D=2000, in flip formula

[Lt. 'Killer' M.]

Yes you are

 

[Gothmog]

You are correct that it does not matter what the % of foreigners is and I agree that it is a bit strange that way. However, if it were % it would be trivial to add workers to a just captured city and reduce the chance of flip to zero with a small G (as you imply). The programers probably saw that as a bit of an exploit (i.e. that an added worker would be almost as useful as an additional garrison in reducing flip chances, and without the chance of losing military in case it does flip).

Re Cartouche Bee and Yndy: I was trying to say what Yndy said, that the value of D was capped at 8000 for a situation where the civ didn't have a city. But it is not unreasonable that the game takes the distance of the ship (or last settler, wherever it is) into account as if it were the capital. Is that what you mean when you say they had a better chance than if they had a lone city? How did you come to that conclusion.

 

[sumthinelse]

Originally posted by Gothmog


P=[(F+T)*Cc*H*(Cte/Cty) - G]/D

where:
P = probability that it will flip this turn
F = # foreigners, with resistors counting double
T = # working tiles under foreign control
Cc = 2 if foreign civ has more local culture than you, 1 otherwise
H = .5 for WLTKD, 2 for disorder, 1 otherwise
Cte = Total culture of the foreign civ
Cty = Total culture of your civ
G = # garrison units
D = factor based on relative distance to capitals

What does "local culture" mean exactly ? Within 15 tiles? 10 tiles? And is "local culture" how much culture the "local" cities are generating now or how much these "local" cities have ever generated?

 

[satchel]

"Local culture" refers to the number of culture points generated in the city in question. Culture points accumulated while the city is in your possession is your local culture; culture points accumulated while the city is in another civ's possession is theirs.

 

[sumthinelse]

In the formula, is D the distance to 1) your capital or 2) the nearest palace? If I build the Forbidden Palace near a city will that help?

 

[Lt. 'Killer' M.]

sumthinelse: I think it is only the Palace, not the Forbidden palace.

This would also easily explain why the flip chance is 0 for the capital.

If I rush the FP in conquered lands, I frequently have flips. If I jump the Palace there I don't. I have been using the jump-Palace strategy for a bout 3 months now and haven't had a single flip with it!

 

[Gothmog]

I agree with killer that it D only counts the Palace and not the Forbiden palace, but this is only from my experience and not hard data. However, this does not explain the 0 flip chance for the capital, because of the 500 min value for D. It does jibe with the fact that a city containing the FP can its self flip.

 

[gozpel]

In my last deity game I captured the Iroquois capitol with the Great Library and Lighthouse...I built a temple right away, paid for it as I was in Monarchy, so no whips.
Then 11 turns later it flipped, swallowing up my 9 immortals, one pikeman and my army of 3 immortals. At the time there was 5 Iroquois citizens and 1 Persian in the city...3 happy, 2 content and 1 unhappy and they had 2 luxuries...
So what's the formula for a stupid flip like this? I thought by taking their capitol I would've chomped off a great bit of their culture, but that didn't help much. I couldn't believe my eyes and just hit quit and went for a nice walk instead

 

[Lt. 'Killer' M.]

gozpel: check the formula in the FAQ thread.

You only kill the ADDITIONAL culture they would get from the town, what HAS BEEN produced stays until the civ dies.

 

[sumthinelse]

Originally posted by Lt. 'Killer' M.
sumthinelse: I think it is only the Palace, not the Forbidden palace.

This would also easily explain why the flip chance is 0 for the capital.

If I rush the FP in conquered lands, I frequently have flips. If I jump the Palace there I don't. I have been using the jump-Palace strategy for a bout 3 months now and haven't had a single flip with it!

It would be nice to have 2 Great Leaders - one for the new Palace and 1 for the new FP. But the game won't let you save 'em.

 

[Lt. 'Killer' M.]

sumthinelse: that is why I build the FP close to home.

if I do not have an FP yet, then the game is so young that noone has serious culture so i cna rush the FP abroad....

 

[gozpel]

Originally posted by Lt. 'Killer' M.
gozpel: check the formula in the FAQ thread.

You only kill the ADDITIONAL culture they would get from the town, what HAS BEEN produced stays until the civ dies.

Ahh, of course. I should have phrased it killed of their "producing" culture or something like that...since I got their GL and Lighthouse?

And the formula doesn't make much sense in a case like this as I've got the temple right away and wouldn't the wonders help to avoid flipping?

Ah well, I took it as one of those "mi****s" that happens now and again...and after all I got at least 10 techs from the library before I lost it again. What I felt bad about was losing all those units really, when the pop-size of the town was less than half of the number of units fortified in there..That was over half of my offensive troops that just went M.I.A

 

[Lt. 'Killer' M.]

gozpel: the wonders won't help as they give you no culture. The temple hepls a bit, but a number of foreigners is a very bad thing, it makes the chance very high here.

Let's see:

P=[(F+T)*Cc*H*(Cte/Cty) - G]/D

where:
P = probability that it will flip this turn
F = # foreigners, with resistors counting double 5
T = # working tiles under foreign control let's say 4?
Cc = 2 if foreign civ has more local culture than you, 1 otherwise 2 - very probably. if it had wonders then it was old, and the other cities around are, too.
H = .5 for WLTKD, 2 for disorder, 1 otherwise 1
Cte = Total culture of the foreign civ let's say you are 1.5:1?
Cty = Total culture of your civ
G = # garrison units 14
D = factor based on relative distance to capitals let's say it is twice as close to theirs: 0.5

gives:
P=[(F+T)*Cc*H*(Cte/Cty) - G]/D
P=[(5+4)*2*1*1.5-14]/0.5=26

that 26 means it will flip i around 25% of cases.... Now this means you were lucky it didn't go earlier.....

 

[gozpel]

Originally posted by Lt. 'Killer' M.
Now this means you were lucky it didn't go earlier.....

Yikes So what's the solution? I was trying to get enough money for a colosseum, but didn't have time...and somehow razing a city with 2 wonders is beyond me...building temples and stuff seems to be just a waste of both time and gold....except when conquered cities are very close to your own main territory.

Hmm...maybe with early wars I should just stay within Despotism and whack those ppl!

 

[Lt. 'Killer' M.]

1) starve them down to one guy
2) LOTS of troops (if you don't have enough, see 5))
3) take/raze adjacent cities (reduced tiles under foreign control)
4) build you Palace in that city (rush with a leader). It should usually be a good place if you previously build your Forbidden palace in your core empire).
5) Stack troops just OUTSIDE the city - you lose fewer then6) kill the enemy all the way - no more flip risk then.

 

[gozpel]

Ok thanks, didn't really continue the war because of all their troops swarming around and deity is quite tricky. But I will start to starve those poor buggers again, I remember that was the best tactic before but I am too soft sometimes

Anyways I'm off to try on the new GOTM and thanks again. Much appreciated info.

 

[TheNiceOne]

Originally posted by Lt. 'Killer' M.

P=[(F+T)*Cc*H*(Cte/Cty) - G]/D
P=[(5+4)*2*1*1.5-14]/0.5=26

that 26 means it will flip i around 25% of cases.... Now this means you were lucky it didn't go earlier.....

Now, using this formula isn't easy (so no wonder the civ3 program miscalculates and flips my cities that should never flip ), and I think you made two errors Killer:

If I understood your base number correctly, you guessed that gozpel's total culture was 1.5 times the AI's, but in the formula you have used an AI culture 1.5 times gozpel's.

You also make an error with the D. This isn't as simple as De/Dy, but rather (2000*De/Dy).

So the revised calculation gives:
P=[(5+4)*2*1/1.5-14]/1000 = < 0. Which means zero chance of a flip. Now, since it flipped, this either means that the formula was wrong, or that some of the guesses (number of enemy-controlled tiles, culture ratio or distance factor) was wrong.

If we go back to the AI hving 1.5 times gozpel's culture (which may well be since the AI had some wonders), then the formula becomes:
P=[(5+4)*2*1*1.5-14]/1000 = 1.3%
This means that the city had 1.3% of flipping each turn. Not big, but possible. 11 turns of this equals close to 15% chance of flipping, so you were still a bit unlucky, but not extermely unlucky.

 

[TheNiceOne]

Originally posted by gozpel
Yikes So what's the solution? I was trying to get enough money for a colosseum, but didn't have time...and somehow razing a city with 2 wonders is beyond me...building temples and stuff seems to be just a waste of both time and gold....except when conquered cities are very close to your own main territory.

There are numerous threads that discuss what to do about flipping. Search for them. Your total culture is the biggest, but that one need a long time building up.

After having captured such I city, I'd suggest starving it down to 1 foreign citizen, making sure none of its 21 squares are controlled by the enemy, and then maybe adding 5 of your workers and getting it in a WLTKD day. If you do these three, even at a 1:4 cultural disadvantage you will be safe from flipping with as few as 4 military units.

 

[Stapel]

Is anyone 100% certain yet about the D value?

I think Both TNO and Killer are reliable posters, but there is a big difference this time...

I would like to know exactly (like many others).

 

[TheNiceOne]

Originally posted by Stapel
Is anyone 100% certain yet about the D value?

I think Both TNO and Killer are reliable posters, but there is a big difference this time...

I would like to know exactly (like many others).

Well, I've got it directly from Soren Johnson's (of Firaxis fame) keyboard that the D value is 2000 on average and ranges from 500 to 8000 dependant of the distance of the two capitals. He wrote this in a thread on Apolyton, which catt linked to on the first page of this thread.

Soren did not write exactly how D is computed, but my educated guess is D=2000*cap(De/Dt). Where cap is a function that caps the De/Dt ratio to be max 4 and min 1/4. Such a formula suits Soren's description perfectly.