Alexman and Qitai, you seem to be thinking the new FP corruption model might be ok. Please check my reasoning in the following - to me this example indicates that there is something very wrong with the new model (whether it is by design or by accident

)
I set up a test with the following mods: standard size map, emperor level, filled the map with plains, changed plains to produce 10s and 10g, and changed plains to zero movement (to make settling easier.) I pre-created two groups of 12 cities - a NW group with the Palace near its center and a central group with FP near its center.
After startiing a game with the test setup I changed every city to have five working citizens. The two groups are clumped as shown in this minimap:
I noted the corruption levels at this point in the game.
Next I settled 12 more cities in the west as shown in this minimap:
I populated the new cities to have five working citizens each so that all cities on the map would be the same in gross production. This is to allow for the possibility of increased corruption in one city being offset by less corruption in another. Some might argue that such a model would be ok. I would argue that it is not since one would have to build and pay maintenance for improvements in more cities. But to take that possibility out of the test for now I made the new cities have five working citizens.
And then I compared the new corruption levels.
The results are as follows:
Code:
tot.income corruption net.income mfg(net.shields)
24 cities 1752 434 1318 1103
36 cities 2556 1296 1260 1050
When I added those 12 cities in the west, corruption in the Palace region remained unchanged. But corruption in the FP region went up noticeably. Although those new cities contributed a tad, their contribution did not offset the loss in the FP region. (And the loss there is of course more important since that's where one would have courthouses, factories, libraries, etc. already in place, vs. the new cities.)
The amount of increase in corruption in the FP region of course depends on the number of other cities added. If I reduce the size of the 12 new cities in the west to 1, and increase the number of new cities there to 24, the result is the new "48 cities" line below:
Code:
tot.income corruption net.income mfg(net.shields)
24 cities 1752 434 1318 1103
36 cities 2556 1296 1260 1050
48 cities 2338 1294 1044 859
The trend seems bad. In a game where one is going for domination I think the results could easily be a fair bit worse again. In the most severe cases (e.g. high score HOF games) the FP region can over time become totally corrupt, without an offsetting gain elsewhere (if such a thing were even possible, considering the infrastructure rebuild costs that would require.)
Do you agree that this is highly undesirable behaviour? To me it seems bad, probably unintended, and very confusing. The last part is what I like least of all - sometimes FP will work well, sometimes it will work somewhat, sometimes it will hardly work at all. Most players will have a very hard time predicting or understanding the difference between those situations. It will just look like it works sometimes and not others. (Or look like it works for some players but not others because as well as varying during the course of a game it will also depend on play style, map size and difficulty level.)
In case anyone wants to reproduce my results or play with other similar tests,
here is the save file for the first position in these tests, the position with 24 cities and a bunch of settlers ready to use.
Warning! When running tests like these watch out for a confusing glitch: After doing something like settling the 12 new cities shown in the second minimap above, the game will NOT show correct values in the F1 screen. Cities in the FP region will look like nothing there has changed. You must end a turn and then check the F1 screen - at that point corruption in the FP region will have been recalculated to show the new and worse values.