Culture Flip Formula

[Fitz]

*curses after reading Deep's post*

Yeah, looks like it is ((N+S)*C*H)-T*R)/D Unless he f***ed it up. Beacuse 2*2 (2 citizens * 2 for culture in city) is 4, and 14/3.5 is 4.

Will edit the fist post to reflect this.

[DeepO]

DeepO

[DeepO]

Originally posted by Fitz
Yeah, looks like it is ((N+S)*C*H)-T*R)/D Unless he f***ed it up. Beacuse 2*2 (2 citizens * 2 for culture in city) is 4, and 14/3.5 is 4.

Well... it could still be the second formula: 4*3.5-14 gives again 0, so if the human players culture was less then 3.5 times the AI's, there is a slight chance in flipping. the only difference between my two formulae is how big the chance is, and as it changes with a factor of R^2 this is rather significant.
My bet is also on the first one, though

DeepO

[Fitz]

Actually, I realized as I was editing the first post that this has a profound affect on the whole reason for posting the formula. The primary reason for knowing the formula is so you can figure out how many troops you would need to prevent or severely hamper a culture flip, not to figure the actuall % chance of a flip. Since it now includes a number that can only estimated, you can only estimate the number of troops. I find that very annoying.

In point of fact, in a recent game I just started I have a culture of ~10 to 1 to most other civs. Now, that could be anywhere from as little as 8 or as much as 12. Now I have no idea if I can hold a size 5 captured city (more culture than me in city) with 1 troop, or if I need more.

Still this is a pretty huge discovery. Before I would have thought I needed 10 troops to hold that size 5 captured city. Now I know have a larger culture, I can possibly use less (5 at 2:1, 1 at 10:1). Definately an improvement, and definately to your advantage if you have more culture.

Conversely, this would explain why people who play on harder levels are annoyed by culture flips, and those one easier levels probably less annoyed. The harder level you play, the more chance the AI has a greater national culture than you, and thus you need many more troops to squash a flip or seriously impact it. I can certainly understand complaints about 20 troops not preventing a flip now. If you were in a size 5 city, they would be insufficient if the AI had more than 2x your culture.

[DeepO]

Originally posted by Fitz
Actually, I realized as I was editing the first post that this has a profound affect on the whole reason for posting the formula. The primary reason for knowing the formula is so you can figure out how many troops you would need to prevent or severely hamper a culture flip, not to figure the actuall % chance of a flip. Since it now includes a number that can only estimated, you can only estimate the number of troops. I find that very annoying.

I know, I already mentioned that. (maybe not so clear, though). Plus, to counter Coracle's usual whining, it makes totalCulture one of the most important factors for flipping, not distance to the capital. In combination with troops, of course...

Conversely, this would explain why people who play on harder levels are annoyed by culture flips, and those one easier levels probably less annoyed. The harder level you play, the more chance the AI has a greater national culture than you, and thus you need many more troops to squash a flip or seriously impact it. I can certainly understand complaints about 20 troops not preventing a flip now. If you were in a size 5 city, they would be insufficient if the AI had more than 2x your culture.

So again, it makes culture an even more important aspect of CivIII. ATM, I'm playing at emperor, and while it is virtually impossible to get more culture than the AI in the ancient era, I never have a ration less then 1/2 to the Babylonians (which, most of the times, are kings of culture). Knowing this, I will gladly sacrifice building a few settlers to building temples ASAP. Apart from having a better chance of winning cities, at least you won't loose any...

Building culture so early is very important, as after 1000 years all buildings produce double culture. Normally, this only takes places with a few temples and libraries at the end of the game, if you could have double counting temples before 1 AD, they're VERY important later on. I used to think this was only needed if you planned for a cultural victory (as you have to have double the culture of the second in place), but it seems everybody should do it, if they want to hold on to their cities!

DeepO

[Soren Johnson]

Originally posted by Fitz


Regardless, SOREN IF YOU ARE LISTENING PLEASE TELL US WHICH WAY IT WORKS. Important info.

actually, the troops will count for each flip probability roll. In other words, if the city has a Greek and French citizen (you being Roman), one spearman could negate both the Greek and the French citizen.

[DeepO]

Soren, thanks for that (so I was right ), is the rest of the formula correct? Or did we still miss something?

DeepO

[Fitz]

Thanks Soren. I really appreciate you taking the time to answer that. It would be cool if you could tell us if the culture ratio affects the # of troops (as in my revised formula) or the number of squares and citizens (as in DeepO's posed formula). I just realized how important that is in determining the chance of a flip.

In the case of: ((N+S)*C*H - T*R) the culture ratio only has an affect if you actually have any troops at all, other wise it has none.

In the case of ((N+S)*C*H*R - T) the culture ratio always has an afffect on the chance of a flip, regardless of the presence of troops.

In case Soren can't take the time to answer, I think we can assume the latter. I can't see that he would set it up so that culture's importance is totally negated by lack of troops. I will edit the first page formula to reflect this change.

[Capt Dizle]

edited out...

[DeepO]

Originally posted by Fitz
In the case of: ((N+S)*C*H - T*R) the culture ratio only has an affect if you actually have any troops at all, other wise it has none.

In the case of ((N+S)*C*H*R - T) the culture ratio always has an afffect on the chance of a flip, regardless of the presence of troops.

In case Soren can't take the time to answer, I think we can assume the latter. I can't see that he would set it up so that culture's importance is totally negated by lack of troops. I will edit the first page formula to reflect this change.

darn... I have to give it to you, you sure know how to foresee consequences of a certain formula! You are right, I think, culture should have an impact on the foreing citizens+tiles, not the troops.

But, doesn't this mean the chance is getting too small to explain the number of flips we see in games? We do know that the capitalfactor should be 2000 if the city is in the middle of the two capitals, given the example of 4 C+S, double foreign city culture, double YourTotalCulture, no WLTKD or revolt, 2 troops it only gives 0.1% chance (as opposed to 0.4% with formula 1). Is this realistic?

Possibly Soren hasn't thought about it so much as we have the past few hours (and it still is formula 1)

DeepO

[Lucilla]

Edited out, because jimmytrick's post has been edited out as well

[Cyclotron]

Originally posted by jimmytrick
Witholding the formula as Soren has, dropping hints, and generally frustrating fans shows the PR mindset of Firaxis.

Soren, you need to grow up.

Uhhh... Soren was the one who gave us this formula in the first place. Within a few hours, he responded to the questions on the thread. This is pretty damn good PR, if you ask me...

Now, I have a new list. I'll check the game credit and if Soren is shown, I won't buy.

Good for you.

[DeepO]

jimmytrick, could you please shut up? Soren is not required in any way to give this info, he is not under Apolyton contract... and we had a nice thread going. Guessing could be a fun game, you know.

okay, it would be very cool if he could just give this formula (and other things too), but why would he do it if every time he answers a question at least someone shows up to bad mouth him? That was not really a constructive post to this thread.

DeepO

[Fitz]

Sorry, in the situation you describe I get .7% with (4*2*1*2 - 2)/2000 and .35% with (4*2*1 - 2/2)/2000. Still small, yes, but I almost never see flips myself until we talk about a situation like this:

Greek city, size 10, recently conquered, 3 squares out of 21 in Greek border, ~1.5:1 culture (mine higher), 3 garrison units. Distance easily 4 times closer to their capitol (assume D=500)

((10+3)*2*2/1.5 - 3)/500 = 6.3%

I can't remember how long it took to flip (1 turn or three), but in three turns, the chance of flip is: 100% - (100%-6.3%)^3 = 17.8%. Maybe that is a little low, but remember I'm still pulling the 1.5:1 ratio form memory. May have been as high as 2:1 against me.

jt: Thanks man. I appreciate the fact that you are willing to respond to politely worded private requests.

[DeepO]

Originally posted by Fitz
Sorry, in the situation you describe I get .7% with (4*2*1*2 - 2)/2000 and .35% with (4*2*1 - 2/2)/2000. Still small, yes, but I almost never see flips myself until we talk about a situation like this:

I stated YourCulture (so I meant human culture) twice enemyCulture, giving (4*2*1/2-2)/2000=0.1% as opposed to (4*2*1-2*2)/2000=0.2%... I agree I was wrong with the second part

Greek city, size 10, recently conquered, 3 squares out of 21 in Greek border, ~1.5:1 culture (mine higher), 3 garrison units. Distance easily 4 times closer to their capitol (assume D=500)

((10+3)*2*2/1.5 - 3)/500 = 6.3%

I can't remember how long it took to flip (1 turn or three), but in three turns, the chance of flip is: 100% - (100%-6.3%)^3 = 17.8%. Maybe that is a little low, but remember I'm still pulling the 1.5:1 ratio form memory. May have been as high as 2:1 against me.

I agree.... wait a moment, shouldn't resisting citizens count double? I remember something like that from Dan's thread a while ago. This would make it even more acceptable.

just a sec, I'll look it up
DeepO

[DeepO]

Originally posted by Dan Magaha FIRAXIS
A) The number of foreign nationals in the city in question (resisters are counted twice),

Yep, here it is. If this info still holds (and it should be, it's quite logical), the chance of your city flipping are even higher then what you had. In a city of 10, there are most likely 5 resisters...

Sorry to mess up your formula again
DeepO

[Fitz]

Don't worry about it. Last time I played with formulas (SMAX eco-damage formula) we had to edit the formula for 4 pages worth of thread. Just so long as it's as close as we can resonably get in the end.

I will edit to reflect resistors. Not sure how to phrase that so it's not too confusing. Maybe just add to the N factor a note that resistors count double.

edit: So to stick with the previous example and numbers, reflecting that there are 5 resistors at double, you would have ((15+3)*2*2*(1/1.5) - 3)/500 = 9% per turn, and over three turns a 24.6% chance. That is fairly significant if you ask me. Even once the resistors are gone and the disorder dealt with, if nothing else changes it is still a 2.8% chance per turn, or ~25% over ten turns.

[Soren Johnson]

Originally posted by Fitz
Thanks Soren. I really appreciate you taking the time to answer that. It would be cool if you could tell us if the culture ratio affects the # of troops (as in my revised formula) or the number of squares and citizens (as in DeepO's posed formula). I just realized how important that is in determining the chance of a flip.

In the case of: ((N+S)*C*H - T*R) the culture ratio only has an affect if you actually have any troops at all, other wise it has none.

In the case of ((N+S)*C*H*R - T) the culture ratio always has an afffect on the chance of a flip, regardless of the presence of troops.

In case Soren can't take the time to answer, I think we can assume the latter. I can't see that he would set it up so that culture's importance is totally negated by lack of troops. I will edit the first page formula to reflect this change.

it is the latter...

[Fitz]

Sweet. Thank you very much Soren. Always nice to know when you hit the mark. Now I'm going to have to learn to start estimating culture ratios ...

I can tell culture must be your personal baby. Always seems like that's what actually draws a remark from you.

[DeepO]

Originally posted by Fitz
Don't worry about it. Last time I played with formulas (SMAX eco-damage formula) we had to edit the formula for 4 pages worth of thread. Just so long as it's as close as we can resonably get in the end.

You had an SMAX eco-damage formula? I should have visited the apolyton site more often in those days... but was too busy with the Black&White forums on Lionhead instead.

I will edit to reflect resistors. Not sure how to phrase that so it's not too confusing. Maybe just add to the N factor a note that resistors count double.

Should do nicely...

edit: So to stick with the previous example and numbers, reflecting that there are 5 resistors at double, you would have ((15+3)*2*2*(1/1.5) - 3)/500 = 9% per turn, and over three turns a 24.6% chance. That is fairly significant if you ask me. Even once the resistors are gone and the disorder dealt with, if nothing else changes it is still a 2.8% chance per turn, or ~25% over ten turns.

Yes, I agree that such a city is very susceptible to flipping, but was more interested in whether the other formula gave other results.
Why is there the second '2', was the example with a city in disorder? That should never happen if you want to keep it... but okay, the other formula gives ((15+3)*2*2-3*1.5)/500=13.5% (!) over 3 turns that gives 35.3% chance... a rough estimate, as the number of resisters will go down over time with 3 troops inside, but still, you'd loose about one third of your cities like this!
Hmmm... the example is not very decisive on which formula to use, whether you have 24.6% or 35.3% you wouldn't notice. We should do a monte-carlo simulation on that one to figure it out (not tonight, thank you )

Taking the above example, if you'd have a WLTKD (just to see whether this is a big factor to consider): ((15+3)*2*0.5/1.5-3)/500=1.8% or 5.3%, taking the other formula: ((15+3)*2*0.5-3*1.5)/500=2.7% or 7.9%
Whichever formula it is, it seems like the WLTKD against disorder is another major factor! If you want to deminish the chance for a flip, making a WLTKD is a very good tactic. Sure, it will require micromanagement to ensure you have enough entertainers, but this pays off!
DeepO

[DeepO]

Damn, cross-posted... Thanks a lot Soren, if only I had noticed it a few minutes earlier, it would have saved me of doing my calculations twice

So no Monte-Carlo's then

DeepO

[Coracle]

In the case of ((N+S)*C*H*R - T) the culture ratio always has an afffect on the chance of a flip, regardless of the presence of troops.

Originally posted by Fitz
Sweet. Thank you very much Soren. Always nice to know when you hit the mark. Now I'm going to have to learn to start estimating culture ratios ...

I can tell culture must be your personal baby. Always seems like that's what actually draws a remark from you.

Why the game really should have been called "Culture" - but NOT "Civ 3". It could also have been called "Ethnic Cleansing", as that is what this culture flipping nonsense encourages the player to do via razing, which is also preposterous: a single unit can make a city of millions and its population disappear in an instant. Yea, surrre.


"REGARDLESS OFTHE PRESENCE OF TROOPS"???

That is the "latter" paragraph Soren refered to.

This is absolutely absurd, non-historical, and ridiculous. No wonder a town of mine of '5' once flipped with EIGHT samurai as garrison, all of whom vanished into thin air without even knocking off a population point or even putting the town in disorder! And that was with 1.21!!

[DeepO]

Originally posted by Coracle

"REGARDLESS OFTHE PRESENCE OF TROOPS"???

It's always easy to pull comments out of their context to gain advantage with, it surprised me it took so long until you found one that you could you use as an argument. If you'd understand what's really behind this formula, you might have noticed that troops are a big factor, but not the most important factor. If you don't like that, go play Quake.

No wonder a town of mine of '5' once flipped with EIGHT samurai as garrison,

Well, if your culture is very low, or you have a disorder on your hand, that culd indeed be the case. If of those 5 citizens, there were 5 resistors, you actually had 10 foreign citizens worth in there. If that's just besides the enemy's capital, far away from you, there is indeed a big chance it flipped. So what?

I can't understand why you don't appreciate the effort were making of understanding flipping, so that thickheads like yourself could do their advantage with it.
And could you please stay on topic? If you have anything to contribute to the formula, please do.

DeepO

[Capt Dizle]

umm, Coracle is right. Presence of significant garrison should be the primary factor. Still if we must have flippin, we have to have some other resolution other than a disappearing stack of troops. No formula will justify this insanity.

[DeepO]

That may be a good suggestion, but it still belongs in another thread

DeepO

[Tuberski]

We have 3 threads on this culture flipping crap.

ENOUGH IS ENOUGH!!!

[Capt Dizle]

Why? Your work on the formula has revealed a great flaw in the game, we are in your debt guys.

[Cyclotron]

Originally posted by jimmytrick
Why? Your work on the formula has revealed a great flaw in the game, we are in your debt guys.

Enough with the "flaw" stuff. It's your opinion, so stick to the other 2 opinion threads.

[Fitz]

My point about the troops factor, since you insist on taking it out of context Coracle, was that if the Ratio of the Cultures of the two civilizations were multiplied by the troops, when there were no troops it would be reduced to 0, and thus become a non-issue in the formula. I couldn't see that Soren or any other Firaxian would consider eliminating the ratio of Cultures from the formula when there were no troops present (correctly apparently). This doesn't mean that troops are unimportant. The fact remains that they are a huge factor.

If you want troops to have a bigger relative squashing effect on the chance of a flip, you have several options available to you. You can do any of the following, all of which have an actual reduction effect on the chance of a flip as well:

1) Reduce the number of foreign nationals in the city. Draft, starve, build workers, build settlers, rush (despotism & communism only).
2) Reduce the number of foreign nationals in resistance in the city. Put more military troops in the city to do this quicker.
3) Reduce the number of squares in the production radii that are in another civ's territory. Enhance the city's culture for quicker border expansion or eliminate neighboring cities.
4) Get more culture in the city than any other civ has had in that city in the past. Build more cultural producing improvements.
5) Stop civil disorder or produce WLTKD in the city.
6) Enhance your total civilization's culture relative to the other civ's total culture.

Now, you can choose to do the negative/militarily offensive of these if you so choose and find them easier (ie raze nearby cities and starve workers), or you can do the more peaceful (enhance luxuries & total civ culture, expand borders quickly). I prefer the second option myself. And I recognize the fact that troops are still critical to stopping a culture flip, and indeed are the primary factor in preventing a flip. It is certainly the easiest to control.

DeepO, think I should add this list to the front page?

Jt,I actually agree that some (even many) of the aspects of the culture flipping are unrealistic and even annoying. I find that in general it pushes the player in one of two directions: Not warmongering or genocide (razing & starving). However, given the fact that it is, and probably will remain, a game mechanic of CivIII, I prefer to understand how it works and try to prevent it and if possible use it to my advantage.

Regardless, I completely agree that troops should not be destroyed. I think they should be returned to the capitol, pushed out of the borders, whatever, or at least have a chance of doing so.

[Tuberski]

Um, I haven't worked on the formula, it doesn't really help all that much.

The only flaw in the game is what you take into it. If you don't like it, it's a flaw. Fine.

But don't call me or others idiots or fools because we DO like it.