D=2000, in flip formula

[Moulton]

If I understand this statement correctly, is it not true...
"If the city doesn't have any tiles under any other nation's control, then the city cannot flip,"

I have had cities flip back when then culture boundaries had been pushed far back -- no tiles under foreign control. Only once with the 1.29 patch, but the Zulu capital was on an island, about 12 squared away.

Also, there is a random factor not stated in the formula. Or maybe the randomness is just the formula applied to the random number...
I had an English city flip to me on one turn -- after all, it was on my turf, about equidistant from the capitals, but they were in awe ofmy culture. Unfortunately, we lost elec power before I could save, and my ups had died.... when I restarted on the beginnig of that turn, from autosave, it never flipped again.

 

[Cartouche Bee]

Originally posted by TheNiceOne

Yes, you misinterpreted the conclusion. In the formula, "foreign" should be changed to the name of the civ you compute the chance of flip towards, and "local/your" should be changed to "current owner".

I thought as much, but people read these threads and then get the wrong idea.

On one of the old versions, 1.16 (IIRC) I had a small city flip that had no foreigner and no tiles under foreign control, it was only 3 tiles from my palace and it flipped. I'm glad they made some changes and exemptions to the culture flip code over time. Anyway it adds some spice to the game and lets a true warmonger keep moving his palace to the front, did that in GOTM15 to counter Egypt.

 

[Globetrotter]

Originally posted by TheNiceOne

Yes, you misinterpreted the conclusion. In the formula, "foreign" should be changed to the name of the civ you compute the chance of flip towards, and "local/your" should be changed to "current owner".

Now consider a case where you take a city with 10 French and 10 Roman citizens. While both of those civs are alive, the flip chance is computed twice each turn, once where France is considered "foreign" and once where Rome is considered "foreign".

When you use the formula, you'll see that when the 10 French are considered the "foreign", the 10 Romans doesn't affect the formula at all (except by help making a WLTKD or civil disorder). So if you kill off Rome, there will never be any flip calculation where the 10 Roman citizens are considered foreign so they will no longer affect flip chances at all - they will effectively be your citizens.

So a city with 20 Roman citizens (and none other) will have no "foreign" citizens any more. If the city doesn't have any tiles under any other nation's control, then the city cannot flip, just as the case would be if the city contained 20 of your citizens instead.

Or said in another, much simpler way: Citizens of an eliminated civ is considered belonging to the city owner for flip calculation purposes.

Exactly what i initially understood, and exactly what i wanted to hear! Thx TNO!

 

[etj4Eagle]

Originally posted by Moulton
If I understand this statement correctly, is it not true...
"If the city doesn't have any tiles under any other nation's control, then the city cannot flip,"

I have had cities flip back when then culture boundaries had been pushed far back -- no tiles under foreign control. Only once with the 1.29 patch, but the Zulu capital was on an island, about 12 squared away.

Also, there is a random factor not stated in the formula. Or maybe the randomness is just the formula applied to the random number...
I had an English city flip to me on one turn -- after all, it was on my turf, about equidistant from the capitals, but they were in awe ofmy culture. Unfortunately, we lost elec power before I could save, and my ups had died.... when I restarted on the beginnig of that turn, from autosave, it never flipped again.

Yep there is a random element, the formula for culture flipping is a formula for calculating a probability. Once the game has calculated the chance of a flip it rolls the "die" and sees if the city flipped or not. This is the same situation with pretty much all the civ formulas, they are probalistic not binary yes/no ones.

 

[TheNiceOne]

Originally posted by Moulton
If I understand this statement correctly, is it not true...
"If the city doesn't have any tiles under any other nation's control, then the city cannot flip,"

I have had cities flip back when then culture boundaries had been pushed far back -- no tiles under foreign control. Only once with the 1.29 patch, but the Zulu capital was on an island, about 12 squared away.

You cut off too much from my previous post. The qoute from me that you included described a situation where the foreign civ was eliminated, so that the city in question only contained citizens of the eliminated civ. In this case my above statement is true. But if the foreign civ isn't eliminated, there is always a flip chance (unless you have enough military in the city).

Also, there is a random factor not stated in the formula. Or maybe the randomness is just the formula applied to the random number...
I had an English city flip to me on one turn -- after all, it was on my turf, about equidistant from the capitals, but they were in awe ofmy culture. Unfortunately, we lost elec power before I could save, and my ups had died.... when I restarted on the beginnig of that turn, from autosave, it never flipped again.

The formula gives a percent chance of a flip each turn, and a random number is drawn to see whether it flips. In your example, assuming you had "save random seed" turned on, if after reloading, you had played the turn in exactly the same way, you would reproduced the flip as well, since it would then use the same formula and the same random number.

 

[one_hoop]

<B>Why not incorporate the 100% into D?</B>
I have no idea why D is being defined as 2000 instead of 20. The results from this formula are fractional values that people are converting into percents without any mention of the additional calculation. Why not just throw the percentage calculation in with the constant???

Sheesh! (grin)
M@

 

[ConanConQ]

OK, check out my HOF try Deity and see how a city flipped with unbelievable odds.

If I add all this in the formula it would be:

P=[(F+T)*Cc*H*(Cte/Cty) - G]/D
F=1
T=0
Cc=1
H=5
Cte= 63K
Cty= 97K
gG= 0
D ?? = It was rather close to my capital, 10 tiles right, 5 down and 1 cross down from MY cpital.
Distance to enemy capital from city in question was 29 tiles east, 19 tiles south..
CAN SOMEONE CALCULATE THE "D" FOR ME ?? and maybe also add all the figures in the formula if u got spare time ?

 

[anarres]

D in your case is 3.17*2000 = 6340, so the odds of a flip are 0.01577% per turn.

FWIW, over 100 turns the odds of a flip is 1.565% (if my calcs are correct), so even cities that seem unflippable are in the realms of possibliity over enough turns...

The easy way to remember the odds is that each "unit" of chance is 0.05%. With no garrison you just take (F+T), and multiply by all the factors, then divide by the distance ratio. This means you can quickly work it out in your head. Your problem took 5 seconds for me to work out it was approx 0.015%, but about 5 mins to get it exactly.

 

[mikezang]

This is old thread, but I want to confirm something, please forgive me
All people are talking about this formula as below:
P = [(F + T) * Cc * H * (Cte / Cty) - G] / (2000 * D)

Assume Player's capital is Tile(x1, y1), AI's Cipital is Tile(x2, y2), the city is Tile(x, y)

I am clear about those:
F = # foreigners, with resistors counting double
T = # working tiles under foreign control
H = .5 for WLTKD, 2 for disorder, 1 otherwise
G = # garrison units

It looks like what I am clear as below:
Cte = Total culture of the foreign civ (show on F8 screen)?
Cty = Total culture of your civ (show on F8 screen)
D = [max(abs(x-x2), abs(y-y2)) + min(abs(x-x2), abs(y-y2)) / 2] / [max(abs(x-x1), abs(y-y1)) + min(abs(x-x1), abs(y-y1)) / 2]

I am not clear as below:
Cc = 2 if foreign civ has more local culture than you, 1 otherwise
What is foreign civ local culture, what is mine? Does it mean it is not mine if a culture building is built before captured?

 

[ainwood]

I am not clear as below:
Cc = 2 if foreign civ has more local culture than you, 1 otherwise
What is foreign civ local culture, what is mine? Does it mean it is not mine if a culture building is built before captured?

This is the culture of the buildings built in this particular city, rather than the culture of the total empire.

If you capture a city that has been founded hundreds of turns previously, then it may have had a lot of cultural improvements built in it. In which case, when you capture the city, the LOCAL culture of that other civ in the city will be higher than yours (until you build up local culture of your own via building cultural improvements etc).

You can't get it accurately from the game - the best you can really do is know what the size of the cultural boundary of the city was before you captured it -
* Border = 1, < 10
* Border = 2, 10 < culture < 100
* Border = 3, 10 0< culture < 1000
etc.

 

[mikezang]

So those are right if I understood, is it right?
Cc=1 if a city is built by me,
Cc=2 if it is captured but culture points by me more than before captured
Cc=2 if it is captured but culture points before captured more than by mine

 

[ainwood]

cc = 2: Culture points before capture are greater than culture built since.
cc = 1: Culture points before capture are less than culture built since.
cc = 1: City built by me.

Think of it this way: A city that has been built by someone else (and has lots of culture) will want to flip back to them: Hence CC is higher (cc=2).

 

[mikezang]

Oh, I made a mistake, Thanks for your confirmation.