Too much corruption
I agree, the model for corruption in Civ3 is totally messed up. Very unrealistic to have most of the world's cities producing only 1 shield. In my game (1st game actually), warlord level with huge map, I'm Americans in South America. Washington is about where Rio really is, and Houston has ForbCity about where Buenos Aires is. I invaded Mexico and all the rest of North America--the corruption is already up to 1 shield in Columbia!!! Never mind that I have beautiful Aztec Cities in Texas and California producing 40-50 shilds per turn but after corruption only yielding 1 real shield! And pol station/courthouse don't help at all. Last night I changed the values in the editor to 100 cities, 800% corruption ( I have 50+ cities)
and it did a much better job of evening it out. I'm thinking that 500% might be a little better for getting more of the effect the authors intended.
But they ought to do better. Distance from the capitol is the main factor until you get to the max number of cities. The rulebook (or was it the strat guide?) says that corruption rises "exponentially" after the number of cities is passed. I just see it go to 1 shield/1 gold. It needs to be a continuous mathematical function of some sort, say for every additional city it goes up 2%.
Corruption should work like this: Say Corruption is C, expressed in %, for example 20% meaning 20% loss in shields. Two values ought to come into play as a function of distance from capitol / forbidden city: some minimum corruption value which is approached far away from the capitol (say 70%) and some radius which governs how fast C goes up as a function of distance. the function could be C=m[1-exp(-(x/r)^2)] where m = max corruption and r = radius. This function falls off only slightly at first, but faster later on as x passes r until, when x is much greater than r, C maxes out at value m. "m" could be a value chosen by government type, with communism having r=infinity to make corruption flat.
It also says in the Civilopedia that harbors and airports fight corruption. This seems realistic, as "things that make the world seem smaller" should lower corruption. It seems to me the way it should work is that the distance of a given city from a capitol/ForbCity should be counted, then adjusted down by 1/3 for roads and say 1/10 for railroads. If the path goes over water, the water should count double if no harbors, but if a link can be counted harbor-to-harbor (maybe with roads/railroads or plain land squares in between on each end) then the distance caused by the water should be decreased to like 1/3 per square. Likewise, if airports are involved, the "corruption" distance between airports could be 1/10 (or 1/20?) the "real" distance in counting this. This all would simulate "making the world smaller thru technology" which has been a real force--a huge empire can manage its resources much better if the government and economic guys can get around quickly.
Then they should multiply the other function by a function of the number of cities, some function that goes up gradually once the max number of cities is passed.
I've noticed that courthouses and police stations only take a little from corruption--they should tweak up the effect of these as well. It also makes since to me for marketplaces and banks to fight corruption, since banks allow laws to govern finances and marketplaces lend an alternative to black-market bartering, etc.
All the values I used above are not playtested obviously--they just serve to make the point of my idea. I would love it if Infrogrames would entertain such an improved model.
Oh yeah, there should also be a factor in the corruption function for map size, so that the radius changes as a function of the map AREA changing, so that corruption is the same proportionally on an earth on a huge map or on a medium map.
Any comments on my dissertation are welcome.
I agree, the model for corruption in Civ3 is totally messed up. Very unrealistic to have most of the world's cities producing only 1 shield. In my game (1st game actually), warlord level with huge map, I'm Americans in South America. Washington is about where Rio really is, and Houston has ForbCity about where Buenos Aires is. I invaded Mexico and all the rest of North America--the corruption is already up to 1 shield in Columbia!!! Never mind that I have beautiful Aztec Cities in Texas and California producing 40-50 shilds per turn but after corruption only yielding 1 real shield! And pol station/courthouse don't help at all. Last night I changed the values in the editor to 100 cities, 800% corruption ( I have 50+ cities)
and it did a much better job of evening it out. I'm thinking that 500% might be a little better for getting more of the effect the authors intended.
But they ought to do better. Distance from the capitol is the main factor until you get to the max number of cities. The rulebook (or was it the strat guide?) says that corruption rises "exponentially" after the number of cities is passed. I just see it go to 1 shield/1 gold. It needs to be a continuous mathematical function of some sort, say for every additional city it goes up 2%.
Corruption should work like this: Say Corruption is C, expressed in %, for example 20% meaning 20% loss in shields. Two values ought to come into play as a function of distance from capitol / forbidden city: some minimum corruption value which is approached far away from the capitol (say 70%) and some radius which governs how fast C goes up as a function of distance. the function could be C=m[1-exp(-(x/r)^2)] where m = max corruption and r = radius. This function falls off only slightly at first, but faster later on as x passes r until, when x is much greater than r, C maxes out at value m. "m" could be a value chosen by government type, with communism having r=infinity to make corruption flat.
It also says in the Civilopedia that harbors and airports fight corruption. This seems realistic, as "things that make the world seem smaller" should lower corruption. It seems to me the way it should work is that the distance of a given city from a capitol/ForbCity should be counted, then adjusted down by 1/3 for roads and say 1/10 for railroads. If the path goes over water, the water should count double if no harbors, but if a link can be counted harbor-to-harbor (maybe with roads/railroads or plain land squares in between on each end) then the distance caused by the water should be decreased to like 1/3 per square. Likewise, if airports are involved, the "corruption" distance between airports could be 1/10 (or 1/20?) the "real" distance in counting this. This all would simulate "making the world smaller thru technology" which has been a real force--a huge empire can manage its resources much better if the government and economic guys can get around quickly.
Then they should multiply the other function by a function of the number of cities, some function that goes up gradually once the max number of cities is passed.
I've noticed that courthouses and police stations only take a little from corruption--they should tweak up the effect of these as well. It also makes since to me for marketplaces and banks to fight corruption, since banks allow laws to govern finances and marketplaces lend an alternative to black-market bartering, etc.
All the values I used above are not playtested obviously--they just serve to make the point of my idea. I would love it if Infrogrames would entertain such an improved model.
Oh yeah, there should also be a factor in the corruption function for map size, so that the radius changes as a function of the map AREA changing, so that corruption is the same proportionally on an earth on a huge map or on a medium map.
Any comments on my dissertation are welcome.
Comment