Thread Tools
Old November 3, 2001, 01:29   #1
Valant
Settler
 
Local Time: 15:58
Local Date: October 31, 2010
Join Date: Nov 2001
Posts: 13
Combat System Explained
I posted this on Civ Fanatics Forum as well...trying to get as many eyes on this as possible because alot of the "UNKNOWNS" I really want to know, heh. Anyone can take this info for posting in other forums or websites, but please do not paste this info in this or Civ Fanatics forum which will create redundant threads.

What I want to do is to be able to make this thread the be all information on how combat works in Civ3. Most of this info I have comes from the instruction manual snipits, civilopedia, various websites and forum feedback. Really wished the manual was more specific on hard numbers used in calculations. If you have anything to add to this thread (like erroneous data or missing data) please do so and I will edit my origional post to import the new data.

HP:
The only different in conscripts, standard, vetran and elite units is the number of hit points they have. During battle each time the unit loses on a battle roll they loose 1 hp.
Conscript 2hp
Standard 3hp
Vetran 4hp
Elite 5hp

Terrain Bonuses:
When combat occurs over a river, the defender receives a 25% bonus.

The defensive terrain bonuses are as follows
10 Floodplains, Grasslands, Plains, Desert, Tundra, Coast, Sea, Ocean
25 Forest, Jungle
50 Hills
100 Mountains

How these numbers are used in the battle calculation are UNKNOWN.

By the looks of it terrain does not give any offensive bonuses what so ever.

Manmade Bonuses:
Cities & Walls
Walls give a ground defensive bombardment of 8. The way this value of 8 is used is UNKNOWN.
Units in cities of population of 6 or below do not receive a defensive bonus.
Units in cities of population of 6 or below that have city walls give a defensive bonus of 50%
Units in cities of population of 7-12 receive a 50% defensive bonus, reguardless of city walls.
Units in cities of population over 12 receive a 100% defensive bonus, reguardless of city walls.

Great wall
Supposed to double the effect of city walls to a 100% bonus to the defense of units, but once your city exceeds a population of 6, the option to build citys walls no longer appears in your production list. So right now its UNKNOWN what effect the great walls have on tows over population of 6.

By the looks of it walls and large cities do not give any offensive bonuses.

Fortresses
Build outside of cities. When a unit is stationed in a fortress it receives a 50% bonus to their defense.

Units stationed in a fortress receive an offensive bonus of being able to get "free shots" on the enemy. The specific definition of "free shots" is UNKNOWN. I have not seen it myself, but based on feedback I think this is what happens. Every unit inside of the fortress will automatically attack any approaching enemy unit (no limit) and not lose their turn. The free shot only involves 1 turn battle where if the unit in the fortress wins the roll, 1hp will be removed from the enemy unit outside of the fortress. If the unit in the fortress loses, no damage will occur on wither unit. Each turn while the enemy unit is next to the fortress, each unit inside of the fortress with get a free shot on each enemy unit next to the fortress. Once again I THINK this is what happens.

Airport
Provides NO attack/defensive bonuses.

Coastal Fortress
Provides a naval defensive bombardment value of 8. The way this value of 8 is used is UNKNOWN.
This provides a 50% defensive bonus to all units stationed in the city against naval attacks. Plus a costal fortress will take pot shots at passing enemy sea vessels. The specific definition of "pot shots" is unknown, but I THINK its alot like fortresses. Each turn the fortress fires on nearby enemy ships using an attack factor of 8. If successful the offending sea vessels suffer 1hp of damage. The fortress will be able to fire on all enemy ships within range, which I believe is 1.

SAM Missle Battery
Civilopedia simply states that it attempts to attack air units that are attacking the city its in. Hard data on how this works specifically is UNKNOWN.

Strategic Missle Defense
Has a 75% chance to destroy incoming ICBMs.

Unit Abilities:
Retreat
Units that have a higher movement modifier than the opponent and have no used all of their movement points can retreat when their hp drops to 1.

Fortify
A fortified unit receives a defensive bonus of 50%, but no offensive bonuses.

Bombarbment/bombing
(Have only tried bombardment a little so its hazy here)
The attack value of the boming unit affects its successful rate, but how exactly is UNKNOWN. Also what factors determine the probability that determines if its a city improvement, unit, city wall, population hit is UNKNOWN. Also I assume all successful bombarbments on a military units only results in 1hp lost. I recently found that bombardment units have a value called rate of fire. From feedback, rate of fire is the number of "projectiles" the unit can fire per attack turn. For example a unit with a RoF of 2 does 1 attack on a city, the city is affected by 2 hits (if both did hit).

Army
Units in an army can only move as fast as the slowest unit in the army. When an army attacks or defends the first unit continues to fight/defend until the unit is down to 1hp. After which the unit gets replace with the next unit. This continues until the army gets to its last unit where that one is destroyed when its hp reaches zero, but the other 1hp units remain undestroyed (albeit heavily damaged).

Nuclear Attacks
Has a 50% chance to destroy each unit caught in its blast area (reguardless if its a friend or foe unit) and reduces a city's population by 50%. Size of the blast area is UNKNOWN.

Other:
The affect of difficulty level has NO effect on attack/defense calculations reguardless if the unit is computer or player controlled.

The penalty (if it exists) for naval units caught in port is UNKNOWN.

The penalty (if it exists) for air units caught on the ground is UNKNOWN.

Final Calculations:
Bonus percentages add, so 2 50% bonuses result in a 100% total bonus. How the game deals with rounding is UNKNOWN.

The final caculation to determine the % chance for an attacking unit to win per turn is quite easy. However, what muddles this is that the vaules of the attackers offense and defenders defense vaules are UNKNOWN because the lack of information of how EXACTLY does ALL of the bonuses are totalled up is UNKNOWN. So we can only (at this time) calculate a crude approximation of turn win percentage. For example where the hell exactly is the 50 point defense bonus of hills used....is it +50 points somewhere in the grand equation or is it +50% bonus. Plus how does the game handle rounding is also UNKNOWN.

Crude Turn Win % = (attackers offensive rating)/ ((attackers offensive rating)+(defenders defensive rating))

If the unit wins the turn, it causes 1hp of damage to the opponent. If it loses the turn, its loses 1hp of damage.

Dusted off my only high school probability and statistics book and I think I found the correct formula to predict the crude final battle win %. After a quick scan I found the binomial equation, abeit if memory serves me right there is a better one when your looking for a specific number turns won and the number of "coin tosses" is a ranged variable.

* I am pretty sure there is better equation out there...help me find it guys *

X = (HP of both units added together - 1)
Y = HP of defender
Z = X - Y
A = Crude turn win % (in a .xx format so 50% would be .50)
B = 1 - A
C = (X-Y)!

Crude final battle win % = (X!/(Y! * C!)) * (A^Y) * (B^Z)

So for example an attacker of hp of 5 (elite) is attacking a defender of hp of 3 (regular). The attacker has a crude turn win % of 66% or .66

So the crude final battle win % would be (7!/(3! * 4!)) * (.66^3) * (.33^4) which equals 11.9% of the outcome being killing the defender in 7 turn battle. The 11.9% seems small but you need you need to add the percentage chance of winning in a 6, 5, 4 and 3 turn battle to get the real overall chance of winning the battle. So for each battle the number of times you have to run through the above equation as many times as the HP of the attacker. So for a 5hp on 5hp unit battle you will have to go through a 5,6,7,8 and 9 turn battles and add up their probability. Hence you would have the do this equation multiple times to add up the total chance to win overall....and also hence I state that there must be a better equation to get the total chance of winning the battle with only using 1 equation.

Well its late and I am tired so going to end it here without proofing.....please add what you can.

Last edited by Valant; November 3, 2001 at 12:19.
Valant is offline  
Old November 3, 2001, 02:55   #2
TheDarkside
Civilization IV Creators
Prince
 
TheDarkside's Avatar
 
Local Time: 11:58
Local Date: October 31, 2010
Join Date: Oct 2001
Location: New York
Posts: 586
dont forget the free shots associated with the former zone of control system. instead of not allowiung enemy units to pass by your unit's 'zone of control', civ3 allows the unit to pass but the unit passing gets to be shot at by the unit who is being passed, even though its not his/her turn.

Example: My cavalry is stationed in a tile accross from an enemy rifleman in an adjacent tile. On the enemy's turn, he moves his rifleman to the next adjacent tile to my cavalry- my cavalry gets a free shot.
TheDarkside is offline  
Old November 3, 2001, 02:59   #3
player1
Emperor
 
player1's Avatar
 
Local Time: 17:58
Local Date: October 31, 2010
Join Date: Sep 2001
Location: Belgrade, Serbia
Posts: 3,218
That 8 value for City Walls is its defense against Catapults & Cannons.
If it is bigger there is less chance to destroy them, if is is lower then chance to destroy them is bigger.

Catapults bombards City Walls: 4/(4+8)=0.33% to hit walls
player1 is offline  
Old November 3, 2001, 03:53   #4
Valant
Settler
 
Local Time: 15:58
Local Date: October 31, 2010
Join Date: Nov 2001
Posts: 13
Ok got some questions before I update....

To TheDarkSide
Could you describe free shot a little more? From what I take it is that any units the move next to the fortress are automatically attacked by the unit(s) in the fortress. But here is my questions
1) Is the any limit to the number of free shots per turn. For example I move 3 units next to a unit in a fortress will this unit free shot against each of the 3 units yet still retain their turn?
2) Does the unit have to be fortified in the fortress for free shot to happen or could the just be inside the fortress for free shot to work?
3) Is free shot automatic?
4) Does the unit doing the free shot ever get damaged?
5) How does the offending unit get damage by free shot? Is it like a 1hp drop like bombardment or is it a full on battle to the death/retreat?

To Player1
1) I just recently discovered that bombardment units have a rate of fire value, do you know what this value does?
2) Do you know what happens for towns over a pop of 6 where city walls seem to disappear?
3) Can bombardment affect more than 1 thing? I.E. hitting wall yet doing 1 hp of damage on a unit within the city?
4) If the wall is hit, does it get destroyed?
5) Does that value of 8 only pertain to bombardment calculations?
6) Is the +50% def bonus or does that value of 8 get applied to a defending unit being attacked by bombardment within a city?


Thanks for the responce.
Valant is offline  
Old November 3, 2001, 05:07   #5
Eli
Civ4 SP Democracy GamePtWDG Vox ControliC4DG VoxCiv4 InterSite DG: Apolyton Team
Emperor
 
Eli's Avatar
 
Local Time: 18:58
Local Date: October 31, 2010
Join Date: Jul 2000
Location: Israel
Posts: 6,480
Re: Combat System Explained
Quote:
Originally posted by Valant
Terrain Bonuses:
When combat occurs over a river, the defender receives a bonus. How much this bonus is, is UNKNOWN.
In the Civilopedia it is stated that if both units stand on the opposite sides of the river the defender gets a 25% bonus.
__________________
"Beware of he who would deny you access to information, for in his heart he dreams himself your master" - Commissioner Pravin Lal.
Eli is offline  
Old November 3, 2001, 05:49   #6
player1
Emperor
 
player1's Avatar
 
Local Time: 17:58
Local Date: October 31, 2010
Join Date: Sep 2001
Location: Belgrade, Serbia
Posts: 3,218
Quote:
Originally posted by Valant
Ok got some questions before I update....

To Player1
1) I just recently discovered that bombardment units have a rate of fire value, do you know what this value does?
2) Do you know what happens for towns over a pop of 6 where city walls seem to disappear?
3) Can bombardment affect more than 1 thing? I.E. hitting wall yet doing 1 hp of damage on a unit within the city?
4) If the wall is hit, does it get destroyed?
5) Does that value of 8 only pertain to bombardment calculations?
6) Is the +50% def bonus or does that value of 8 get applied to a defending unit being attacked by bombardment within a city?
1) I think it is how much damage can bombardment do in a single turn,
for instance Artillery has rate of fire 2. When you say bombard it fires to shots, if both are succesfull then enemy unit loses 2 hit points, if only one is succesful enemy loses 1HP, etc... In fact it is the MOST IMPORTATNT ability of bombardment unit (more important then bombard ratink if you ask me)
2) They will have bouns for being a CITIES (7-12 pop) of 50+, I don't know do Great Wall give anything to them.
3) Maybe, I don't know (if unit has fire rate of 1, it will affect only one thing, if not then I don't know)
4) I don't know I haven't played game much, if walls have some sort of HP rating then, they get hit, if hot then they are probably destroyed
5) I thnik so, otherwise they would be to powerfull
6) they don't get +8 defese (you play too much CTP ), I think thay get only +50%, 8 is "probably" defense of walls itself
player1 is offline  
Old November 3, 2001, 07:51   #7
Grim Legacy
Prince
 
Local Time: 16:58
Local Date: October 31, 2010
Join Date: Aug 2001
Posts: 624
Difficulty level does not affect combat, Soren Johnson said. Which is surprising but welcome in a way.
Grim Legacy is offline  
Old November 3, 2001, 15:03   #8
TheDarkside
Civilization IV Creators
Prince
 
TheDarkside's Avatar
 
Local Time: 11:58
Local Date: October 31, 2010
Join Date: Oct 2001
Location: New York
Posts: 586
Quote:
Originally posted by Valant
Ok got some questions before I update....

To TheDarkSide
Could you describe free shot a little more? From what I take it is that any units the move next to the fortress are automatically attacked by the unit(s) in the fortress. But here is my questions
1) Is the any limit to the number of free shots per turn. For example I move 3 units next to a unit in a fortress will this unit free shot against each of the 3 units yet still retain their turn?
2) Does the unit have to be fortified in the fortress for free shot to happen or could the just be inside the fortress for free shot to work?
3) Is free shot automatic?
4) Does the unit doing the free shot ever get damaged?
5) How does the offending unit get damage by free shot? Is it like a 1hp drop like bombardment or is it a full on battle to the death/retreat?
Thanks for the responce.
1) From what I've seen, yes, only one free shot then they can mvoe as many units passed as they want without getting shot at.

2) Preliminarily speaking- Yes, from my play experience the unit has to be fortified to be able to take free shots.

3) Yes, if you blink you might not even notice! Theres no prompt or confirmation. (on a side note: i tried looking in script.txt for an entry of "FREESHOT" or whatever so maybe I can try making a dialogue to notify the player that a unit is taking a free shot, but there is no such entry )

4) No.

5) It gets hit once, and damage depends upon unit doing free shot. I've seen artillery take free shots- so they took quite a bit of health as you expect.
TheDarkside is offline  
Old November 3, 2001, 20:02   #9
heardie
Prince
 
heardie's Avatar
 
Local Time: 02:58
Local Date: November 1, 2010
Join Date: Aug 1999
Posts: 684
Anyone know a way to make this into one equation. I could get my maths book, but its a bit far away. Where's Victor Galais? He seems to be able to answer all my questions.

Or else I could use a loop in Visual Basic.
heardie is offline  
Old November 4, 2001, 02:04   #10
billybaruch1
Chieftain
 
Local Time: 10:58
Local Date: October 31, 2010
Join Date: Nov 2001
Posts: 48
First time poster...

Regarding the probability of an elite unit defeating a regular unit, the originator of this thread was definitely on the right track. I only have two corrections to make. One is that the "morale" of the units (to use SMAC terminology) or the Hit Points, does not determine what was called the "crude turn win %." Instead, that percentage is determined just as the author stated, that is

Crude Turn Win % = (attackers offensive rating)/ [(attackers offensive rating)+(defenders defensive rating)]

The second point I'd like to make is that to determine the probability of winning over an opponent with more than one hit point, the last "roll" must be won. Let me explain.

Let's say there are no defensive adjustments and it is an 5 HP attacker with an offensive rating of 4 versus a 3 HP defender with a defensive rating of 2. The attacker's Crude Turn Win % = 4 / (4 + 2) = 2/3. Okay, to get the probability of the attacker winning the battle, we do have to calculate how likely it is the attacker will win in 3 rolls, 4 rolls, etc. up to 7 rolls. (It cannot go 8 rolls because then both attacker and defender would be dead.) But in each calculation for figuring the attacker's winning [Pr(A win)], the attacker must win the final roll. That's it. Very simple.

So we have
Pr(A win in 3) + Pr(A win in 4) + Pr(A win in 5) + Pr(A win in 6) + Pr(A win in 7) = [(2/3)^3] + [3 x (2/3)^3 x (1/3)^1] + [6 x (2/3)^3 x (1/3)^2] + [10 x (2/3)^3 x (1/3)^3] + [15 x (2/3)^3 x (1/3)^4] = 95.5%.

As a check, we can figure the chance of the defender winning (it'll take at least 5 rolls):
Pr(D win in 5) + Pr(D win in 6) + Pr(D win in 7) = [(1/3)^5] + [5 x (1/3)^5 x (2/3)^1] + [15 x (1/3)^5 x (2/3)^2] = 4.5%

After all that, my question is: in Civ3 how can we get a message to popup advising us of the odds and prompting us to decide to continue with the attack or not? SMAC has that and I use it all the time.
billybaruch1 is offline  
Old November 4, 2001, 03:24   #11
heardie
Prince
 
heardie's Avatar
 
Local Time: 02:58
Local Date: November 1, 2010
Join Date: Aug 1999
Posts: 684
I would love it if you could make that a bit more generic.

Say
Attacker 1 Hitpoints : X
Defender 1 Hitpoints: Y
Attacker 1 Offense : A
Defender 1 Defenders Rating: B

I'll try myself

Crude Turn Win % = (attackers offensive rating)/ [(attackers offensive rating)+(defenders defensive rating)]
Giving

Crude Turn Win %(C) = A/(A+B)

Victory Probability
Pr(Win in Min) + Pr(Win in Min +1) ...Pr(Win in Max)
[What are the values of Min and Max here?]

= C^Win in Min + ?

Could you help?

I want to get this into a program. If you could tell me the formula's it would be much appreciated.

BTW: Are you using the binomail theorem or the hypergeometric random distrubition varibale or something else here?
heardie is offline  
Old November 4, 2001, 04:37   #12
billybaruch1
Chieftain
 
Local Time: 10:58
Local Date: October 31, 2010
Join Date: Nov 2001
Posts: 48
Yeah, I couldn't remember that name, "hypergeometric." I can generalize it for the 4 basic variables of

Attacker 1 Hitpoints : X
Defender 1 Hitpoints: Y
Attacker 1 Offense : A
Defender 1 Defenders Rating: B.

But Civ3 has already done this! I want them to give me the odds, not have to calculate them myself. I'm here to have fun. (Well, arithmetic, especially combinations and permutations is fun, but anyway...) Valant's point, and it was a good one, was that so many of the adjustments to the basic variables are UNKNOWN.

Let's try this. Please check me to see if I'm wrong. We agree on the attacker's "Crude Turn Win % = A / (A+B). What Civ3 does is apply that percentage until someone loses, okay? In my understanding, their program does not, for example, determine in advance that the attacker has a 95.5% probability of winning. I think their program simply "rolls the dice" with a 66.67% chance (from the previous example) of the attacker taking away a defender hit point.

To get you started: for the attacker the MIN and MAX are MIN = Y and MAX = X - 1 + Y. And for the defender it is MIN = X and MAX = Y - 1 + X. The reasons for these are, at a minimum, the winner has to beat the loser the number of times in a row equal to the loser's hit points, and, at the most, all the hit points except one will be gone at the end of the battle.

But to write a general formula and explain the combinatorial math is, as they say, beyond the scope of this message. If you can tell me, based on an assumption of the Yankees having a 2/3 probability of winning any given World Series game, and the Diamondbacks having a 1/3 probability, what's the overall probability of the NY team winning the World Series, then you should be able to do a similar calculation for Civ3 Combat.
billybaruch1 is offline  
Old November 4, 2001, 07:20   #13
heardie
Prince
 
heardie's Avatar
 
Local Time: 02:58
Local Date: November 1, 2010
Join Date: Aug 1999
Posts: 684
Okay. I'll still try and make a program.

Yankee question: I hate baseball

Last edited by heardie; November 4, 2001 at 07:32.
heardie is offline  
Old November 4, 2001, 15:01   #14
jones
Chieftain
 
jones's Avatar
 
Local Time: 15:58
Local Date: October 31, 2010
Join Date: Jun 2000
Location: USA
Posts: 81
Quote:
Originally posted by heardie

Pr(Win in Min) + Pr(Win in Min +1) ...Pr(Win in Max)
[What are the values of Min and Max here?]

= C^Win in Min + ?
principle of choice (combinations)
Pr(
(2/3)^3 - base chance of a winning 3
(
3c1 (3)- to go 4 rounds, d must win 1
4c2 (6)- " 5 " 2
)
(1/3)^n++ - pr d will win 1 more round each time

Bleh..i'm terrible at explaining math. However,
Attached Files:
File Type: zip combat.pl.zip (706 Bytes, 123 views)

Last edited by jones; November 4, 2001 at 16:11.
jones is offline  
Old November 6, 2001, 14:14   #15
CubsFan915
Settler
 
Local Time: 10:58
Local Date: October 31, 2010
Join Date: Aug 2001
Posts: 12
Re: Combat System Explained
Quote:
Originally posted by Valant
The penalty (if it exists) for naval units caught in port is UNKNOWN
I think that the penalty is that the unit gets sunk - one of my first games I lost a caravel when it was the only unit left in my city and the city got captured...
CubsFan915 is offline  
Old November 6, 2001, 14:37   #16
FeathersMcGraw
Settler
 
Local Time: 15:58
Local Date: October 31, 2010
Join Date: Nov 2001
Posts: 13
Re: Re: Combat System Explained
Quote:
Originally posted by CubsFan915

I think that the penalty is that the unit gets sunk - one of my first games I lost a caravel when it was the only unit left in my city and the city got captured...
I've noted this behavior with galleys and caravels -- I'm not sure if it also extends to "offensive" naval units such as frigates, ironclads, etc. (I would suspect that it would, so you might want to evacuate those battleships if your port city's about to be taken!).

Capturing a city with workers housed inside results in the capture of those workers, as well.

I seem to recall reading that artillery units can be captured only if your civilization has the technology which enables its construction (e.g. mathematics for catapults, etc.), but I don't recall if I've ever captured one along with a city I've conquered militarily.
FeathersMcGraw is offline  
Old November 6, 2001, 16:31   #17
Yohan
Chieftain
 
Local Time: 15:58
Local Date: October 31, 2010
Join Date: Apr 1999
Location: Jackson, MS
Posts: 64
I've discovered something odd about the combat model.

I tried this a few times and it seems to be consistent. I tried attacking a regular spearman with a veteran swordsman and lost. I reloaded the game, tried the same attack, and lost again. What was strange was the sequence of hit point losses was EXACTLY IDENTICAL. I thought this was a strange coincidence, so I tried it a few more times, and the hit point loss sequence was identical every time. Curious, I tried attacking with a different veteran swordsman instead, and the results were once again identical.

But then I tried attacking with an elite swordsman and I won.

I have this strange suspicion that prior to each turn a table of combat results is precalculated. So depending on the attack/defense ratings of the units involved in combat and the number of hitpoints, the game looks up the results in a precalculated table to determine who wins. I have no other plausible explanation for this, and it strikes me as a very strange way to resolve combat.
Yohan is offline  
Old November 6, 2001, 16:39   #18
Yohan
Chieftain
 
Local Time: 15:58
Local Date: October 31, 2010
Join Date: Apr 1999
Location: Jackson, MS
Posts: 64
double-post

Last edited by Yohan; November 6, 2001 at 16:47.
Yohan is offline  
Old November 6, 2001, 17:59   #19
rah
lifer
Apolytoners Hall of FameCivilization IV: Multiplayer
Just another peon
 
rah's Avatar
 
Local Time: 10:58
Local Date: October 31, 2010
Join Date: Nov 1999
Location: who killed Poly
Posts: 22,919
Quote:
Originally posted by Yohan
I've discovered something odd about the combat model.

I tried this a few times and it seems to be consistent. I tried attacking a regular spearman with a veteran swordsman and lost. I reloaded the game, tried the same attack, and lost again. What was strange was the sequence of hit point losses was EXACTLY IDENTICAL. I thought this was a strange coincidence, so I tried it a few more times, and the hit point loss sequence was identical every time. Curious, I tried attacking with a different veteran swordsman instead, and the results were once again identical.

But then I tried attacking with an elite swordsman and I won.

I have this strange suspicion that prior to each turn a table of combat results is precalculated. So depending on the attack/defense ratings of the units involved in combat and the number of hitpoints, the game looks up the results in a precalculated table to determine who wins. I have no other plausible explanation for this, and it strikes me as a very strange way to resolve combat.

What you're seeing is a common trick by game designers these days. All Combats are decided by a random number generator. The generator uses a seed number to start the sequence and the resulting number becomes the seed number for the next random number. So if you always start with the same seed number, the results will always be the same. The game stores the seed number so if you reload the same thing will occur. To change this you must do one of two things.
1. stop the game and restart it (which will clear the seed)
2. do something else that requires a random number to generate, which will change the seed number. (I believe diplomacy uses random numbers some times) or do another attack first.

Why you won with the elite unit is probably because it had another hit point to survive until it could win the later rolls.

RAH
__________________
The OT at APOLYTON is like watching the Special Olympics. Certain people try so hard to debate despite their handicaps.
rah is offline  
Old November 6, 2001, 22:02   #20
billybaruch1
Chieftain
 
Local Time: 10:58
Local Date: October 31, 2010
Join Date: Nov 2001
Posts: 48
Okay, here's the general formula for the simplest combat situation. Thanks go to Valant and heardie - hey, maybe we can publish this and make some money?

First, let me correct what I wrote before about how SMAC units' morale corresponds to Civ3 Hit Points. That's not right. I think it's the SMAC "power" that corresponds to Civ3 Hit Points. If you know about the fission, fusion, quantum, singularity progression in SMAC, well, I think that to defeat a singularity unit with 40 power, it has to be "killed" 4 times as much as a fission unit with 10 power.

Next, let me say I got Civ3, but after spending a good part of last weekend reading the manual and then reading these boards, I returned the disk unopened. I may buy it again after some bugs are resolved and a patch comes out because I don't want to develop incorrect playing habits. This may seem strange, but the Air Superiority is one of the best things in SMAC. If it doesn't happen in Civ3, I'm gonna wait.

To heardie: I wasn't referring to baseball, per se. I was trying to say that Civ3 combat with both attacker and defender in veteran status with 4 Hit Points is exactly like the best 4 out of 7 baseball World Series. Or at least it should be.

Let's define the following:
A = Attacker's Offensive Rating
B = Defender's Defensive Rating
X = Attacker's Hit Points
Y = Defender's Hit Points

Also, p = A / (A + B) and q = 1 - p = B / (A + B).

And 0! = 1, 1! = 1, 2! = 2 x 1 = 2, 3! = 3 x 2 x 1 = 6, etc.

Since the maximum Hit Points is 5, we need to sum 5 calculations:

For n = 0, 1, 2, 3, 4:

If X > n,
Pr(A win in Y+n) = {(Y + n - 1)! / [(Y - 1)! x n!]} x p^Y x q^n.
If X <= n, 0.

That's it, my friends. All adjustments for defensive, terrain-related bonuses, and bombardments and much, much more need to be applied to the p's and q's before the above can approach becoming a comprehensive formula.

Check me on this: if A, B, X, Y = 4, 2, 5, 3, you should get 95.5%.
billybaruch1 is offline  
Old November 7, 2001, 13:48   #21
xmax
Settler
 
Local Time: 15:58
Local Date: October 31, 2010
Join Date: Nov 2001
Posts: 3
Here's an Excel table with the same calculations - thanks to billy for spelling it out.

I suspect all of this is theoretical as the game applies a lot of modifiers to ADM numbers. For example I noticed that attacking after bombardment, even if defending unit sustained no damage and population remained the same, has much better chance of success than attacking without one.

Then of course there are bizarre cases, at least when you don't know exact odds, when my 2 elite cavalry units attack a knight with 1HP stationed across the river and BOTH get killed. Must have been headless horseman or smth
Attached Files:
File Type: zip civ3.zip (3.2 KB, 202 views)

Last edited by xmax; November 7, 2001 at 14:12.
xmax is offline  
Old November 9, 2001, 13:38   #22
gfrazieror
Chieftain
 
Local Time: 07:58
Local Date: October 31, 2010
Join Date: Nov 2001
Location: Eugene OR
Posts: 31
My experience thusfar indicates that removing the firepower factor from combat was a mistake; ridiculous battles like those in the original Civ, where a defending phalanx defeats armor, are showing up in Civ III. I lost an elite ironclad that was attacking a veteran caravel that already had damage due to shore bombardment.

This is just not right...technologically advanced units should not be losing battles like this. I can tolerate something like a rush of impi on a sole defending rifleman and having the primitives win, but this is ridiculous. I want firepower back!
__________________
Gary Frazier
Civ Freak from way back

Last edited by gfrazieror; November 9, 2001 at 14:06.
gfrazieror is offline  
Old November 9, 2001, 15:08   #23
TheDarkside
Civilization IV Creators
Prince
 
TheDarkside's Avatar
 
Local Time: 11:58
Local Date: October 31, 2010
Join Date: Oct 2001
Location: New York
Posts: 586
Firepower

Battleship getting sunk by frigate

Armor ratings

Tank getting destroyed by army of knights
TheDarkside is offline  
Old November 9, 2001, 17:28   #24
Karhgath
Chieftain
 
Karhgath's Avatar
 
Local Time: 10:58
Local Date: October 31, 2010
Join Date: Dec 1999
Posts: 82
Yeah, sure, and make it so technology rules all like in Civ2, if you have the edge, you win any war without really any chance of losing a unit or too with some massive rush.

Now in Civ3, you MUST use strategy and tactics if you want to win. Just stop complaining that, one in like 20 times, you get beaten by a lower tech unit. Just use strategy instead of just attacking, one unit after another. Flatten them with artillery first. Don't just send one tank to conquer a city, or one ship to destroy another. Anything can happen, I don't think a General could say to his superior that :

"I sent one squad of marines to take over Paris, and those stupid musketeer cornered us and use the city has a battleground, they knew where to hide. In my calculation, they should just have surrendered because, you know, our marines are better than them!"

He would be kill before finishing his explanations =)

Also, I THINK I read somewhere that having an elite unit in the stack gives you defensive bonus, not 100% sure tho.
__________________
-Karhgath
Karhgath is offline  
Old November 9, 2001, 18:08   #25
gfrazieror
Chieftain
 
Local Time: 07:58
Local Date: October 31, 2010
Join Date: Nov 2001
Location: Eugene OR
Posts: 31
You can't use tactics in ANY Civ; there are no bonuses in combat for enveloping cities or for flanking manuvers at sea. If a band of warriors in rags can defeat soldiers in AFVs and this happens in abstracted strategic combat, something is seriously wrong with the abstracted combat, and that is what you have in Civ games.

I don't disagree at all with you point about using bombardment units to soften up the opposition; that part is quite realistic. But a bunch of guys with spears and shields holding up against 120mm HE fired from armored vehicles? I think not.

Why develop tanks when your horsemen have nearly the same chance of winning a battle? The whole idea is to use technological superiority to subdue your backwards opponents...not only militarily, but culturally as well in Civ III.
__________________
Gary Frazier
Civ Freak from way back
gfrazieror is offline  
Old November 9, 2001, 18:25   #26
Karhgath
Chieftain
 
Karhgath's Avatar
 
Local Time: 10:58
Local Date: October 31, 2010
Join Date: Dec 1999
Posts: 82
Why would they have the same chance? Clearly, the tanks have about 95% chance of winning over a phalanx, and probably much more, so why are you saying it makes no difference? I dunno, in history, in happened a couple of time(that we know off) where less tech won over bigger tech. Frankly, tanks are very useless against infantry, they can easely get near the tank and sabotage it and all. Ok, there are no such thing you can do in civ, but the 5% chance might be because something like that happened. Or, what if they ambushed you in their city, they have the terrain advantage. In wars like that, not in open terrain(where only numbers matters), one single man can do a lot if he knows what to do. Wars in cities, especially in WWII, could take months, and have lots and lots of unexpected events. Ok, they had mostly similar tech, but, one time out of a hundred, some unexpected events can happen. Your marines where dinning in the park when impis attack them, overwhelming them, hehe.

You know, how much time have you defeated less tech units? You know, humans have the bad habits of always remembering bad event more than happy ones =) I could count on my fingers the number of time I lost to a unit several tech step down, and I've played about 5-6 games yet, half conquer games. Most of these times even, I lost to an Elite unit, and mine where regular or conscript.

Also, would it be fun if you where ertain to win with a better tech unit(civ2 anyone?)? I know in Civ2, and even more so in SMAC, if you had the tech advantage, you were sure ot win. Now it was just tedious to declare war and build an army to conquer them, so I usually never finished those games. Now war can be unexpected(as it should be).

Also, just another quick one. Why aren't the US sending all his troops to Afghanistan? They clearly have the tech advantage. They just bomb them for months before, and send just elite squads. Why? Well, because they have the terrain advantage, which is a BIG one. Sending their troops there would mean casualties, even for better equiped and trained US soldiers.

War is not an easy calculation of numbers. It can be unexpected and dangerous. If it was that easy, afghanistan would probably be filled with US soldiers by now.

And, if you want to comment about Desert Storm, and how they won easely, well, it was in a desertic area, the defenders never had much terrain advantage. Over clear terrain, numbers/tech win hands down. But usually, it's not as easy as that(Vietnam, Guerilla Warfare, etc.)
__________________
-Karhgath
Karhgath is offline  
Old November 9, 2001, 18:57   #27
gfrazieror
Chieftain
 
Local Time: 07:58
Local Date: October 31, 2010
Join Date: Nov 2001
Location: Eugene OR
Posts: 31
The Taliban troops are not all that "low tech" compared to any US Troops; they are armed with automatic weapons, SAMs, and ATGMs. Not spears and shields. It is much more evenly matched than the Spanish vs. the Aztecs.

In Civ you're dealing with massive disparities in technologies. While Pikemen and Musketmen are not that far apart, remember that early firearms were a far cry from the sort of weapons wielded even in Napoleanic times in terms of accuracy and rate of fire. Which is why I chose the impi example specifically; superior numbers can carry the day, even against a technologically superior foe.

But a lone warrior overcoming trained Infantry with standoff weapons like firearms? Wooden ships firing cannonballs against armored steamships with rifled cannon?
__________________
Gary Frazier
Civ Freak from way back
gfrazieror is offline  
Old November 9, 2001, 22:00   #28
Azrael-42
Settler
 
Local Time: 15:58
Local Date: October 31, 2010
Join Date: Nov 2001
Posts: 3
Arg
Khargath, i agree with a lot of what you said but your comment on tanks is absolutely ridiculous. 100 tanks could fight 10 000 modern soldiers, out in the open, and probably hardly even take a few scratches. Modern tanks have 4 heavy machine-guns mounted on them... Not to be confused with assault-rifles/submachine guns, which is what infantry carry around with them. Machine guns are the ones that need to be set up on bipods (light) or tripods (heavy) before use. They have a range far greater than assault rifles and mow down infantry like so much wheat. They spit out heavy calibre bullets at insane firing speeds. The only way the infantry have a chance is if they are trying to hold a defensive position. That way, they can a) use cover, and b) Set up anti-tank missile systems, notably the Sager russian system and the TOW american equivalant. This is actually a post WWII development. Probably, the absolute peak of usage of tanks in warfare was the Six-Days war, fought by the Israelis in 1967. They used classic, hard hitting, flanking tactics in perfect combination with air support (ironically, similar to German panzer tactics). However, when the Israelis tried to use the same tactics in 1973, the Yom Kippur War, they took heavy tank losses because the Egyptians (which relatively relied more on infantry) were prepared, and using Sager missiles.
Azrael-42 is offline  
Old November 9, 2001, 22:31   #29
Marquis de Sodaq
King
 
Marquis de Sodaq's Avatar
 
Local Time: 09:58
Local Date: October 31, 2010
Join Date: Jul 2000
Location: ... no, a Marquis.
Posts: 2,179
Regarding the statistics...
Altho firepower is gone, please go to the CivII Strategy forum. The thread Info: Combat (GL) (edit: link added) explains how the probability math works in CivII combat. I suspect this can be easily modified to your needs. To figure the odds of an attacker winning, use a SUM of COMBinations. You end up with the cumulative odds of success. The defender wins (1-attacker's odds). Again, what is input to determine each round will differ, but the full battle odds may work with the same math.

Regarding the simplified formula given in the game documentation: Don't believe it. The CivII thread will show you why. The documentation of Civ has always simplified for explanation, it hasn't worked for mathematics! Also, CivII has only one additive combat adjustment, all others are multiplicative. I'd suspect that holds true in Civ3, too, just because of how programming works.

When I get the game I plan to work out the answers you seek in the first post. If anyone has questions about how CivII combat works, see that thread! Or PM me for further details...
__________________
The first President of the first Apolyton Democracy Game (CivII, that is)

The gift of speech is given to many,
intelligence to few.
Marquis de Sodaq is offline  
Old November 10, 2001, 09:17   #30
Growler
Settler
 
Local Time: 10:58
Local Date: October 31, 2010
Join Date: Nov 2001
Posts: 13
Haste penalties
I haven't played any Civ games before but I haved played SMAC.

In SMAC, a unit which moves before attacking and has only a fraction of a movement point left when it attackes is penalized in its attack strength. Ex. an infantry unit (1 mp) moves 2 road squares than attacks a third, its attack strength is 1/3 of normal.
I am calling this a "haste penalty", pardon me if there is already an established jargon for this.

It is my opinion that haste penalties apply in Civ3. Even though combat odds calculations are hidden, it explains some combat results in the open ground using horseman or knights versus strength 1 defenders, when the horsemen had moved.
__________________
eof
Growler is offline  
 

Bookmarks

Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is On

Forum Jump


All times are GMT -4. The time now is 11:58.


Design by Vjacheslav Trushkin, color scheme by ColorizeIt!.
Powered by vBulletin® Version 3.8.2
Copyright ©2000 - 2010, Jelsoft Enterprises Ltd.
Apolyton Civilization Site | Copyright © The Apolyton Team