tweaks to the garrison formula

[etj4Eagle]

As we know the basic garrison formula is:

(foreignors + non-controlled work tiles) X Culture_factor = required garrison to prevent flip

with culture_factor being 2 if you have less culture in that city than other civ, and 1/2 if you have more. (all the other are applied after the garrison factor in changing this into a probability).

For the most part this is ok. However with large cities and before the first cultural expansion the number of required garrison to supress a flip can get rediculous. Since if you just took the city you are going to have 12 work tiles factored in, which means that you need 24 more garrison units. (ouch!). This means that a pop 1 city needs 26 units to prevent a flip till the first cultural expansion (and assuming it fully expands) then the number drops to 2.

While this may only exist for a few turns, I think the effect of tile you don't control is a bit too large. Some ideas might be to count them as only half a foreignor (much as a resistor is counted as two) or not to ably the cultural factor to them. If you did both, this would drop the units required for territory control down to only 6, a much more reasonable number than the 24.

The other issue comes with capturing any large city, as it become quite impratical to garrison enough defenders (time to mass produce workers to drop the pop). A pop 6 town would require 12 defenders till you managed to surpress the previous culture. And a metropolis would be near impossible to control with at least a garrison of 26 units required.

My suggestion would be to have the garrison work much the same way that luxeries do with a market place.
ie the first defender would be equivelent to 1 garrison unit
the second defender would be equivelent to 2 garrison units (for a total of 3)

With this method you are still required to put a fair number of defenders into newly acquired enemy cities, but you don't have to put ubsurd numbers into the huge cities.

Now this effect could require a small wonder or an improvemet to be built. Also it might be an interesting idea to make the effect of the garrison to be at least partially dependent on government. Maybe a unit acting as a military police give a bonus garrison of 1. Hence if you are allowed 4 military police, and stack in four defenders you would have a garrison strength of 14 verse 10 for zero military police.


A note to those replying: try to keep the thread on topic. This is not a thread for complaining about how "unrealistic" cultural flipping is or what not. It is about how to improve the garrison factor while remaining within the current design parameters.

 

[Zouave]

. . .My suggestion would be to have the garrison work much the same way that luxeries do with a market place.
ie the first defender would be equivelent to 1 garrison unit
the second defender would be equivelent to 2 garrison units (for a total of 3. . .)

So you don't want us to talk about how unrealistic it is. OK. So, I then assume from the above you want Firaxis to PATCH this problem since we cannot make garrisons function in that manner.
Your idea is GOOD - and don't hold your breath waiting fo Firaxis to change it. BTW, they can put an end to "razing" cities and "disapearing garrisons", too.

What the game really needs is a FEAR factor. Towns and cities have surrendered to overwhelming and brutal invading forces, and towns/cities have been cowed into non-resistance by efficient brutal garrisons NOT in absurdly high numbers.

 

[cracker]

etj4eagle,

Are you sure you are using the number of tile part of the equation correctly?

I thought this part of the equation applied to the number of tiles of your base area that were clipped by the cultrual boundary of another civ. So if you were at the base size of 9 tiles and one corner was clipped off by another civs boundary, then you whould have and increased garrison requirement to compensate. If you were just at an unimpeded 9 size, then there would be no penalty.

If the city expanded from 9 to 21 without countering the overlap from the other civ, then the culture incursion in the expanded case would jump from 1 tile up to 3 or 4 tiles depending on the overlap arrangement.

I am fairly sure that a nine tile city has no garrison penalty associated with not being a 21 tile city as long as there are no outside overlap losses that impede the 9 tiles.

(And yes Zouave, you just totally farted in et's soup by ignoring his hopes that we could discuss this issue without more of the same foot pounding rants we see in almost every post on this subject. Don't do this if you want us to respect you in the morning, especially when someone asks you specifically not to.)

 

[etj4Eagle]

Cracker it is possible that I miss-understood the post over at Apolynet on how cultural flipping works. If there is no penelty for not having had that first cultural expansion yet, then that is good. (have not really seen enough flips to say one way or the other).

So if you are right, then that second layer of 14 tiles does not matter till you get that second culture expansion. (guess I should go try to dig up that post on look at it more, it was complicated enough that I could have messed things up).

Even still I think some kind of tweak like I suggested for the garrison would be good for taking large cities.

 

[cracker]

Knowing that Soren is no longer updating features of the CIV3 AI and that the culture flipping has seemed to be in the same code area, I doubt we will see any updates or changes of this type.

The best we could hope for would be to beg for some sort of Garrison effectiveness number to be displayed in the city display screen so that at least we could see what our exposure to culture flipping might be in any given city. This would not be asking for secret forewarning of the event, which has been off the table, but would look for something like "percent chance of a culture flip" or something like that.

 

[etj4Eagle]

Cracker, after reading the Apolynet thread on the topic I still can't exactly make heads or tails of when to count those tiles, but you probably are correct. I also realized that I had left a few terms out of the formula. (so for those here who like me don't regularly read the threads over there or who otherwise are confused on this concept, here is the correct formula):

P=[(F+T)*Cc*(Ce/Cy)*H - G]/D
where:
F = # foreignors (resistors counted twice)
T = # working tiles under foreign control (ones that you should be able to work if not for enemy culture)
Cc = 2 if the civ in question has more local culture than you. 1 otherwise
Ce = The total culture of the civ in question
Cy = The total culture of your civ
H = .5 for WLTKD, 2 for disorder, 1 otherwise
D = distance factor

So now with the correction that enemy controlled tiles that are not in your civ radius are not counted, the resulting number of troops is much better. Since in most cases you will only have to be dealing with the local population. Of course for any decent size city that can be an issue (especially if the enemy civ has substantially more culture than you). Though, at least initiating a WLTKD can halve the required garrison size.

Even still I would probably engage in a policy of worker production for any large city I take and repopulate it with my own people. Since even under a WLTKD, a pop 7 city still requires 7 garrison if you have equivelant national culture.