Combat System Explained

[FeathersMcGraw]

Re: Re: Combat System Explained

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.

[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.

[Yohan]

double-post

[rah]

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

[billybaruch1]

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%.

[xmax]

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

[gfrazieror]

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!

[TheDarkside]

Firepower

Battleship getting sunk by frigate

Armor ratings

Tank getting destroyed by army of knights

[Karhgath]

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.

[gfrazieror]

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.

[Karhgath]

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.)

[gfrazieror]

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?

[Azrael-42]

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.

[Marquis de Sodaq]

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...

[Growler]

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.