View Full Version : How much improvements really help corruption


Bamspeedy
Mar 16, 2002, 10:21 PM
Did a study on a tiny map. No opponents. To see how much the improvements really affected corruption. Stopped at 12 cities (# of ideal cities for a tiny map), although I did notice a couple cities lose 1 shield or 1 gold when I added the 12th city. The map is entirely bonus grassland so when maxed at size 12 city all cities would have potential to produce the same amount. developed just the outside squares so when the border expanded all citizens would work the outer perimeter so I wouldn't have to connect any cities right away. Mined and roaded the worked squares. Built marketplaces, banks, temples, cathedrals, colleseums in all cities.

I found out five squares straight east is not the same as say five squares southeast or northwest. Makes sense, never realized that. So 10 squares straight east or west is the same as 15 squares northeast or southwest, etc.

Results for Democracy:

Connecting the cities by road had minimal impact, amounted to 4% in most cases (corruption and waste). Although if I had luxuries this would help kick off a WTKD.

Courthouses: Nearby cities only saw a 4% improvement. Far away cities saw as much as a 27% improvement.

Police Stations: Nearby cities saw 4% improvement just like the courthouse. Far cities 20%.

Results for communism:

Roads: 0% for corruption, 4% for waste
Courthouses: 7%
Police stations 14% corruption, 7% waste

This was with science at 100%. When I bumped it down to 0% the corruption percentages were much smaller (because of the marketplaces and banks multiplying the uncorrupted gold that is going straight to the treasury) so the difference would be about half as much as the above results.

Here's a file for more detailed results: (hopefully this works), it's in microsoft works spreadsheet format.

Edit: typos and errors (railroads, what was I thinking I didn't have no coal :crazyeye: )

Note: I deleted a clone of this thread and the attachments got deleted as well. I managed to re-upload the attachments. ~TF

EDIT Please look at Alexman's study. His study which was done after this one, is more accurate, and has been updated for the latest patches. Do you think you understand corruption (http://forums.civfanatics.com/showthread.php?s=&threadid=19922)

Park Ranger
Mar 17, 2002, 06:34 AM
Hey, this is excellent! Thanks for doing this little study. Most intersting (to me) bit of info there is the difference between diagonal distances and perpendicular ones.

I have two questions: is it possible that the effects of courthouses inrease over time? I thought I read this somewhere and have been looking for evidence in my games but since I don't have "scientific method" yet my undiscliplined observations have produced no results. ;)

And, does WLTK work better if there's already a courthouse there?

chiefpaco
Mar 17, 2002, 09:51 AM
Cool stuff, Bamspeedy. A good study & well presented. I think we can really learn more about corruption/waste management from this kind of study. A few questions:

I was wondering about your diagonal distances. Did you measure the diagonals by counting over on the diagonal? I wondered how the diagonal calc worked because of the triangle theorem (it should be further) & the funny shape of the world figured in. It looks by your calcs in some cases (the 5 E vs. 5 NE)that diagonal is actually considered closer to the palace than straight across or down!

Under communism, you list police stations as a 14% reducer of corruption. I wondered, is it double that of a courthouse because there is already a courthouse in the city? Or, if just a police station was added, would it still give a 14% improvement. I guess I just wondered if the results of your study would be the same if you added the police station 1st.

Does your corruption/waste factor remain consistent after you sell a building? i.e. does selling the police station make the corruption/waste go back to the previous level with just the courthouse?

BTW, I always thought Communism would yield a good corruption/waste % after putting in Police Stations/Courthouses. Thanks for proving me right! :)

Bamspeedy
Mar 17, 2002, 03:12 PM
I was wondering about your diagonal distances. Did you measure the diagonals by counting over on the diagonal? I wondered how the diagonal calc worked because of the triangle theorem (it should be further) & the funny shape of the world figured in. It looks by your calcs in some cases (the 5 E vs. 5 NE)that diagonal is actually considered closer to the palace than straight across or down!

I was imagining that the right of your screen is E and the top of your screen is N. I counted them by how many tiles I would have to use to get a unit there. So going horizontal or vertical it would still take 5 moves like going 5 moves diagonally. However, going horizontal or vertical you are actually further away. Easier to see this when you zoom out. So a city 5 tiles to the east is like being 7.5 tiles to the northeast.

Does your corruption/waste factor remain consistent after you sell a building? i.e. does selling the police station make the corruption/waste go back to the previous level with just the courthouse?

I went and sold all courthouses and it went back to all the same stats as when I had courthouses, but now I have police station there. Either they work better together, or (more likely) it has to do with rounding the numbers off. With only 27 gold possible, there is a margin of error of 3%. Without courthouse it lost 5, with courthouse 4 (this could have been 3.5), and with police station 2.

And, does WLTK work better if there's already a courthouse there?

Found out WTKD reduces waste (lost shields), but not gold (corruption). With courthouse and police station cities gained back 1-3 shields out of 26. Without a courthouse or police station cities gained back 1-5 out of 26. However without the courthouse, the 2 farthest cities weren't able to go into WTKD, even with luxury expense at 100%, because they didn't have enough uncorrupted money to put into entertainment for their city.

I have two questions: is it possible that the effects of courthouses inrease over time? I thought I read this somewhere and have been looking for evidence in my games but since I don't have "scientific method" yet my undiscliplined observations have produced no results.

Nope, no increase with time. I sold the police stations about 100 turns after I had built the courthouses and the stats were the same as when I had freshly built the courthouses.

Beard Rinker
Mar 19, 2002, 02:51 PM
Originally posted by Bamspeedy

Found out WTKD reduces waste (lost shields), but not gold (corruption). With courthouse and police station cities gained back 1-3 shields out of 26. Without a courthouse or police station cities gained back 1-5 out of 26. However without the courthouse, the 2 farthest cities weren't able to go into WTKD, even with luxury expense at 100%, because they didn't have enough uncorrupted money to put into entertainment for their city.


I assume the 1-3 shields is in addition to the shields gained back by the courthouse and police station.

If I interpret your numbers correctly, it shows that WLTK is as effective as a courhouse for waste.

Bamspeedy
Mar 19, 2002, 04:01 PM
Yes, it appears that WTKD has the same effect as a courthouse. Without a courthouse all the cities that did go into WTKD showed the same results (in waste) as if they had built a courthouse.

Keep in mind, although courthouses and police stations seem to have a small impact on corruption (gold), if you have the money and science improvements, those improvements multiply the uncorrupted gold. So a city with 30% corruption with no improvements would have 15% if it had marketplace, bank, library, and university. It depends on what improvements you have and what your putting your money into.

I added 12 more cities to test the optimal number of cities corruption (doubling the cities). Basically, with courthouse and police station, all the cities almost doubled their corruption. A city with both the corruption improvements would have the same corruption as a city one step closer without any of the corruption improvements. In democracy, this was the case in both corruption and waste. In communism waste went from 19% to 31%, corruption doubled, going from 14.3 to 28.6%

Grey Knight
Mar 19, 2002, 05:32 PM
Great research, however I think the analysis is a bit flawed. Look instead at the reduction in the waste.

Looking at the change from none > courthouse, and courthouse > police station you don't really see anything glaring (something around 30% for each), however if you look at the numbers for none > (courthouse + police station) a very clear picture arises!

Courthouse + Police Station = 50% reduction in corruption

FWIW, the reduction afforded by each is actually 1 - ( 1 / Sqrt(2) ) -- roughly 29.29% :D

Cheers,
Shawn

Excalibur_Z
Mar 23, 2002, 08:02 PM
Gray Knight,

An excellent idea. Now, I just installed 1.17f about a half-hour ago so I haven't tried it yet, but now it sounds like it might have been a mistake =) I actually haven't played much Civ3 at all since the gameplay is so much different from the first two Civs, so the majority of information I've received here is foreign to me. Is the tech cost in gold really 1/10 of its research cost in beakers?! Ridiculous! With that data, my initial strategy would be to simply invest 100% into commerce and leave the teching to the AI players, then buy techs off of them. From what I've read so far, this is already how a lot of players progress through the game.

Teching is so horribly slow in Civ3 as compared to the other Civ games. Part of the reason I enjoyed Civ2 so much was because I was able to win on Deity fairly regularly due to tech advantages and the espionage system. Apparently Sid has determined that this was a cheap way to win. However, I think out-teching your opponents should remain a valid strategy due to the new combat system.

In Civ2 I was able to focus on teching up before going into any extended war, eventually catching my opponent by surprise with my higher-level units. Fast Elephants and Crusaders were able to storm and decimate enemy cities. In Civ3, I was expecting to utilize roughly the same strategy using Swordsmen, but there is such a significant tech penalty. Mind you, I realize Firaxis is encouraging a different style of play that requires more strategy and planning, but giving the AI distinct advantages even on Regent level is unfair. Corruption and AI advantages are clearly rampant, leaving the human player in a constant catch-up position - a position I would rather not be forced into as a result of game mechanics and designer intention.

So, I advocate higher tech value in gold, as well as removal of any AI advantages that may be in place. Early expansion is fine. AIs that tech up fast are fine. AIs that use defensive expansion are fine (Perfectionist states in Civ2). AIs that use militaristic suppression are fine. However, don't give each individual AI all of these then supplementing those behaviors with unfair production/economic advantages, as it puts the human player against the ropes for the majority of the game.

JoseM
Mar 23, 2002, 08:52 PM
what the hell is a "WTKD" ? can u people be especific when ur talking about something?

Excalibur_Z
Mar 23, 2002, 08:57 PM
We Love the King Day - state of a city where you have more happy citizens than content citizens/specialists, and no unhappy citizens. Population prerequisite is 5 I believe.

Switch625
Mar 24, 2002, 08:07 PM
Originally posted by Excalibur_Z
We Love the King Day - state of a city where you have more happy citizens than content citizens/specialists, and no unhappy citizens. Population prerequisite is 5 I believe.

The pop requriement for WTLKD is 6. It is adjustable in the editor, though.

Michelangelo
Mar 25, 2002, 03:06 AM
Great study Bamspeedy.

Originally posted by Bamspeedy
developed just the outside squares so when the border expanded all citizens would work the outer perimeter so I wouldn't have to connect any cities right away.


What do you mean that working the outside squares would deminish the need to connect the cities with roads. Do you mean that working the outside squares is the same as connecting them with roads.

Shabbaman
Mar 25, 2002, 04:00 AM
Originally posted by Bamspeedy

Nope, no increase with time. I sold the police stations about 100 turns after I had built the courthouses and the stats were the same as when I had freshly built the courthouses.

Well, how sure are you about that? I've got a b*load of totally corrupted cities, and over time corruption is lessening: instead of 1 unwasted shield they produce 2 unwasted shields :D

Anyways, in these (remote) cities nothing seems to decrease corruption: no courthouse, no police station and no luxuries, not even the 'We love the Dark Lord'-day ;) ...any suggestions? I guess I can't move my capital 'cause not enough production 'close' enough, and no leader...

warpstorm
Mar 25, 2002, 06:46 AM
The only problem I see withthe study is that it stops just when corruption and waste become real problems. Once you get above the ideal number of cities these rules don't quite apply. In huge empires (ones where more that half of your cities are not within the beneficial effects of your palaces), communism is the only way to go. When Democracy is giving less than 50% of your shields empire wide with nothing that can be done about it, it's time to switch.

Bamspeedy
Mar 25, 2002, 07:01 AM
What do you mean that working the outside squares would deminish the need to connect the cities with roads. Do you mean that working the outside squares is the same as connecting them with roads.

What I did was have them work the outer 12 squres not the inner 9 squares so I didn't have to connect them by roads to test if connecting them by roads actually had any effect. The outer perimeter is perfect for the test because I made sure the cities were maxed at 12 so each citizen was working the same tiles, producing exactly the same shields/commerce. I've heard many times where people have said that to help cut corruption, make sure they are connected to the capital by road. This actually doesn't help at all (or very little, like 4%, which might be a 10% REDUCTION in corruption, but in most cases added just 1 gold to a city if any). It does help in kicking off a WLTK day because they have access to luxuries if actually connected by road.

Well, how sure are you about that? I've got a b*load of totally corrupted cities, and over time corruption is lessening: instead of 1 unwasted shield they produce 2 unwasted shields

Perhaps they were producing more shields (corrupted ones, too) than before. Example: City producing 1 good shield out of 5, later on you look at it and it's producing 2 out of 10. It's still 80% corruption.

Anyways, in these (remote) cities nothing seems to decrease corruption: no courthouse, no police station and no luxuries, not even the 'We love the Dark Lord'-day ...any suggestions? I guess I can't move my capital 'cause not enough production 'close' enough, and no leader...

I don't really have any suggestions, when some cities are just way too far from the capital AND you have way too many cities, nothing can improve it. If you haven't exceed the optimal number of cities by a large margin, they should eventually produce more than 1gold, 1shield/turn if the population gets big enough, regardless of how far away they are. I once had 6 times the optimal number of cities (so corruption/waste was 6X as bad), so some far cities produced only 2 good shields and 2 uncorrupted gold, even when it was a size 22 city, with courthouse, marketplace, bank, police station, hospital, factory, WLTKD etc. These towns would be better off just being a specialists town. Gold and beakers from Taxmen and Scientists are immune to corruption, and have them build Wealth, that's also immune to corruption, even if it is only 1 gold/turn.

However, there is the 'palace jumping' trick if you really want to move your palace to more ideal site. It's really risky, because you don't alway know exactly where it will jump. Plus, you'd lose any improvements you had in your capital.

The only problem I see withthe study is that it stops just when corruption and waste become real problems. Once you get above the ideal number of cities these rules don't quite apply. In huge empires (ones where more that half of your cities are not within the beneficial effects of your palaces), communism is the only way to go. When Democracy is giving less than 50% of your shields empire wide with nothing that can be done about it, it's time to switch.

Yes, I stopped at 12, so I wouldn't have to worry about optimal number of cities being formulated in there. If you look at one of my previous responses I showed the results from doubling the number of cities (corruption basically doubled). So I imagine instead of courthouse + police station reducing corruption by 50%, it would seem more like 25%.

chiefpaco
Mar 25, 2002, 10:02 AM
Originally posted by warpstorm
The only problem I see withthe study is that it stops just when corruption and waste become real problems. Once you get above the ideal number of cities these rules don't quite apply.

I think Bamspeedy was trying to single out the corruption by distance factor to show its behaviour. This, he did well.

You're right, the 2nd factor is # cities. I'd recommend Aeson's study over at Apolyton for a look at how it works. It's buried in the " Corruption from number of cities" thread in their strategy section. The two factors work independently, so knowing both can help you make the most of your empire's value.

Originally posted by warpstorm
In huge empires (ones where more that half of your cities are not within the beneficial effects of your palaces), communism is the only way to go. When Democracy is giving less than 50% of your shields empire wide with nothing that can be done about it, it's time to switch.

Communism suffers from # cities only, not distance. However, every new city you make past the optimal cities # under Communism adds corruption/waste to every city equally. Your total corruption/waste percentage will still be like that of other govts. (i.e. adding a new far-flung city under Democracy adds a ton of c/w in that city. Adding it under Communism adds a small amount of c/w over all your cities.)

PS. This will not be true for the case where you have fewer cities than the optimal # and your empire is spread out past the limits of corruption by distance. Then will Communism be definitely more efficient.

chiefpaco
Apr 04, 2002, 04:15 PM
Which Civ did you use for the study? Did I miss it somewhere? Is it Commercial? I would like to extend the study to judge the value of a Commercial civ with regards to Corruption by distance, unless it's been done before.

Aeson did a study for corruption by # cities that included Commercial & non-commercial civs. He determined you get 1-2 extra cities before hitting the corruption limit. I think knowing the commercial effect on the distance of cities would be the final piece of the puzzle.

Could you post your save too so I can compare the maps correctly? Unless you'd like to study it yourself & show the results :)

Otherwise, I can try to do the 2 tests myself...

Bamspeedy
Apr 05, 2002, 07:22 AM
I used Babylon. Religious, so you can easily switch between governments and study the various effects of governments on corruption. I'll PM you the saved game, so it won't use up space on the forum's server.

Bamspeedy
Apr 05, 2002, 07:26 AM
Aargh! you can't PM files! Ok, I guess I'll post it here: Here's the map just after the roads are connected:

Bamspeedy
Apr 05, 2002, 07:28 AM
Mistake, look at next page

Bamspeedy
Apr 05, 2002, 07:29 AM
Here it is

chiefpaco
Apr 05, 2002, 09:06 AM
thanks

Edit: You could delete the save now, if you wish. Unless you feel it would be valuable for anyone who wishes to doubt your study. :)

chiefpaco
Apr 09, 2002, 03:27 PM
Hey. I just did a minitest on the affect of being a commercial civ on the corruption by distance. I used Bamspeedy's map & set up a similar city pattern. I found that being commercial really didn't help the corruption by distance, maybe 1 difference in some cities. Here's my results.

Bamspeedy: I'm posting my save too because I haven't figured out why my non-commercial corruption & waste percentages are not identical to yours. Perhaps because I did not make as many cities?

I'll try to figure it out. If you want to try too, that would be great. If I pull up your game, I get the exact same numbers as you had in your sheet.

Bamspeedy
Apr 09, 2002, 03:45 PM
I might download it later this month. I'm kinda weary about Loading other saves, thinking they might screw up the current saves in my directory since I'm playing the GOTM. I'm sure I could put a new directory in, but don't want to take the chance that my autosaves get deleted or something. But how many cities did you use? Optimal number on tiny is 12, but I did notice a change in some cities when I added the 12th city. I'm also wondering if the placement of some cities would affect the corruption of other cities. Like the general lay-out of your civ. Would placing cities in a perfect circle have any difference to a specific city's corruption than if all the cities were placed in one direction? There is a study posted in the General Discussion forum you might find interesting (he found no difference in commercial, either), and it has more statistical numbers for figuring corruption with both distance and #of cities factors.

Bamspeedy
Apr 09, 2002, 03:52 PM
Oops, I mean it's in the Strategy & Tips section, it's titled "Do you think you understand corruption?" or something like that.

chiefpaco
Apr 09, 2002, 11:09 PM
We're right. It was most probably the cities. I assumed (wrongly) before that since 12 was the optimal limit, there would be no affect on the # of cities up to 12. This is not the case. I took your game & abandoned a few cities & your numbers started to look more like mine.

Thanks for the input. :) & yes, I'm very much following their progress.

Strider
Jun 30, 2003, 11:05 PM
Could you possibly find out how much being a commercial civilization actually drops corruption?

Bamspeedy
Jul 01, 2003, 09:34 AM
This little study was done before Alexman's awesome, more detailed study (and his study has been updated for the patches), so I recommend looking to him and his thread for any corruption questions:

Do you think you understand corruption? (http://forums.civfanatics.com/showthread.php?s=&threadid=19922)

I believe commercial civs have the effect of having +25% more cities for the 'optimal number of cities' corruption, with the latest patches or PTW, and 12.5% with the out-of-the box unpatched version of civ3.

Tigerclaw
Sep 09, 2003, 11:48 AM
Would it be possible to use the ptw editor to essentially "clone" the Forbidden Palace Sm Wonder several times (maybe 4-5 times, all differently named of course), to give you a number of "Regional Capitals" (which could then be strategically placed to have the greatest effect...) Would this even have any effect? (sorry if this is like, the wrong forum or something, but this is my 1st post...*begs forgiveness*)

Norlamand
Sep 11, 2003, 04:16 PM
Tigerclaw, yes it can be done. In fact if you play the DyP mod there are 3 or 4 additional palaces (winter, summer and others that I don't remember). With them you can set up several productive core cities. I don't play DyP so I haven't tried it for effect.

Hygro
Dec 08, 2003, 05:20 AM
I'm curious: did you involuntarily use a RCP varient due to symmetrical terrain? I have not looked at the save.