Fortress on a mountain

[Digz]

What's the defense bonus?

I know being on a mountain gives a 100% defense bonus, but what happens when a mountain has a fort on it?

For example, let's pick a spearman:

Regular defense: 2

Mountain: 4

Mountain w/ fort: 6 or 8?

I could see it being 6 because it's a 100% from the unit's defense (which would be 2), but i'm thinking it would probably be 8.

Anybody know the answer to this?

 

[Chamnix]

Bonuses are added first then multiplied by the base defense. I think a fortress is +50% defense, so a spearman in a mountain fortress would be:

2 * (1 + 100% for mountain + 50% for fortress) = 2*(2.5) = 5

 

[Flak509]

Haha i works well with 2-3 Spearmen and 2-3 Catapults normally the AI aviods them like the plauge

 

[ZzarkLinux]

Echoing what Chamnix said. I know what the OP is thinking, and it's tricky.
You just gotta add all the bonuses into a total bonus:

BASE * (1 + %bonus + % bonus + %bonus etc...) = INTEGER RESULT

Like a numidian merc fortified in a hills city:
3 * (1 + .25 + .5 + .5) = 6.75 = 6
Fortification didn't make a difference in this case (125% bonus vs. 100% bonus).

EDIT:
Proven wrong below

 

[Northen Wolf]

Like a numidian merc fortified in a hills city:
3 * (1 + .25 + .5 + .5) = 6.75 = 6
Fortification didn't make a difference in this case (125% bonus vs. 100% bonus).

It is rounded down? not rounded to nearest integer (7)? (just wondering)

 

[Othniel]

BASE * (1 + %bonus + % bonus + %bonus etc...) = INTEGER RESULT

Are you sure the decimal is truncated off? Combat calculators such as this one keep the decimals. I've always assumed this calculator was correct.

 

[ZzarkLinux]

RE: Northern Wolf and Othniel

I have assumed my calculations based on 2 things:
a) my game experience
b) my programming experience

That's my reasoning, which is by no means in-fallable. I could be wrong. If I am please let me know.

My reasoning:
a) my warriors do pretty well against enemy warriors across a river
CASE 1: There is no 50% bonus (DEF 1.5 -> DEF 1)
CASE 2: Cognitive bias
b) I have coded C++ programs for 6 years now, and have am doing it professionally.
--I have never written a proper Rounding function. I rely on INT-TRUNCATION,
--which simply means computer just "drops the decimal" eg 6.583 -> 6
CASE 1: I'm an average programmer, just like FIRAXIS, and they use easy INT-TRUNCATE rounding
CASE 2: I'm just lazy programmer, not like FIRAXIS, and they use a proper Rouding function

So I stand by CASE 1 myself
EDIT: I guess I'm just a lazy programmer !!

EDIT:
For the programmers out there, I use tolerances. This method applies to more than just "INT-TRUNCATION", it also works when combining decimals.

Remember, binary doesn't always perfectly convert to DECIMAL
ie the standard number 5.258 (DECIMAL) may be an infinite-repeating number 101.1010110110101101010.... (BINARY)

This is annoying because simple stuff may fail if BINARY <-> DECIMAL conversion is involved
ie the program would think 5.258 + 2.242 != 5.5 because the program compares 5.5000001984 != 5.5
But having a tolerance would fix the above comparison

 

[narsqrd]

Remember, binary doesn't always perfectly convert to DECIMAL
ie the standard number 5.258 (DECIMAL) may be an infinite-repeating number 101.1010110110101101010.... (BINARY)

This is annoying because simple stuff may fail if BINARY <-> DECIMAL conversion is involved
ie the program would think 5.258 + 2.242 != 5.5 because the program compares 5.5000001984 != 5.5

this is why computers will never conquer humanity and rule the world

 

[anaxagoras]

I have assumed my calculations

Why assume anything? If you have access to the editor, this is a testable hypothesis. Do warriors defending behind a river or on a hill get a bonus, or don't they. I use a Mac, and I don't plan to buy Windows, so I can't run this test. But the test would be easy to design and run.

 

[Rodent]

I thought it was simple, Add the Bonus Percentages (In this Case, 100% + 50%=150%) and apply that to the base value

150% of 2 is 3 and 2+3=5

 

[ZzarkLinux]

anaxagoras I'd love to test but I'm very busy (as I mentioned earlier).
Besides, I don't like working with confidence integrals I'd have to dig through my old binder...
But still... dang... that test sounds pretty fun though... Maybe tonight.
I'll even make it so the AI warriors attack my warriors, so it's easier on me.
(This way we can see if the RNG hates players too )


Rodent, it's all in the wording:

"Spearmen gets 50% bonus"

Spoiler :

Spear + 50% = 2 + 1 = 3

"Spearmen fights at 100% strength"

Spoiler :

Spear = 100% + 0% = 2 + 0 = 2

"Spearmen fights at 150% strength"

Spoiler :

Spear = 100% + 50% = 2 + 1 = 3

"Spearmen receives 150% bonus"

Spoiler :

Spear + 150% = 2 + 3 = 5 = 250% original strength

 

[ZzarkLinux]

Here is a C3C test file.

Game Stats:
- Chariots only have 1 move to avoid retreat
- Promotions all = 3HP ie promotions are meaningless
- Preserve Random Seed is off
- Respawn AI is off
- Culture link-locations is on
- Animated battles are on

Units Stats:
- Human has 3 cities, enemy has 1.
- Human has 25 warriors and 25 Chariots (w/o retreat)
- Enemy has 50 warriors (+2 for the token city defenders)

How To Test saved game: (zipped)
- save the delMe.sav game to Conquests\Saves folder
- Open C3C and load game
- Your units are defense 1 units fortified on a hills = +75% bonus (total strength = 175% of 1)
- Declare war on France before hitting enter.
- France will march 50 warriors next to your 50 units.
- Press enter and France will attack
- Analyze: did your units win as if they had defense 1, defense 2, or defense 1.75?

The actual scenario file is zipped as well, so you can check the rules.

My 1st test:
- France 15 warriors survived (out of 50)

My 2nd test:
- France 12 warriors survived (out of 50)

I swear, shouldn't France be winning more agains my fortified 1-defense on a hill?


BTW I'm pretty much out of pocket now. Ocean City with 20 friends.
In fact I haven't really had a free weekend since mid-may.

 

[Mathias]

Your results clearly indicate that the defending unit has a strength of 1.75, which gives the attacking units, with a strength of 1.0, a 74.33% casualty rate. On each roll, the attacker has a 1 / (1 + 1.75) = 4/11 chance of winning, while the defender has a 1.75 / (1 + 1.75) = 7/11 chance of winning. In the breakdown of the possible combat results below, A = 4/11 and D = 7/11.

Attacker wins, unscathed:
A*A*A = 64/1331 = 4.81%

Attacker wins with 1 HP lost:
A*A*D*A = 448/14641 = 3.06%
A*D*A*A = 448/14641 = 3.06%
D*A*A*A = 448/14641 = 3.06%
Total = 1344/14641 = 9.18%

Attacker wins with 2 HP lost:
A*A*D*D*A = 3136/161051 = 1.95%
A*D*A*D*A = 3136/161051 = 1.95%
A*D*D*A*A = 3136/161051 = 1.95%
D*A*A*D*A = 3136/161051 = 1.95%
D*A*D*A*A = 3136/161051 = 1.95%
D*D*A*A*A = 3136/161051 = 1.95%
Total = 18816/161051 = 11.68%

Overall odds for attacker winning:
41344/161051 = 25.67%

Similarly, for the defender winning:
Unscathed = 343/1331 = 25.77%
1 HP lost = 4116/14641 = 28.11%
2 HP lost = 32928/161051 = 20.45%
Overall = 119707/161051 = 74.33%

 

[ZzarkLinux]

Thanks Mathias !!

I had thought that fortifying spears in towns never made a difference, since the 25% bonus of 2 just became a lost decimal.

Guess I should pay more attention to fortified vs. unfortified spears now...

 

[AutomatedTeller]

yeah, if it didn't take into account non-integer bonuses, then a lot of things woudl be silly - like 10% on plain, or ever fortifying before gunpowder.