unless I'm very much mistaken, dumping is different:  If it costs me 100$ to make 100 widgets, and I can sell 100 widgets for 150$ a piece, that is good.  If I can make 200 widgets for 90$ a widget(efficiencies of scale), as long as I can 'dump' the 100 excess widgets overseas in third world markets for 81$ a piece, I've made money.  This is considered unfair, because the overseas producers are unable to make widgets for 81$, they cost them 90$ to make same as the original practicioner

Can we get back on topic now?

[Squirrelloid]

The difference between random and simulating events is that the latter has *persistence* and *logical effects* rather than being totally arbitrary.

Persistence is the more important.  For instance, merchants requesting totally different item types every time is bizarre.  Unfulfilled requests should remain unfulfilled for some period of time, and be repeated.

Further, one of the things i've been advocating is different persistent economies with each world.  Ie, the price of various goods will not be the same when you first embark depending on the nature of the world you embark in.  While this is obviously initially somewhat random (at worldgen, and with settlement placement), that particular random seed would dictate the basic nature of the world economy persistently from year 1 onwards.  This makes different worlds behave differently, something that 'random at merchant generation' cannot model.  A world with 100% volcanic activity would have much less sedimentary rock, and therefore less iron, so the price of iron in such a world should be relatively higher to one with lots of sedimentary rock.

Basically, the persistence of a given random event is the key differentiating factor.  You can't fake a more persistent system with a less persistent random system. 

We can also talk about persistence across caravans.  If a randomly generated famine event happens, it may be that *all* caravans will offer you more for foodstuffs, not just the human caravan.  Again, not modeled by a 'random at merchant generation' system.  Similarly, a persistent event could lead to persistent demand.  Wars which last multiple years would create additional demand for arms and armor in all of those years.

Second, the benefit of modelling the causes leads to rational expectation which leads to informed player decision-making.  Part of this is about immersion.  When the caravan tells us that there's a war going on, the fact that they're willing to pay more for arms and armor becomes unsurprising and logical.  Players may reach other conclusions about future demand without needing to be told.  If the war persists through harvest time or planting time, individuals who would otherwise have farmed fields will be engaged in fighting and thus demand for food may go up for next year, so the player might plant more crops than usual.

Basically, there is a substantial difference in the nature of the output, and thus modeling the system is worthwhile and interesting.

[praguepride]

Again the question is, why should starting economies be random by default.

Why, playing a vanilla game, would I want a game where Iron is 10x the price of platinum. Why not make environmental factors randomized too so I could start a vanilla game and all my dwarves suffocate because the oxygen levels were set too low.

There needs to be some solidity and if you want a bizarro world where Platinum is the new Iron, edit the RAWS, that's what they're there for.

And don't get me wrong, I'm not saying that everything should be completely random 100% of the time. But instead of building a proper S/D model, you can fudge it. You can have your persistence but at it's core its based off a random seed, so why not just use the raw game engine for 90% of the work, and then in that last little 10% that the player will actually see, then you can flesh out the events/persistence/whatever.

[Squirrelloid]

Lets just say that I agree with all 5 of your points Granite

(although I think the genre debate was mildly meaningful - and I wasn't looking at it as a genre thing so much as a 'level at which control is exerted' concept, but lets move on).


(For completeness, i'd like to comment on your response to dumping)

Spoiler (click to show/hide)

Unless I'm very much mistaken, dumping is different:  If it costs me 100$ to make 100 widgets, and I can sell 100 widgets for 150$ a piece, that is good.  If I can make 200 widgets for 90$ a widget(efficiencies of scale), as long as I can 'dump' the 100 excess widgets overseas in third world markets for 81$ a piece, I've made money.  This is considered unfair, because the overseas producers are unable to make widgets for 81$, they cost them 90$ to make same as the original practicioner

Can we get back on topic now?

Spoiler (click to show/hide)

The correct response of companies in the 'victimized' market is too increase their production and 'dump' in the home market of the victimizer.  Eventually prices go up in both markets to the reasonable price of $90+(shipping cost to farthest market) for a widget. The correct response is not to go whining to your government 'boohoo, they're stopping me from earning excess profits!'  Dumping is a scare word used to discourage competition via differentiating markets, a perfectly reasonable thing to do, and legal in most other contexts.

[praguepride]

This is a neat idea! Kudos for granite as being the first to show me how to engage in a private debate while keeping things on topic!

Spoiler (click to show/hide)

Here's a big issue that other companies have with Japan. The Japanese government has had no problems with monopoly and has worked very hard to help it's companies gain market. Almost aggressively so to the point that it puts foreign companies at a disadvantage. Shaky patent laws, red tape, the works which their companies can avoid because they're not foreign.

So what happens when an American company, that is closely regulated so that it "plays fair" and isn't a monoply and has to allow competition to survive is thrown against Japanese companies that don't play by all the rules, who will undercut and "dump" and their government, instead of opposing it like most other governments do, will actually subsidize them to help them gain market dominance.

But you're right. Nothing's been proven because it's not technically illegal. It's wrong, but very difficult to prove in the short term. The long-term effect has been dramatic though. Just 50 years ago America was the manufactoring powerhouse of the world, and now Japan has under cut many American Industries: Cars & Electronics suffering the most. Now, I'm not in either industry so I'm not someone whining that the Japanese stole my job. Nor am I saying that it was a conspiracy, a plot to destroy America, just that it's a different way of doing business. And because their government is backing them instead of fighting them like many other governments do, it's given them a unique ability to exert control into other marketplaces.

[Squirrelloid]

Again the question is, why should starting economies be random by default.

Why, playing a vanilla game, would I want a game where Iron is 10x the price of platinum. Why not make environmental factors randomized too so I could start a vanilla game and all my dwarves suffocate because the oxygen levels were set too low.

There needs to be some solidity and if you want a bizarro world where Platinum is the new Iron, edit the RAWS, that's what they're there for.

And don't get me wrong, I'm not saying that everything should be completely random 100% of the time. But instead of building a proper S/D model, you can fudge it. You can have your persistence but at it's core its based off a random seed, so why not just use the raw game engine for 90% of the work, and then in that last little 10% that the player will actually see, then you can flesh out the events/persistence/whatever.

You seem to be misunderstanding what I'm saying.  I want prices conditional on world gen.  I certainly don't know how much iron vs. platinum there is in any one world, and I shouldn't be expected to.  How can I achieve my concept by altering the raws?  And given a particular ratio of platinum to iron, how do we know what the price should be without some idea for demand for platinum and iron?  You'll note the OP talks about specific resource profiles of settlements, so deposits far from any settlement aren't counted, making the requisite knowledge for RAW modification even harder to acquire in some ad hoc fashion.

Prices predicated on real information is not the same as prices chosen at random out of a hat.  Especially when old player fortresses will be able to become permanent settlements that effect the nature of the S/D model in later games (pretty sure 'abandoned' fortresses continuing as actual settlements is at least on the dev list).

Edit:

Spoiler (click to show/hide)

Here's a big issue that other companies have with Japan. The Japanese government has had no problems with monopoly and has worked very hard to help it's companies gain market. Almost aggressively so to the point that it puts foreign companies at a disadvantage. Shaky patent laws, red tape, the works which their companies can avoid because they're not foreign.

So what happens when an American company, that is closely regulated so that it "plays fair" and isn't a monoply and has to allow competition to survive is thrown against Japanese companies that don't play by all the rules, who will undercut and "dump" and their government, instead of opposing it like most other governments do, will actually subsidize them to help them gain market dominance.

But you're right. Nothing's been proven because it's not technically illegal. It's wrong, but very difficult to prove in the short term. The long-term effect has been dramatic though. Just 50 years ago America was the manufactoring powerhouse of the world, and now Japan has under cut many American Industries: Cars & Electronics suffering the most. Now, I'm not in either industry so I'm not someone whining that the Japanese stole my job. Nor am I saying that it was a conspiracy, a plot to destroy America, just that it's a different way of doing business. And because their government is backing them instead of fighting them like many other governments do, it's given them a unique ability to exert control into other marketplaces.

Spoiler (click to show/hide)

Just because America controls less market share than it used to is not evidence of unfair competition.  American companies may be inefficient competitors.  Certainly the American auto industry is.  They have no one to blame but themselves, for reasons I already covered in another post.  I know less about the history of the other industries, so I can't comment specifically, but it could just be that Japanese companies also offer better products or are more efficient competitors in those industries as well.

Furthermore, government subsidies means taking money from taxpayers and giving it to businesses, which ultimately hurts the Japanese economy in the long run, and is not a sustainable tactic.  I mean, consider your claim.  Japan takes money from its citizens to make cars cheaper in *other* countries?  So not only do the japanese pay for their own car, they pay for some foreigner's car as well?  Not only does this make terrible economic sense (you're killing GNP by reducing national consumption and therefore national demand), but short selling by american companies in Japan should force Japanese companies to cut local prices and be forced to raise prices acrossed the board.

Furthermore, I refer you to the model T evidence in the CATO paper I linked - Ford set a price which then forced his engineers to find a way to make it profitable at that price.  He then set a lower price.  etc... Actually engaging the Japanese in competition rather than whining about it would force American car companies to innovate - which is far better for the consumer (lower prices) than government intervention is (artificially high prices).

[praguepride]

Hmmm... it might be interesting, but then again it might also be a pain. I'm not a DF code expert, but I can see how doing a detailed analysis of each zone might take time during Worldgen, especially if its not needed. If I'm on a little island all by myself I don't care what the rest of the world is doing.

Look at how much time the site finder takes to go through each region piece by piece. If it had to do that to calculate variable prices on resources... PASS! Worldgen takes long enough. If it's a simple extension of an existing calculation, eh! it won't be that bad.

[Granite26]

(switches hat)

Some people think that their role in the game is to make sense of the randomness.  When the human caravan offers you more for arms or armor, the story IS that there's a war on.  There are some obvious weaknesses (why just leather armour?), but it plays out neat on the LPs.

Prague is right about the randomness causing problems.  The same thing comes up with the random creatures.  Sure you can randomize the starting stats of the creatures, but if dwarves aren't tougher and shorter than humans, there'll be some confusion.  Hell, vanilla's use of unseelie for goblins confuses the hell out of modern Tolkienites...

For world economies, diamonds should always be valuable, as should iron.  If iron is too common, it depresses the values of other materials down to nearly worthless because of it's innate superiority.  As for diamonds, it wouldn't make any sense to players if diamonds were common as dirt (and really, they are pretty common in the universe)

One of the advantages to hard setting values is that it allows you to fudge the numbers.  World Gen doesn't have to worry about how many diamonds are in a fort, because the dwarves are incapable of destroying the world market.  How much fun would it be trying to learn an entirely new scale of values with every game?  (Which is worth more?  Copper? Silver? Platinum?  Gold?)  Sure it could be fun if one game Plat was worth 20 * Gold and the next 2.  It just gets hard to play if its ever 1/20.

Spoiler (click to show/hide)

Squirrelloid : That's true, but if the home country (poor third world) doesn't have the capital to last in a long term price war like that, it severely depresses their ability to stop being third world.  Nobody does this to the US or Europe or Japan.  It's Africa and Indonesia and the poorer parts of Asia where they don't have the economic muscle to fight back.  Small third world producers can produce small numbers of units for sale locally, but don't have the contacts or capacity to sell globally.  Dumping prevents them from building the capacity to do so.

Prague:  I haven't followed this in a while.  I *THINK* you're right about the not playing fair.  The problem isn't Japan, however, it's the US letting them break these rules, and letting them play currency games (keeping the yen relatively low).  China is doing the same thing.

Another difference is Unions and Labor practices.  They don't have anywhere near the same labour overhead the big 3 have.

Basically, you can't blame another country for cheating, only your own for letting them...  Economic theory has the solution for someone imposing Tariffs(or other unfair trade practices.)  Unfortunately it's just the prisoner dilema, and we keep letting them rat on us.

[praguepride]

I wouldn't be opposed to minor fluctations, but granite has my support on his last statement. Having to "relearn" the value system for each object in the game would be a pain unless your sole interest is economics.

Having them fluctuate slightly is fine, but completely re-writing the books is a whole other issue.

So maybe platinum is worth ~15-25 times as much as gold, the point is that it's still worth a lot more then gold and I can remember that.

So for those of you who want to challenge yourself to see who can make the most wealth in X period of time, you could carefully monitor the markets to try and maximize your earnings. But the rest of use would see little effect. Platinum > Gold > Silver > Copper. Simple.

[Squirrelloid]

Hmmm... it might be interesting, but then again it might also be a pain. I'm not a DF code expert, but I can see how doing a detailed analysis of each zone might take time during Worldgen, especially if its not needed. If I'm on a little island all by myself I don't care what the rest of the world is doing.

Look at how much time the site finder takes to go through each region piece by piece. If it had to do that to calculate variable prices on resources... PASS! Worldgen takes long enough. If it's a simple extension of an existing calculation, eh! it won't be that bad.

I honestly don't understand why the site finder takes so long to do it, since it doesn't need to look (fortress level) tile by tile.  I suppose the long part might be a packing problem - trying to fit an nxm fortress to include everything is harder than just finding everything.  And I think it may be trying to fit an nxm site on every world tile rather than just those that have all the necessary features (since it takes the same length of time for each tile based on the tile count update frequency), so it has no sorting algorithm, effectively.

If i were writing the site finder code, I'd take world-size matrices (256x256 for large?), and use 1s and 0s for present/absent for each specified criteria.  I'd multiply same-location elements across all matrices for all specified criteria, and then look for 1s (should be really fast).  Heck, I could probably program such a thing in Matlab in ~15 minutes, and it should run in ~5 seconds.  Then you get a list of locations which have all the things you want, and possibly options like 'go to this location' or 'fit an nxm site on this location' - fitting would take longer, whereas those of us who play with everything revealed can just go through the location list and see if we can fit a fortress of a size we want to play on the features we want.

Regardless, since my proposal only involves looking at settlement locations, which the game should be able to find really easily, it should take less time than the site finder currently does because it has to investigate fewer locations.

Edit:
On relative pricing - for most world's this can be controlled by choice of the slope and intercept for demand curves.  Ie, gold has higher innate demand than most things.  Similarly, iron has more possible uses than nickel, so demand for iron should be higher.  You may occasionally get a world where things are *really weird* (how weird depends on how the world gen code works), but those worlds would have to have truly strange resource distributions and would probably be epic. 

[Granite26]

He's talking world gen, not site finder.

Site finder is slightly more complicated because it's checking multiple size squares for multiple features.  I imagine speed improvements could be done by only considering possibles near rare features (magma) but who knows.  There's a topic for this somewhere.

World gen = totally something else

[praguepride]

But that's again assuming that fixing finding codes is one of Toady's priorities.

As it stands, I'd imagine that it would take the same amount of time to calculate resources per civ as it does to find an embark location.

And then there's the question of how many of those resources does a town actually have access to? Do you just auto assume that a town has 100% access to all resources in its area, or do you only do a percentage?

Dwarven settlements should have more access to deeply buried stones and gems then humans, and more so then elves. On the flip side elves should have more access to surface plants (berries and wood) then humans and more so then dwarves.