Combat System, Take 2

[scavenger]

D Day invasions can be neccisary even on a huge map

I am on a huge map, and early in the game I got boxed in on both sides by two major superpowers, so no room to expand out, but just a short transport hop away was the japenese civ, who were, at the time, at war with 3 other civs and getting their butts kicked. I needed room to grow, couldn't go left, couldn't go right, so joined other nations, filled two transports with infantry (most advanced unit I had at the time) and raided the western coast of Japan and chewed through em.

in my current game im about to make another D day landing, the russians have their own continent, but I want them to die, so right now im in the process of loading transports (6 of them) full of modern armor to rush em...

 

[Sophist]

Good thread; I hope it remains a civil discussion about this issue....

The spearman/tank thing has definitely happened to me, in just about every game I've played. But let's just say that's anecdotal, that it doesn't happen often, and so forth. The real problem is what that does to the game and to the calculus in the game. Simply put, losing too often in combat has definitely dampened my enthusiasm for making of war. If this was the purpose of the combat system, well, it was damn successful.

What I do wonder about, however, is whether anyone has ever tested the combat system. Occurs to me that perhaps a mod can be made to use the underlying combat system engine for a test. If someone on the forum has the skills to run some tests, and post the results, that'd be great.

What I have in mind is something like this:

-Somehow set up a large number (100? 1000? 10000?) of one-on-one fights between different units on different terrain/conditions (i.e., Warrior vs. Spearman on Grasslands, across a river, etc.) and see how the results come out.

-Run some statistical analysis on the results and see if they adhere to the description of combat in the manual. That is, in a fight between a Cavalry (Att 6) and a Spearman (Def 2), the Cavalry should come out ahead 75% of the time (6/8=3/4=75%). Does it in fact come out ahead 75% of the time?

What do you think? Is it possible to set up such a mod?

-Sophist

 

[Akka]

Originally posted by Sgt Zimm
1. Pikemen doubling their defence rating against mounted units (horses that is.) - Civ II

2. Infantry get a bonus when attacking a city (I think it was 25%) - SMAC.

And a +25 % defending, too.

3. Not really a bonus per se, but units from different era's had increased firepower. Gunpowder gave a FP of 2 vs. 1 for units without it. I think the Battleship had a firepower of 4, but I'm not positive. - Civ II

It was the hit points, not the firepower :
- gunpowder units had 2 HP
- armored units had 3 HP
- heavily armored units (BB and Carrier) had 4 HP

4. Is the "Pearl Harbour" combat modifier (from Civ II) still in Civ III? If it isn't, it should be.


Can't think of anything else at the moment. Feel free to add more or trash the ones I've suggested.

Aircraft fighters had attack doubled against helicopters.
Aircraft fighters had defense halved against land attacks when they were in a city.
Ships had defense halved against land attacks when in a city.
Aircraft scrambling to defend the city against air attack (tripling the defense of the defending unit). I don't exactly remember how this one worked, I may be wrong.

 

[Lt. 'Killer' M.]

had "strange" combat results as well. Trippled hitpoints -- GONE!!!! Chances are just too small.

Still i wonder: Elite spearman on mountain, fortified attacked by 2 veteran warriors. first gets him down to red, second kills him off and stays green..... hardly seen the AI take on these odds before, but if it does, my units dead - does the puter KNOW results beforehand and THEn decides if it should take the "chance" ?

 

[Lt. 'Killer' M.]

Sophist: will do if I find time. Will also test Tank/Spearman....

 

[simwiz2]

the tank/spearman thing happens way too often. i was over at my firends house and he was complaining about losing tanks to ancient units alot, and i didnt believe him but i watched him play his game for 2 turns. in those 2 turns he attacked about 5 times, and a regular spearman defending grassland towns/cities forced 2 tanks to retreat. then he attacked again and it was a close battle, but the tank won with 1 hp. theres 3 results, in a row, that should NEVER happen in civ3

 

[Zachriel]

Tanks and urban warfare!

People defending their homes can be very dangerous. If it was your home, what would you do?

It is best to have plenty of firepower when attacking cities. Be sure to bombard first to destory their morale and ability to fight. If tanks were enough, Germany would own the world by now. No one has ever conquered the entire world, unless you count Genghis Khan, but he was a "primitive."

 

[simwiz2]

if it was my home i would take a cast iron pan and start beating on the tank with it... uh oh the tank didnt blow up yet and i think the people inside may be getting mad...

ya know, i think shells could hurt in this situation

 

[Gruntboy]

I don't know about any bugs but here is some maths that gives you an idea of what attackers are up against.

There are 2 examples of siege combat for each era, each involving regular units. The first (a) is a fortified town (or city) on a plain with a garrisoned unit. The second (b) is a metropolis on hills behind a river with a garrisoned unit. The latter may be anachronistic for ancient or medieval era combat but it could be possible in games where the AI does not upgrade troops, the player is technologically advanced or just representative of a difficult fight – a mountain top fortress perhaps.

The calculations for each round of combat in the examples are as follows:

(a) Probability of attacker inflicting 1 hit point of damage =
A / (A + D + D *(0.5 + 0.1 + 0.25))

Probability of defender inflicting 1 HP damage = 1-(a)


(b) Probability of attacker inflicting 1 HP damage =
A / (A + D + D*(1 + 0.5 + 0.25 + 0.25))

Probability of defender inflicting 1 HP damage = 1-(b)


Era /Units /A /D
Ancient Swordsman Vs Spearman 3 2
Medieval Longbowman Vs Pikeman 4 3
Industrial Cavalry Vs Rifleman 6 4
Modern Tank Vs Infantry 16 10


The probability of a particular outcome is thus calculated as ((product of all Hit Probabilities) * (X)), where X = number of different outcomes.

Example a Attacker wins
Damage /0 /1 /2 /Any Win
Ancient 9.0% 14.9% 16.4% 40.3%
Medieval 7.3% 12.8% 14.9% 35.0%
Industrial 9.0% 14.9% 16.4% 40.3%
Modern 10.0% 16.0% 17.2% 43.2%


Example b attacker wins
Damage /0 /1 /2 /Any Win
Ancient 3.7% 7.4% 9.9% 21.0%
Medieval 2.9% 6.1% 8.4% 17.3%
Industrial 3.7% 7.4% 9.9% 21.0%
Modern 4.2% 8.2% 10.7% 23.2%


Even with a huge tech advantage, there is the possibility of defeat (using example 1, a spearman will destroy an attacking tank unit 4.9% of the time).

 

[DarkwingGT]

Nice post Gruntboy. A little complicated because of formatting issues, but nice.

As for some replies a little while back, I'm positive it was a swordsman who had just moved onto the grassland with one health and who didn't level up. Which surprised me since the computer seems to level up every single battle compared to my units which can defend against 10 horsemen and not level up once because the f***in buggers retreat before getting killed. Seems like I should get at least one promotion, no?

Here's a thought, make unit promotions based on relative strength and damage dealt. If my tank kills a warrior, no big deal, relative strength is too high. But if my warrior kills a tank, he should level up every battle since he does damage to a much superior unit. Not that a warrior can kill a tank, but you get the idea. This way a defending spearman who beats back a bunch of horseman would get at least one promotion because of the sheer amount of damage caused.

 

[Zachriel]

Probably not very effective, simwiz2.

May I suggest attacking their fuel depot, tearing up rail lines, slitting their throats in the night, attacking their command and control, building hidden pits in roads, burning down the town rather than surrendering it, befriending their enemies, taking advantage of the arrogance, etc.

Make them pay in blood (hp). You will probably lose, but maybe if you make the price high enough, they will withdraw.

 

[Moulton]

It begins to look to me like there is a bias for ancient units against modern. Flukes aside.
When I first got Modern armor, I immediately upgraded my tanks, and lost these units -- all -- to knights. I had a few tanks left, and they did better, but did not always win.. didnt expect always, but they took a lot more damage than I would hav thought. So I called in my cavalry that had been standing garrison duty.... they rocked. IT seems like the more distance in time, the better chance the older unit has.
Now I am having battleships fall to ironclads and frigates. Frigates with an attack of 2? It's like the ancient units get enhanced attack bonuses for each era they are bypassed by...
On the other hand, I can turn out battle ships in two turns, and it took me 5 or six to make Men-of-war when they came in...

 

[lonestar]

To avoid these silly combat results it would be cool if say a spearman being attacked by a tank, fleed the battle by randomly moving (strickened with terror) one square away! think about, you have a stick with a bit of kind of sharp metal on the end and a little wooden shield, are you going to face down a 10 ton steel tank? and it would be fun to see the enemy running from you!

 

[Gruntboy]

Darkwing - thanks. I know, they were word tables, I'm writing a strategy document for waging war in Civ3.

Moulton - I think it is not a bias towards ancient but a bias towards the defensive.

Odds for attacking even a small city on a plain are 2.5:1 in any era. 4-5:1 odds for tech gaps or rough terrain/big cities.

If you read military doctrine, odds of 3:1 or more are usually accepted as the minimum for offensive operations. But I'm not going to drag the real world into this.

PS can anyone fault my maths? Las time I did this cr*p was 8 years ago. But it still doesn't explain some combat situations. Odds of 3 tanks losing Vs spearman (scenario a above) is like (5/100)*(5/100)*(5/100) = 0.0125%?!

 

[Gruntboy]

Yikes! Posted them as html tables. Looked fine in IE but they were strangely elongated in the thread? I'l leave it.

 

[Zachriel]

Great analysis, Gruntboy.

I just don't see the problem you guys see. Could it be a technical problem? I believe that a lot of you are having troubles, and take your concerns to heart, but I believe that it is mostly due to the necessary learning curve.

Everytime I attack, sure I lose some, but I always bombard first, have infantry support and plenty of backup. There is so much lead flying when I attack that no spearman will survive long. And yes, I always upgrade my units as soon as practical.

Armor without infantry/artillery support is useless even in the real world -- especially against a fortified position such as a city. Bombard, bombard, bombard, turn after turn if need be. Then when you enter the city virtually unopposed, the inhabitants will stand in awe of your power. (You can let them think it was easy, if you like.)

Anyway, it works for me.

Many a battle has been lost because someone takes their "superiority" too seriously, and their enemy not so seriously.

 

[Nikua]

I think there just needs to be another factor involvled in more modern units. Because, the battle system seems to take Hit Points too seriously. After a round of battling, my tanks have like 1 - 2 bars left. Well, they come in with their Elite Impi's and horsemen, and WIPE out my armies.

I got SO sick of this, I tried to look for a cure via the CivEdit. But, all it allows is more Hit points, attack and defense and ?rate of fire?. Who CARES about attack and defense when my 24 is losing out to a 4?! LOL! Hit points is a cheap fix, and I don't know wth rate of fire does.

They need to add a definite modifier for each age. There should be a LOW *like 5% not the current 50%* chance of something lower killing a higher unit, NOT considering other factors. I still play and enjoy the game, but for ME, this is highly annoying and anti-fun. If I get to the point if I constantly run at a higher tech adnavtage, I will go up in difficulty. But IF I AM higher, I EXPECT to have an ADVANTAGE.

They make a game to be anti-cheating. That's fine a dandy. I personally don't reload games if something didn't go my way. But IF someone wanted too, let them. Now, they keep that cheater from cheating, while screwing MY game. Not cool, and makes no sense. Let the customer CHOOSE how he wants to play. If he wants to win every game via combat, let him on cheiftan. If he wants a challenge, give them alternate ways to win, plus added difficulty levels.

 

[Gruntboy]

Nikua, sure there is a definite "historical" innacuracy with horsemen taking out tanks. But in the game mechanics an elite horseman (A2) fighting a damaged tank (say 1 HP at D8) on plains (+10% D) has the followning probability of winning

Each successful attack [1 needed] 2/(2+8+8*(0.1))=10.8
Probability of hit = 18.5%

That's almost 1 in 5 chance of hitting. With 5 hitpoints, there is a 64.1% (8.2% + 10.0% + 12.3% + 15.1% + 18.5%) chance of the horseman getting a hit in before he loses.

Maybe you should be defending against counter attacks? Save one tank in a stack with full HP, or better yet, infantry with D10.

I know it sucks seeing horsemen do that but with this combat system what do you expect? You either hate it or live with it.

 

[DarkwingGT]

Gruntboy, I hate to ask this of you, but you seem good at this. Last night I got really frustrated because my Persian friends just up and declared war on me for no reason. I had 4 Pikemen in each city, was more advanced then them (though they did have Knights). Ok, so what are the odds of this:

Town of size 12, with Walls. Fortified Pikeman.

Persians come in, attacks my town with a Knight. In four turns proceeds to kill everyone of my Pikeman. Takes at most, 2 HP of damage (he was Veteran to start with).

Correct me if I'm wrong, but don't Pikeman get a massive bonus against Knights? Why do I bother upgrading to Pikeman if they're no more effective than Spearman? Oh well, it just really frustrated me so I was wondering just what the odds where that would happen.

Thanks.


EDIT: BTW, I'll ignore the fact that my knights were being beaten back by Longbowman, Spearman and once a Warrior. In fact, my crowning acheivement was when a Warrior defeated my Pikeman. That was what caused me to quit that particular game. And yes, my Pikeman was in a city of size 12 fortified with walls.

 

[lonestar]

I'll try to calculate this for you, think it's right.
assuming your pikemen were regulars and city is on grasslands:
pikemen defense is : 3 + 3 *( 0.1 ( grasslands) +0.25(town) + 0.5 (fortified) ) = 5.55
knights attack: 4
probability of knights winning a point is 4 / ( 4 + 5.55 ) = 0.418

probability of doing 12 (4*3) dmg and losing 2 hp is:
0.418 exp ( 12 - 2 ) = 1.661 e-4 = 0.0001661
which is pretty small
:crazyeyes