Darn the blasted combat system

IIRC Civ2 prevented it with Fire Power. Just need something to attenuate the rng rolls to a smaller range. Combat should not be able to swing so wildly that A beats B with no damage, B beats A with no damage. It is fine for A or B to win, just seems nuts that it could vary by that much. I could be wrong, but it is annoying.

This is exactly it. It's not that it's RNG, it's that it's the variance level of the RNG.

So a system which permits rolls of 1-1,000 can indeed roll you 10 1s in a row - that's RNG - it's just that 10 1s in a row would be ultra-rare.

But a system which permits rolls of 1-2 is also RNG - it just makes rolling 10 1s in a row much more likely.

So just saying "that's RNG folks" is entirely erroneous. As if RNG is some kind of answer-all to any awfully designed algorithms is immature at best and reckless at worst. And it's always been evident that there's something very wrong with Civ3's combat algorithm.
 
Agreed that the combat values of units with different technological values in Civ 3 are far from realistic, but if these values would be far more realistic (here a lot could be done at least to the result of the battle by adding an overwhelming number of additional hitpoints to the technological more advanced unit) the game in my eyes would become boring by battles between technological different advanced units without any risk.

When such an event happens, that a galley would really sink an ironclad in Civ 3, the ironclad run on a riff, or the machines or ammunition on board did explode, or one ironclad rammed the other, or the ironclad run out of ammunition and coal, or the complete crew was drunken, or all these events happened all together. :)
 
Last edited:
People who moan about RNG are mostly taking unprepared gambles where a slight bit of bad luck would change the outcome of the game. They attack without proper massing of troops or fire support. Who cares if you lose a tank or 2 to spears when you have another 10 tanks attacking that same city? And if you're attacking rifles without artillery support that's on you.

As for lower tech units being able to destroy higher tech ones, this happened regularly in real life. The technologically inferior factions may have lost wars or fought long insurgencies, but they nevertheless inflicted casualties on their tech superior opponents. Civ3's combat is actually quite accurate. War is frustrating. Esp. for would-be conquerors expecting quick and easy victory met with fiercer than expected resistance.

Another factor would be game play balance. Civ3 is meant to be played with the AI roughly 1 age ahead of you. If that's not the case you're playing on a too easy difficulty. That's just how bad the AI performs. It needs the tech advantage. But that also means that the tech inferior human player needs to be able to beat the AI through smart tactics and strategy. Having things like cavalry killing mech inf (with fire support) is a necessity.
 
Bad RNG is a problem when it comes early in the game. Later as Predator145 observes it barely makes a difference. In my last game I experienced a redlined regular pikeman defending in a plains unwalled size 6 city (not across a river and the city most definitely did not have civil defense) kill one attacking veteran cavalry and beat off a second (it retreated with 1hp remaining). My rough calculation of the odds on this is somewhere around 2000:1 against. It made no difference as I had several more cavalry available and the next finished off the valiant pikeman.

Had something this unlikely occurred earlier in the game I would have been much less tolerant. Forces are smaller so bad RNG losses are more significant. Moreover it is usually not possible to assemble overwhelming force and if I had I should really have attacked a lot sooner.

At the start of a game bad RNG can be a disaster and the fight that led to it unavoidable. Does one accept the result and the consequent delay to your victory or even possible defeat? I suppose it comes down to your chosen play style and the rules you set yourself.

Technology is sometimes quite fragile so the availability of tanks could be quite low. In Civ mechanical problems are probably not factored in. But then every unit has a set of ratings, is given a name and has a set of images related to it. The images and name influence players' perception of the unit's capabilities even though the actual numbers are in plain sight. The disconnect that hit me most was in Civ4 where a spear formation inflicted some small damage to an attacking helicopter gunship. How, I asked myself? It also jolted me out from immersion in the game.
 
In Finland last century we saw battles where tanks lost to guys on skis with rifles. I'm happy with the variance and I see the AI get just as bad luck as humans (except for barbs where default settings have them being gentler on the AI).
 
except for barbs where default settings have them being gentler on the AI.
Barbarians are for AI as for humans on regent: Regent setting gives +200% against barbarians. Below regent humans have an even larger combat bonus against barbarians than on regent.
 
In Finland last century we saw battles where tanks lost to guys on skis with rifles. I'm happy with the variance and I see the AI get just as bad luck as humans (except for barbs where default settings have them being gentler on the AI).
You make the classic mistake of assuming the combat units actually represent micro-versions of real military units.

They don't.

Like everything in a Grand Strategy Game of this type, everything you see is just an abstract representation of power-level.

Societies don't suddenly wake up one day and 'discover' things like horseback riding, it's just an abstract way of saying this tribe has now 'mastered' horseback riding & it implies that their power-level over a tribe without has increased.

When you see a unit on the screen, it's not 'a battalion of Horsemen', it's a visual representation of the local expansion of the military influence of the tribe that wields them.

This is why in Civ2, units had power-ratings more than just hit-points. Civ3 fails on both counts by having no power ratings, so disrupting the abstraction, but also fails to take in to account unit volume as a replacement, which, ironically, means that the only way to 'strategise' in this strategy game is to increase volume, but not because increased volume increases overall power-level, but because increased volume mitigates the poor rationale behind the combat system.

The whole point of having Horsemen is a reward for advancing your empire at a speedier rate than the opposition, they represent a significant power increase of everything. To then have that nullified by a tribe who has, almost literally, done eff-all for 1000s of years, and then say "well that's RNG folks! Ever heard of General Custer?" is missing the whole point by quite some country mile.

You are supposedly 1000s of years more advanced, have better cash-flow, better societal organisation, stronger culture, the whole shebang, and yet a tribe who's pretty much still in the early bronze-age can field not only just as many men as you, but have them be more than adequate at completely nullifying any great advantage you've made, you're not significantly better off fielding 300 shield's worth of Horsemen than you are fielding 300 shield's worth of Archers. You could even argue that you're worse off militarily as you'd only have 10 Horsemen rather than 15 Archers.

You could have a wonderful pedantic debate about what's better, 15 Archers or 10 Horsemen, but, in reality, there shouldn't even be a debate. The AI itself rates 15 archers as stronger in it's own programming as it just sees 15 x 2 versus 10 x 2. It's absurd from quite a few perspectives. The fact that you can get to Knights and/Medieval Infantry and still lose or red-line huge numbers of them when attacking a Bronze-age tribe, is the next level of absurdity which takes it into the 'obviously a bit wonky' territory rather than 'Lol random'.

That guy who said he likes it because it allows him to game the system is a completely different topic & anyone is free to like such a system, but liking the system doesn't mean it's not a technical train wreck. :lol:
 
I posted the game's CRT ("Combat Results Tables") HERE. You'll see that, even at the highest possible odds in favor of the Attacker, that the Defender has a 1% chance of winning a round of combat. Also, don't forget that combat in the game, using heavy Defense modifiers, e.g., a Metropolis on a Hill = (Defender Base Strength x 150%) PLUS a 1 column shift for both an attack across a River AND a 1 column shift if the Defender is fortified = (at highest Odds, and depending upon the number of HPs each, for Defender and Attacker) a 1% - 13% chance of -

:spear:
 
That is one reason why I change attack and defense values, but also add hitpoints for higher tech level units.
 
Horsemen are not a step up from archers if it's static combat. They're contemporaries. If you've blown your cash trying to get HBR without thinking it through it's on you. There are times when you'd be better off with more archers supported with catapults or pure archers if it's very early game where you only have like 100 shields worth of units for your first rush. Just like there are times you'd be better off with a cheap foot unit medieval composition instead of expending resources on expensive knights that require an optional tech. The choice of tech advancement is not a no-brainer.

Techs aren't auto win in Civ3. From ancient age to medieval age the units scale proportionally stat and cost wise. Your bronze age spearmen can hold vs knights if you have the numbers. The advantage from pikes and even more so muskets (because they suck stat-cost wise) is support cost. Otherwise the emperor dif. and above AI would absolutely stomp you. There is simply no way to keep up in tech with the AI and have a balanced game. If the AI is on par with you, you've already long won.

"The fact that you can get to Knights and/Medieval Infantry and still lose or red-line huge numbers of them when attacking a Bronze-age tribe, is the next level of absurdity which takes it into the 'obviously a bit wonky' territory rather than 'Lol random'."

That guy's preparation for war must be horrible. Sending in like 3 knights vs a city with 3 spears is asking for that kind of frustration. The guy sending in 10 is not gonna mind losing 2 if he's unlucky with RNG. And if you bring trebs you'd be losing zero most of the time.

The game already has enough tech induced power spikes like cavalry, tanks, bombers and all modern age units. That's more than enough to reward those that sink shields into multipliers to keep up with tech. These units do not scale proportionally stat-cost wise. 100 shields for tank vs 90 shield infantry? A fast offensive unit with more than 50% atk stat compared to the defender's def stat? That'd be like having mounted warriors costing 22 shields compared to spears. What a bargain. That's your reward for having sunk resources into tech right there. Most modders would actually curb those spikes to balance gameplay out because it's very predictable and repetitive to play the game around them in order to win in the fastest time possible.
 
You make the classic mistake of assuming the combat units actually represent micro-versions of real military units.

They don't.

Like everything in a Grand Strategy Game of this type, everything you see is just an abstract representation of power-level.

Societies don't suddenly wake up one day and 'discover' things like horseback riding, it's just an abstract way of saying this tribe has now 'mastered' horseback riding & it implies that their power-level over a tribe without has increased.

When you see a unit on the screen, it's not 'a battalion of Horsemen', it's a visual representation of the local expansion of the military influence of the tribe that wields them.

This is why in Civ2, units had power-ratings more than just hit-points. Civ3 fails on both counts by having no power ratings, so disrupting the abstraction, but also fails to take in to account unit volume as a replacement, which, ironically, means that the only way to 'strategise' in this strategy game is to increase volume, but not because increased volume increases overall power-level, but because increased volume mitigates the poor rationale behind the combat system.

The whole point of having Horsemen is a reward for advancing your empire at a speedier rate than the opposition, they represent a significant power increase of everything. To then have that nullified by a tribe who has, almost literally, done eff-all for 1000s of years, and then say "well that's RNG folks! Ever heard of General Custer?" is missing the whole point by quite some country mile.

You are supposedly 1000s of years more advanced, have better cash-flow, better societal organisation, stronger culture, the whole shebang, and yet a tribe who's pretty much still in the early bronze-age can field not only just as many men as you, but have them be more than adequate at completely nullifying any great advantage you've made, you're not significantly better off fielding 300 shield's worth of Horsemen than you are fielding 300 shield's worth of Archers. You could even argue that you're worse off militarily as you'd only have 10 Horsemen rather than 15 Archers.

You could have a wonderful pedantic debate about what's better, 15 Archers or 10 Horsemen, but, in reality, there shouldn't even be a debate. The AI itself rates 15 archers as stronger in it's own programming as it just sees 15 x 2 versus 10 x 2. It's absurd from quite a few perspectives. The fact that you can get to Knights and/Medieval Infantry and still lose or red-line huge numbers of them when attacking a Bronze-age tribe, is the next level of absurdity which takes it into the 'obviously a bit wonky' territory rather than 'Lol random'.

That guy who said he likes it because it allows him to game the system is a completely different topic & anyone is free to like such a system, but liking the system doesn't mean it's not a technical train wreck. :lol:
I'm unaware of any reference that would back up your theory. If it was all about mastery of a technology or a military strategy (e.g. horsemen) rather than an actual division or regiment of humans then why would drafting a military unit reduce the population if it doesn't represent a physical regiment of human beings? Similarly, but on a tangent for workers, settlers and slaves.

Why would a galley be able to carry 3x units only regardless of whether they were all horsemen, all archers or a combination of both?

I appreicate that non drafted units do not have a population implication, but I'm not presenting my arguments as the undeniable 'truth', I'm simply saying there are grounds for suspecting that interpreting units as representing actual regiments of little people fulfilling a specific military function and that to do so is a perfectly reasonable interpretation than a mistake. Particularly as in Civ3 you can stack units of different functions in the same stack to effectively build an army of little people that consists of units with different offensive and defensive values. If it was Civ1 where you can't stack units I'd perhaps have more time for your theory.

So I'd argue that a group of Finns on skis repelling a Soviet armoured division could be represented in Civ3 by 1x rifleman or infantry icon repelling an attack by 1x tank icon.

That said, I'm happy to be corrected if this is referenced anywhere. I'd also completely forgotten about the variation with Civ2 combat and I don't think I had any issues with that either (Civ1 was more of an unsatisfying lottery given the lack of healthbar). I suppose it is an argument for perhaps modifying health points across the board for units based on their era (-2 ancient, -1 medieval, +1 modern), so your post has given me food for thought.

The problem is reduce health too much and combat becomes more of a lottery (like Civ1), removes the tactical consideration around recovering health and devalues barracks.

Conversely, to increase health makes combat overly tedious and significantly buffs the militaristic trait for the AI (assuming the AI prioritise sending injured units to cities with barracks).

Ergo, I think the default combat system in Civ3 is simple, understandable, fair and more than satisfactory.
 
I suppose it is an argument for perhaps modifying health points across the board for units based on their era (-2 ancient, -1 medieval, +1 modern), so your post has given me food for thought.

The problem is reduce health too much and combat becomes more of a lottery (like Civ1), removes the tactical consideration around recovering health and devalues barracks.

Conversely, to increase health makes combat overly tedious and significantly buffs the militaristic trait for the AI (assuming the AI prioritise sending injured units to cities with barracks).
I would argue that barracks are a standard in the modern age. Recruiting in a city without barracks will be the exception by then.

It is of course possible to change the standard hp. It could be 2 for conscript, 4 for regular, 6 for veteran and 8 for elite. That way militaristic would be strengthened.

If we follow this, then -1 for ancient units, + 0 for mediaval units, +1 for industrial units and + 2 for modern units may make up a reasonable whole.
 
I would argue that barracks are a standard in the modern age. Recruiting in a city without barracks will be the exception by then.

It is of course possible to change the standard hp. It could be 2 for conscript, 4 for regular, 6 for veteran and 8 for elite. That way militaristic would be strengthened.

If we follow this, then -1 for ancient units, + 0 for mediaval units, +1 for industrial units and + 2 for modern units may make up a reasonable whole.
I have combat experience set to 2 for conscript, 5 for regular, 8 for veteran, and 12 for elite. This is based on my study of military history.
 
In theory the increase in HP is fine, but in game play I hate it. My memory of some of the mods with larger HP was annoying as combat all took so much longer and a lot of extra actions by me. Maps with thousands of units, adding 2 or more extra HP to deal with is no joy. I agree it makes units "fit" better. Later era units should be harder to take down by prior era units. This
additional HP is great on a std map on most difficulty levels. Less so for highest levels and larger maps and especially AW type games. IOW the more combat, the less one will enjoy the extra HP.
 
Top Bottom