View Full Version : Years
Apr 22, 2004, 04:22 PM
While the current progression of years worked fine for previous civ games, isn't it time for a more reactive and realistic system. This would also change how long game where.
Here is the idea; the amount of time that passes each turn would be based on tech levels. Why you ask? Cause it doesn't take 50 years(10 years a turn) for a city to build a division of tanks.
To solves this, each turn the computer would use a formula that took the relative tech levels of all the surviving civs and found some kind of average point. This point would be compared to a table that told how many years would pass next turn. Over time the passage of time would go from 50 years to 6 months.
Apr 22, 2004, 06:45 PM
This would of course start a tiny snowball effect: if you start gathering techs early, you will be in modern age in like 1000AD, or earlier. On the other hand, if you start slowly, you will not be able to finish the tech race in time. So this would ease the task of good players to win the space-race, whereas it would decrease the chances of the not-so-good players still more.
Other than that, I see no other problem with this but how the calculations are actually made. This could be the developers' problem though, or could be solved somewhat by as so as to ease the already burdened developers.
Apr 23, 2004, 10:39 AM
To me a turn is a turn. The only time I am aware of the year is when it is closing in on 2050 and I am behind on the histrograph. Other than that, it is irrelevant.
Apr 23, 2004, 02:57 PM
Here is part of how the calculation would be based.
All the civs tech levels would be quanticized somehow, maybe even in multiple dimensions(ie. an A and B number, etc.). These quantities would be arranged on some kind of Deviation curve. Those on the outlying sides, highly advanced or primative civs, would be eliminated. Then all remaining numbers would be weighted, putting more weight on more central quantities. The new weighted qunatities would then be averaged once more and then plugged into a formula.
i have not decided which kind of curve the actual formula should be modeled off of. Linear modeling seemed to ignore the rapid acceleration of the growth of technology over time. Here are some basic ideas
Years Per Turn = Y
Starting Constant = C
natural number or limit of (1+(1/x))^x as x -> infinity = e
Rate Constant = k
Global Tech Level = T
Y = Ce^kt
This model is based on the idea of compound growth at a continuous rate. Year decrease would be slow througout much of the ancient and middle ages, but would really be quick in modern times. Benefits include an easily set upper and lower limit. Unfortunately this regression can easily snowball, making games thousands of turns if you reach the middle-industrial age unusually early.
Y = Years per Turn
C = Constant
T = Global Tech Level
S = Minimum Years per Turn
Y = C/T + S
This model would likely make only the early game go slow. A higher C would make it later until the rate really picked up. Once again, possibility for snowballing.
There are other mathematical models i could consider, but I'm too tired to remember them right now. Antoher factor may include some demographicsm primarily Life span. Maybe the T ratio could determine what fraction of an Average generation would be the years per turn.
Under this model, the average life expectancy(not average of civs, but of all the citizens combined and wieghted for population) could be G.
Y = GT
Where T would go from 1.00 to .05
May 25, 2004, 10:52 PM
woah.. I think I understand what your saying even though I have no idea how that match works out. Firstly though for simplicity I think the turn system should be representative of 1year. With the ability to skip to the next event marker to save the space bar.
However there should definately be some correlation between technology and time it takes to build such as it seems you've suggested.
A civ with the ability to build a tank would certainly indicate that factories are present and thus would be able to churn out a tank or more per turn. (assuming a turn means 1year) This however leads the problem of funding such an expense, both in the purchase and the running, not just in access to resources or shields to build. Once factories are present shield should largely be unused as a method to calculating the ability to build quickly. If a factory is constructed it should determain the speed of production based on availiable funds.
((Also Individual cities shouldn't care that one or more units are in the city or out in the field, but what its costing to keep them on active duty. If I want a large army as defence I should have it without the local populous thinking that I want to conquer the world. If a the start I have a spending habit of 80% on defence, then at the end of the game the populous should be used to that level.))
The pregressive years should also effect the way Settler unit acts. A settler built in 2000BC would have the technology to build a primitive town based on the parent city current tech level and slowly grow. A settler unit in the 1800's has more tools and greater access to resources and populous, and have less logistical issues. So the growth of the city would be more accellerated be based on access to those things which make them grow. A Civ that has the attributes of expansism should benefit from that attribute and continue to take advantage of it thoughout the game. Even provide government insentives for populous to move to a new city on there own. The ability to build a library, should not take 80 turns in the 1800's; it should take maybe 5 or 10 turns (turns being 1 year each) regardless of the city size or access to shields. The story would be that WE provide the city with the resources/labour to build the library.
Jul 08, 2004, 12:10 PM
I would love for the game to simply work on a one turn, one year basis in single player, but it would obviously expand the game into way too many turns.
It has never made sense to me for it to take millenia for a troop of warriors to go over a mountain range and return, though...so maybe we should leave the turns as they stand, but base unit move rates, build rates, and population growth on years, rather than turns.
Jul 08, 2004, 02:10 PM
I disagree. A turn should be an arbitrary amount of time. The way this game is there won't be a perfect mapping of a turn to unit of time or algorithm.
Jul 08, 2004, 02:30 PM
Although the nay-sayers to this idea are usually suggesting that the year marker does not really count, that is incorrect. The amount of time that a culture-producing building has existed (in years, not turns) matters a great deal! After 1000 years, the culture doubles. In addition, if the building is a Great Wonder, then the building produces 2 coins for each 1000 years it has existed. The year marker is not arbitrary.
I do agree that something needs to be done about the year system, but I am not sure if the above system will work the way sir schwick intends.
Jul 09, 2004, 05:20 AM
I think if we want to see the years to more or less reflect real world history, there could be a different setting of the calendar for each difficulty level.
I think it is not too hard to calculate an "average" speed of development for the AI on each level - the turn/year formula could be arranged according to that - have one scheme for all levels: from chieftain up to Sid! :)
Jul 09, 2004, 05:22 AM
Oh, yes, and the number of players, size of map, and other pre-sets of the game could alter the scheme... so it is many schemes for each level