Questions about fortified defending

[maxbacsi]

Hi all!

I have a few questions about the defense-improving techniques of the game. Although I play CIV for a long time, I still dont know all the algorythms and trivia about it.

I have read several manuals and discussions mentioning the factors modifying the defense strenght of a unit. All of them states that the bonus from the veteran status and terrain defense modifier are stacked with each other and any other defense-improving factors. It is also stated clearly, that a Fort and a City (thus the City Walls as well) cannot exist on the same square, which means that Fort 2x modifier ("mod." below) and City Walls 3x mod. dont have a chance to stack.
Which remains unclear, that if the 1,5x mod. of the units Fortified status stackes with Fort or City Walls mod. Some guides say it wont, some other say it will. How does this exactly work?

It is also said that two units, Artillery and Bomber ignore the City Walls mod; the unit will not get the 3x for its defense when attacked by one of them. However, it is not mentioned that these units will ignore the Fort as well, or not. The same question comes for the Fortification status mod.

I also have a bit more complicated question about this. If these special units do not ignore Fortification status mod., and Fortification status mod. do not stack with City Walls or Fort mod., what will happen when a fortified unit standing in a city with Walls is attacked? Will the Fortification status mod. take place instead of the ignored City Wall (or maybe Fort) mod.?

Thank you for your help

 

[loopinaloop]

I don't know exact numbers as well but from my experience all bonuses are summarised as in example:
+50% for river
+50% for fortification
+50% for veteran status
+200% for city walls
gives +350% and that means that Phalax has 7 defense, giving it a little bit better then catapult. From my experience, non-veteran catapult more often looses against such fortified veteran Phalanx behind a City Wall then wins.

If I'm wrong the prove it, please.

 

[Mize]

Actually, I think it's more along the lines of:
base defence for phalanx = 2
* 1.5 (50%) for veterancy = 3
* 1.5 for fortification = 4.5, rounded down to 4
* 1.5 for river bonus = 6
* 3 for city walls = 18

Which, if so, is quite formidable and explains why we see the phalanx-sinks-battleship phenomenon so often.

Otherwise, some pretty good questions in this sadly neglected thread. I also would like to know whether or not bombers and artillery ignore forts as they do city walls, but that would require extensive testing, a task I am not up to. It's not really a big gameplay issue, since the AI doesn't build forts and only attacks your forts under certain conditions, like when it thinks it has a chance. The AI seems to be very aware of its odds and often chooses not to attack a well entrenched unit.

Oh, and here's some more calculations on defence ratings in a very ancient thread on this forum:

http://forums.civfanatics.com/archive/index.php/t-2036.html

 

[Last_Evolution]

Actually, I think it's more along the lines of:
base defence for phalanx = 2
* 1.5 (50%) for veterancy = 3
* 1.5 for fortification = 4.5, rounded down to 4
* 1.5 for river bonus = 6
* 3 for city walls = 18

Which, if so, is quite formidable and explains why we see the phalanx-sinks-battleship phenomenon so often.

Otherwise, some pretty good questions in this sadly neglected thread. I also would like to know whether or not bombers and artillery ignore forts as they do city walls, but that would require extensive testing, a task I am not up to. It's not really a big gameplay issue, since the AI doesn't build forts and only attacks your forts under certain conditions, like when it thinks it has a chance. The AI seems to be very aware of its odds and often chooses not to attack a well entrenched unit.

Oh, and here's some more calculations on defence ratings in a very ancient thread on this forum:

http://forums.civfanatics.com/archive/index.php/t-2036.html

Ever heard of random numbers? That's what matters.

 

[Mize]

Ever heard of random numbers? That's what matters.

Ummm... Yes. But we were talking about the numbers at play when defence bonuses stack, those are most decidedly not random. The only "random" thing in Civ is the outcome of the battle and that too is based on simple mathematical odds, which should, in theory, be testable. In the particular case of the phalanx-on-a-river-behind-city-walls vs. a battleship the odds are even or 27/18 in favour of the battleship if it has veteran status.

 

[killertux]

so Here are my answers based on "Sid Meier's Civilization or Rome on 640k a day"

Page 256 it says that defender can get only ONE of following bonuses.
-fortified (using F key), defence*1.5
-in fortress square, defence*2
-in city that has city walls(unless defending unit is air unit or attacker is bomber or artillery), defence*3

Page 257 says that if unit is behind city walls(and fortified by F key) but attacker is artillery or bomber then fortified bonus(using F key) will apply.

Page 258 has possible error in it because example 3 says fortified bonus(using F key) would not apply on units behind city walls(if city walls are ignored).

 

[Mize]

Killertux, you just might be right. I've read this both here somewhere and on Apolyton. It does make some sense, the 'F' key being some sort of temporary fortification like trenches, palisades etc., made redundant by a fort, made redundant by a fortified city. I think there's an easy way to test it. Build a fort or a city with walls on plains or grassland (no defence bonus) and put inside a veteran rifleman and a green mech inf. I'll use the fort for my calculations (100% defence bonus).

The veteran rifleman has 7.5 defence = 15 with fort, or 14, doesn't matter in this particular case
The green mech inf. has 6 = 12 with fort

When the enemy attacks the rifleman should defend. Now fortify the mech inf. and leave the rifleman active. If you and Wilson and Emrich are right, the rifleman should defend the next attack too. If I'm right and 'F' key bonus stacks with walls, the mech inf. should defend, because
6 * 1.5 ('F') = 9 * 2 (fort) = 18

---EDIT---

Just tested a better case. I fortified a veteran rifleman on a hill in a fort. If 'F'-key stacks, the rifleman should have 7.5 * 2 (fort) = 15 * 2 (hills) = 30 * 1.5 ('F') = 45
I got a veteran mech. inf. in the fort and left it active. 9 * 4 = 36
As expected, the rifleman defended. Which is already a confirmation in and on itself, because if 'F'-key didn't stack, the veteran mech. inf. should defend as it outclasses the rifleman by 1.5 base defence. The second check also came out positive: I fortified the mech. inf. on the next turn and, as predicted, it defended. 36 * 1.5 = 54 > 45

Maybe I should thoroughly officialise this in the "Where Wilson and Emrich got it wrong" thread... Some other time.

 

[Mize]

I'm necroing this, because I'm not satisfied with just editing my last post. There are still unanswered questions in this thread. Like f. ex. do bombers and artillery ignore forts as they do city walls? Can any of you think of a proper test for this? The only thing I can imagine is making a custom map, giving the AI lots of bombers on its landlocked island city and giving the player lots of *insert proper test-unit* to put in a fort and in the open respectively. And then you can just try a statistical approach. If the fort defends against bombers and artillery, they should be losing twice as much as when attacking units in the open. Would this test work, or are odds in Civ more skewed than we assume?