Combat System Explained

[vmxa1]

Sorry for being so verbose, I have no probelm with the position, just I hate the phantom reasons that are used. If one feels a spearmen should win some battles, fine, just no made up reason like bribes.

[Blue Moose]

Originally posted by vmxa1


I like your style.

thanks! I like yours as well.

[Zachriel]

Originally posted by vmxa1
This stuff is just hokkum, I do not see any of these mentioned in the doc. The are a figment of someones imagination. They are for real life, not Civ.
. . . It is a straight up computation, nothing more. Leave the romantic ideas out of the equation.

The randomizer represents those factors which are beyond the direct control of the player. However, if you treat it as strictly a matter of "computation," then the degree of variation in combat outcomes doesn't really matter.

One roving spearmen, does not require me to send a wave of combined arms to defend.

If the defeat of the spearman is of strategic significance, you certainly should consider it. However, if you don't, then don't blame the combat system when it turns out badly.

Custer had access to a machine gun at Little Big Horn, but decided it would inhibit his mobility. A reasonable, but obviously flawed decision in this instance which was completely out of the control of the strategic planners in Washington.

My whole interest is that it would be less likely to annoy new players to the game. When it first came out I read alot of people on many boards complain about this feature and say they took the game back.

Contrariwise, other posters have complained that the game has already been oversimplified for the mass market in order to make it more appealing to new players. Admittedly, it is a hard balance to obtain, and many improvements will undoubtedly be made in the future. However, I am a "glass half full" person and well-able to enjoy the combat system and Civ3 as it is.

I have not yet begun to fight!
John Paul Jones just before boarding and capturing the British Frigate Serapis.

[Theseus]

I'm OK, conceptually, both ways.

First, clearly, this is a game with mechanistic and probabilistic rules. Fine.

Second, however, I do admit that like many of the screamers (whiners^nth), there are times when the RESULTS irk me... I actually enjoy doing what Zachriel does, making up a little fiction to account for the results.

I mostly imagine human error (meaning, my units screwed up); I was in the Marines, and believe me, until military personnel / units are trained to a very high level of proficiency, there is at least a 20% factor for mistakes.

[Zachriel]

Originally posted by Theseus
I mostly imagine human error (meaning, my units screwed up); I was in the Marines, and believe me, until military personnel / units are trained to a very high level of proficiency, there is at least a 20% factor for mistakes.

Historically speaking, they ARE trained to a very high level of proficiency.

SNAFU

http://www.britannica.com/dictionary?book=Dictionary&va=snafu

Wellington's opinion of his army
Ours (our army) is composed of the scum of the earth - the mere scum of the earth.

Baron Von Steuben, quoted while training the American Revolutionary Army
@#^$ @#$ and you G*&^( sons of &*%$ ^%#$#!!!
(It's a rather long quote, actually.)

[Lucarse]

Come on guys we all know CIV 3 does this all the time you just live with it and build your forces accordingly

I lost 2 BB and a 3 Nuke subs to a handful of ironclads and one single poxy destoyer until i got wise and weakend them wih airpower then destoyed them.

Sure a Battleship should NEVER lose to an Ironclad and the old Ironclad should have no idea a boomer is under them!! but sure enough they actualy sort out my sub attacked and won!.

If you want to win these type of engagements use combined arms soften them up first!. the game in this respect is not fair or balanced!.

[Zachriel]

Originally posted by Lucarse
Sure a Battleship should NEVER lose to an Ironclad and the old Ironclad should have no idea a boomer is under them!! but sure enough they actualy sort out my sub attacked and won!.

If you want to win these type of engagements use combined arms soften them up first!. the game in this respect is not fair or balanced!.

I almost never lose a modern ship to ironclads -- anymore. I used to think I was supposed to just ram them, but my Destroyers kept getting dented. Then I discovered how to use those cannon-thingies. O, the Joy!

Bombard, bombard, bombard.

[vulture]

Originally posted by vmxa1
Sorry for being so verbose, I have no probelm with the position, just I hate the phantom reasons that are used. If one feels a spearmen should win some battles, fine, just no made up reason like bribes.

Each to their own - they're not remotely phantom reasons to me. For me, the idea of spearmen lining up on a plain with tanks rolling towards them is ludicrous (and the tanks would win that one essentially 100% of the time). Any spearman with some kind of survival instinct isn't going to be trying to impale a tank with his spear. He is going to be digging pits, making deadfalls, launching raids on the enemy camp, trying to destroy fuel, ammunition, food and possible the tanks. He is going to try and kill the enemy *before* they get into their tank. In the real world, occasionally overconfidence, carelessness or stupidity on the part of the attackers is going to let this happen. Civ doesn't make any attempt to model this kind of stuff of course, but for me (as for others) that is part of what makes up the combat odds and gives the spearman his 1% chance of winning a vastly unfair fight. The 99% of the time that he loses, he is unable to do this and ends up trying to impale spears with tanks, so much the worse for him. If your view of what civ combat 'represents' is different, then fair enough. Neither is automatically right or wrong.

I was reading a book t'other day (gosh!). The author gave a hypothetical situation of a unit being ordered to move from rally point A to advanced rally point B, 20 km up the road. Not something terribly difficult to acheieve, you'd suspect. He then gave a list of some of the thing that he'd seen go wrong with this in training maneuvers (where things are comparatively stress free due to the low chance of being killed), ranging from getting lost (pretty easy and common) or half the unit never getting the order and so not moving (or going to the wrong place) up to people being hospitalised in accidents. In real life, once the tanks actually get loaded up and engage the spearmen in battle, it is all over, but there is a real chance of the losing it before that. I gather that your view of civ 3 combat is that it 'starts' at the point when they get in their tanks and launch the attack, and the chance of screwing up the logistics or anything else is not part of the game.

[Zachriel]

Originally posted by vulture
I was reading a book t'other day (gosh!). The author gave a hypothetical situation of a unit being ordered to move from rally point A to advanced rally point B, 20 km up the road. Not something terribly difficult to acheieve, you'd suspect.

It's not always as easy as it looks:

During the Black Hawk War, Lincoln was marching his militia company in a formation with twenty men across the front, he came to a fence with a small gate in it. “I could not for the life of me remember the proper word of command for getting my company endwise. Finally, as we came near I shouted: ‘This company is dismissed for two minutes, when it will fall in again on the other side of the gate.’”

http://www.suite101.com/article.cfm/presidents_and_first_ladies/94636

[Gen.Dragolen]

Could Someone Explain This to Me?

Originally posted by Hurry
Only the terrain of the defending unit is counted during combat, so it doesn´t matter where your units are.

And PLEASE, if you (Gen.Dragolen for example) want us to take your arguments seriously, POST A SAVEGAME!!! You have it in the autosave folder.

Hurry,

As you wish. The 360 AD save in the zip is after the army was destroyed the first time. The 350 AD save is from the Autosave folder.

According to the combat calculators out there, there should be about an 88% chance for a veteran horseman to destroy an archer in open gound.

I did 9 runs to see what would happen, since that first match up was a shocker.
Random Seed is NOT preserved selected in the game options
Aztec: Army, Horsemen 2/1/2, 11/12 hp
Roman: Archers 2/1/1 3 hp
Results:
7 wins for the horseman army and 2 losses:
1. Army -11 hp attacking 1st archer, archer takes no damage
2. Army -2 hp attacking 1st archer, -5 hp attacking second archer
3. Army -5 hp attacking 1st archer, -6 hp attacking second archer
4. Army -6 hp attacking 1st archer, -2 hp attacking second archer
5. Army -3 hp attacking 1st archer, -2 hp attacking second archer
6. Army -0 hp attacking 1st archer, -1 hp attacking second archer
7. Army -5 hp attacking 1st archer, -0 hp attacking second archer
8. Army -2 hp attacking 1st archer, -0 hp attacking second archer
9. Army -3 hp attacking 1st archer, -3 hp attacking second archer

I did exctly the same moves with the workers that moved before the army each time. I did the same attack each time as I had intended in the first attack, and that was to destroy both archers and then lead the nearby legions on a wild goose chase.

My stats textbook is about 20 years out of date, and the subject was never my strong suit, but the probability of the first outcome and the 3rd must have been extremely small.

Having seen results like this in almost every game I've played, especially when I have had to take it to the enemy early, you can see where I doubt the way the game does combat calculations. I know the AI has a threshold overwhich it will not attack units, because the odd of winning are too low.

Vulture has gone and checked the math for the Psuedo Random Number Generator the game is using, and I am satisfied that it is not at fault. My paranoia suggests that there are other factors used to see who loses the hp each round of combat. In any random environment, you don't expect to see extreme results with such frequency.

As nice as it is to see things like a pack of 6 Ironclad tear down a battleship and carrier using only 5 attacks in a desperate charge, the effect of the game is to insult one's intelligence by reducing rational thought and planning that should decide who wins the day down to a mere coin toss. Ironclads wouldn't even make a dent in a WW2 vintage battle ship or cruiser. Even at night, their cannons would not have been able to dent the armour plating, while the close action deck guns (3"-5") would have shredded any ironclad design from the Monitor to the HMS Warrior. HE solves alot of gunnery problems.

I played wargames all through high school that used a simple table with the difference between the attacker's offensive strength and the defender's defensive strength, and a 6 sided dice. The attack strength was a measure of firepower and morale and for the defense, armour, morale and terrain. At least in those games you could plan and concentrate your forces and get a reasonable outcome to the battles. And it worked both ways. If your strategy was unsound, you lost. But not in as spectacular a fashion as in this game.

Considering the combat system is based on a "division" scale unit (roughly 10000 men per division) they should have gone for a simplified, abstract system, instead of the current Monte Carlo model. Otherwise, they should have plagarized the CTP 2 model and used stacks. Give me some trained troops with good morale and some firepower, and I will show you what I can do with an Army...

D.

[vmxa1]

Well I am no great fan of the combat system, but 212 vs 211 is a toss up. What you have is 3 (212) vs 2(211), where one of the horsemen was already down 1 HP. I am not so sure I would expect to win that fight all of the time.

[Hurricane]

That first result really sounds unbelievable. That's really bad luck.

It will be impossible to replicate, however, since preserve random seed is turned off.

[Alantus]

Originally posted by Hurricane
It will be impossible to replicate, however, since preserve random seed is turned off.

Not impossible, just not very likely. If you load it up and run it enough times I'm sure it will come up again. If enough people load it up and run it once somebody will come up with devastating army loss he experienced the first time.

Myself, I will take the view the leader of the Army was either smoking something or drinking something when he "led" the army to that nasty loss. Then again maybe he just discovered he was allergic to horses.

A great leader with poor leadership, what a shame.

Alantus

[Zachriel]

Re: Could Someone Explain This to Me?

Originally posted by Gen.Dragolen
According to the combat calculators out there, there should be about an 88% chance for a veteran horseman to destroy an archer in open gound.

79%, assuming veterans and disregarding the effect of retreat.

1. Army -11 hp attacking 1st archer

One in 800. However, the average hitpoints lost in your trials was about 3. So, you start with an very odd result, similar to being dealt a "royal flush," then deciding at that point to start accumulating statistics. And you still end up with a very reasonable result. Good example of "trend to the mean." Meanwhile, no one has yet duplicated the "royal flush," further confirming that combat is working as expected. Can you duplicate the results? That might be a clue.

Considering the combat system is based on a "division" scale unit (roughly 10000 men per division) they should have gone for a simplified, abstract system, instead of the current Monte Carlo model.

Where did you learn that? In the Ancient Age, I often have 20 or 30 units on the attack, not including garrisons and support troops. Caesar conquered Gaul with six Legions, but failed to hold Britain with a like number. Monte Carlo anyone?

This is a typical example with 24 Horsemen, plus a few Spearmen.

http://www.zachriel.com/gotm6/0090bc.htm

Give me some trained troops with good morale and some firepower, and I will show you what I can do with an Army...

You already have the tools. Part of being a Great Leader is knowing how to deal with set backs. So let's see what you can do.

PS. Sorry about your Army. I pray they died well.

[Gen.Dragolen]

Re: Re: Could Someone Explain This to Me?

Originally posted by Zachriel


PS. Sorry about your Army. I pray they died well.

Zachriel,

I do not commit troops to battle unless they know they are already dead. With nothing to loose, they always die well... just too frequently sometimes for my liking.

The thing I find interesting is the sheer number of units you are able to field. Given the less than perfect starting locations I usually start from, I rarely have more than 15 units in Ancient times that aren't tied down as the Happiness Patrol in each city. Granted I tend to play on a standard size map with 8 opponents.

There are a few other oddities I have noticed about archers in the game: they seem to do better against spearmen and warriors than swordsmen. It seems a little odd, but having lost large numbers of swordsmen attacking warriors in previous campaigns, I usually keep 4-5 as shock troops before sending in the swordsmen.

I know that the Roman Legions were a powerful organization and battles like Cummae show what well led men can do. However battles like that were few and far between. And Romans fielded a large number of support troops in addition to the typical soldier. They had small cavalry detachments, skirmishers, and some siege engineers and their best weapon: professional officers.

I can understand a horde of horsemen going down to a legion, but to a lowly 2/1/1 archer ?

And the size of the units being set to 10000 persons, I was working on a mod and it occurred to me that this was the scale they intended, mostly since a Roman Legion was about 10,000 men, the Egyptians organized in divisions of 10,000 men, and a modern full strength infantry division is about 10,000 persons, though only about 6,000 are actual combat troops.

Anyways I will check out your save and compare out strategies. Should be very interesting. Thanks.

D.