|
Author
|
Topic: A few thoughts about science output and computer’s AI. Current Date/Time:
April 08, 2000 21:16:51
|
IVANMV
Settler
From: Russia
Registered: Mar 2000
|
 |
posted March 13, 2000 13:41
   
|
 |
|
I bet that most computer fans will agree with me that the best wining strategy in strategic games like civilization is to build or capture as much cities as possible. The game algorithm supposes that the number of cities and population you control generates science so that whoever builds more cities consequently generates more science and wins the game. The most annoying thing for me is that the computer followed by the mentioned algorithm builds cities everywhere without stopping. (That’s the only way to survive and win). AI’s strategies become pretty irritating and very often spoils the game play because as countermeasure by the end of the game you usually find yourself managing a huge empire with hundred of cities and units. One of the solutions to this situation may be a new algorithm.I believe that the current algorithm (cities and population give science) is not logically correct. I suggest using money instead. If we look into the real world and compare for example Japan or USA with India we can conclude that the number of population does not necessary mean the great scientific output. (The number of population of India is two times more that both in the USA and Japan) A strong economy is what that makes countries to invest in education and science and consequently generates scientific breakthroughs. Christopher Columbus wouldn’t discover American continent without Spanish king’s investment. In the game with such an algorithm we can control a small territory but still rule the world. Like in the real world strong economics will determine the winner and the loser. The new simple science formula could be something like this: Technological breakthrough (game turns left) = tech factor / money invested in current project. Money invested is the amount of money invested in the scientific project. (In the ancient times before the invention of money food can be used instead.) Tech factor is a predefined constant assigned to technologies. Like in the real world huge investments do not necessary mean the immediate success in research. The more difficult the invention the more fundamental research is needed and the more time should pass before your scientists can discover anything. The offered algorithm may give us very interesting strategies to play. Money income should be changed and balanced accordingly to match a new algorithm. The idea is that a small but economically efficient nation should generate the same amount of science as a huge but economically poor empire. The bigger does not have to mean the better (moreover big empires require more bureaucrats increasing the costs and corruption level.) If well balanced this new algorithm can suggest many new strategies and change AI’s behavior as well. That’s the idea, I wonder what do you think?
|
VetLegion
Warlord
From:
Registered: Sep 1999
|
 |
posted March 13, 2000 18:55
|
|
|
Agreed. I have some questions for you:Since more population means more tiles worked, which means more money...obviously amount of trade/money should not be connected to terrain. How should trade be represented then? Perhaps a city should act as a gravitational center, with invisible "roads" from other cities and resources (mines, willages, oil platforms) streaming through it. Then these roads are calculated (summed) and we have number of arrows, or something. Tech model looks good, parts of it are already implemented in other civ games like CTP (money for science). |
Yuvo
Settler
From: Australia
Registered: Feb 2000
|
 |
posted March 14, 2000 00:56
|
|
|
I agree too.But when you look at the real world system, there's really only one way to go: Instead of the Taxes/Science/Luxuries trade division, there should be a simple "budget", where you allocate taxes earnt to science, luxuries etc., with sliders and %ages. For example you might have a 20% tax rate, which might give you 120 gold each turn(if you include trade route income). Then you might decide to spend 30% of it on science, 30% on luxuries, 20% on health&education and also 15 gold on improvements upkeep, leaving you 9 gold surplus. This system might seem complicated compared to the previous one, but it's still pretty simple. Obviously IVANMV's algorithm could be put into this.
[This message has been edited by Yuvo (edited March 14, 2000).] |
Dienstag
Chieftain
From: Sinking Spring, PA, USA
Registered: Feb 2000
|
|
posted March 14, 2000 05:21
|
|
|
Good ideas, guys.IVANMV-
I'm not sure about this, but I got the impression somewhere that India has had and continues to have a fair amount of really respected scientists. I've seen lists of how countries currently rank in terms of scientific output and the U.S. and Japan were at the top of the list (I forget I saw this or exactly what the order was). I wouldn't be surprised if India was pretty high up there (didn't they just build their own nuclear weapons from scratch?). So I don't know if the India example was all that good. However, there is the Civ II feature of building libraries, universities, and research centers. Higher populations help these work better, but you need the good economy to get them in the first place. Could this count as money/food indirectly helping research, instead of making direct payments for techs? I don't feel that a really rich Civ with no universities or reasearch labs should be able to "buy" advanced technology. VetLegion-
Your "invisible roads" idea sounds interesting, but I'm not sure I understand. Would you give an example? Yuvo-
I totally agree with you on the "budget" idea. Good example, and I for one do not think this idea is too complicated. I'm surprised no one suggested this before, as far as I know. It would help Civ III a lot. Dienstag |
korn469
Prince
From: the dark wastelands of apathy
Registered: Apr 99
|
|
posted March 14, 2000 05:39
|
|
|
i think the best way to seperate research and city numbers would be to make research specialist very important to a civs researchto do this would require a few changes - specialist consume three food
- they give high amounts of research points
- science buildings should give specialist maybe twice the benefit (so for example instead of +50% to research they get +100%)
- make their be a ratio of specialist to workers until a city gets to be a really large size
i think that greatly emphasizing specialists would decrease the reliance on hundreds of cities and would increase the need to have large well developed cities to support your scientific output The OpenCiv3 website
korn469
[This message has been edited by korn469 (edited March 14, 2000).] |
VetLegion
Warlord
From:
Registered: Sep 1999
|
|
posted March 15, 2000 18:53
|
|
|
Dienstag
Yes, I like this idea too, but it is not originaly mine. Someone made it in detail and it was posted on these forums, while origination on sidgames forums. I remember that it was quite elaborated, but I cannot remember in which thread
 |
The Joker
Warlord
From: Copenhagen, Denmark
Registered: Aug 1999
|
|
posted March 16, 2000 09:15
|
|
|
I agree with korns idea.This problem is almost exactly what we have been discussing at the ICS thread (check it out). There is a lot of great ideas about food output, base squares, bureaucracy points etc. |
tniem
Chieftain
From: GR, MI USA
Registered: Apr 2000
|
|
posted April 08, 2000 21:16
|
|
|
What if after getting research points you then had to spend them. For instance, every child needs an education in order to gain a tech and so the bigger your empire the more tech points are used for teaching. There could be other items that you have to pay for using your research points and so a larger empire would not be ideal in gaining research. A larger empire still may be the best at gaining research, it just would be harder. I know the idea needs work but any comments. |
Apolyton Civilization Site Forums
