culture flipping in 1.21

[LaRusso]

how does culture flipping work for you in the new patch?
i usually manage to keep conquered cities, but in my last game, my size 18 city flipped to my allies, indians, while WW was killing my democracy in a prolonged war. it was only logical that it flipped, so many unhappy people, on the brink of revolt, and depleted of any units (they all went to die at a beachhead in germany). occassionally, i'd flip a small AI city

[notyoueither]

And your 'allies' accepted it?

Kill them. They are not good allies.

[Harovan]

It's hard to lose a size 18 city. The war must have been really long and your culture pretty low. Why didn't you switch to monarchy as soon as you noticed, that WW gets out of hand?

[LaRusso]

i could have switched to communism, sillly of me that i did not. my culture was pretty okay, i had several wonders and was leading the tech race by a whisker

[Fitz]

There's a thread around that Soren from Firaxis actually posted to explaining it recently, but IIRC it boils down to:

A chance for each foreign citizen not countered by a military unit.
A chance for each square inside the cities maximum production radius (21 squares) that is actually in another civs territory, and not countered by a military unit.

If the city had more culture produced in it by the other civ in question than by you, it takes two military units to negate the chance produced by each citizen/square.

The size of the "chance" is determined by the distance to your capitol, and the distance to their capitol.

So lets say that size 18 city was a captured one that originally belonged to the Indians, so all the citizens are still Indian (not the civ you were fighting. If you contolled 8 of the 21 squares, the indians 3 (one side of the "square") and the civ you are fighting the remaining 9. Then it takes 36 military units to prevent the flip to the indians due to citizens (18*2 since they almost certainly produced more culture there than you. It takes 6 military to stop the squares flip (again doubles due to culture), and probably 9 to stop the flip back to the civ you took it from (probably they produced no culture yet).

Now for the actual chances of the flip, Soren said the distance to the capitols has a range of dividing by 500 to 8000 (caps on the chances), averaging about 2000 (I assume for equal distances). Then you muliply that chance by the ratio of their culture to your culture.

So, chance to flip to indians with no military = {range of 8.4% to 0.5%, avg 2%}*{Indian culture/your culture}, with each military unit subtracting .2% to .01%, avg .05%, times the ratio of culture.

Now assume you had left 4 military of the 42 required, that means a range of 7.6% to 0.45% with an average of 1.8%, times the ratio of culture.

So if that huge city was much closer to the indians (assume /1000), and you had 1.5 times their total culture, you might see a 2-3% chance per turn, and if they had 1.5 times your culture, you might resonable expect 5-6% chance per turn. Not too high, but certainly high enough to see a flip within a couple of turns (5 turns at 5.5% means 25% chance of a flip, 10 turns means 44%).

So, probably too many numbers, but the key point is to squash all chance of a flip, make sure you have enough military units (1 per foriegn national, 1 per square not yours that should be, double if the national/territory threatening your continued ownership has produced more culture in the city than you) or get rid of as many foriegn citizens and intruding borders as possible (my favored approach.

[Coracle]

We don't need explantions from Soren.

he's the one who got so much wrong in the first place.

[Cyclotron]

Originally posted by Coracle
We don't need explantions from Soren.

Well, those of us still playing the game would like those explanations.

The next time you use "we," make sure everybody agrees with you. Until then, speak for yourself.

he's the one who got so much wrong in the first place.

Says you. Stop using the same old meaningless, baseless, fact-starved half-baked opinions that everybody knows you by, and try an intelligent discussion sometimes.

[Grrr]

Culture flipping still occurs occasionally when not nesecary, but generally it is a lot better in 1.21

[LaRusso]

thanx!
just one note. 10 times 5.5 % is still only 5.5% probability each turn
btw, that was a huge city with all thingies, 5 tiles from the FP but very close to their big city. i took it over from chinese 1000yrs ago and all citizens were assimilated. i guess WW and unit depletion did it.

[Fitz]

10 turns of 5.5% is only 5.5% each turn, but there is a 44% chance that a flip will occur at least ONCE in those 10 turns.

Chance of not flipping in 10 turns = (100% - 5.5%)¹⁰ = ~66%, therefore chance of flip occuring at least once (which is all it takes) is ~44%.

Probabilities is one of my stronger points. Comes from being a old time D&D fanatic.

Edit: Of course I say that and then I make a math error.

Now if only Soren would give us a more detailed explanation of how the distance factor is determined (500 to 8000 avg 2000 is very vague and not very mathematical) and the necessary steps to be able to see another civs culture points, I could write it out as a formula that would actually allow you to calculate exact percentages. I miss the days of SMAC and actually being able to be precise. You listening Soren?