Unit Strengths by Era

[gamma]

I don't do much fighting in the ancient era. If I do, it's horsemen and spearmen, possibly with swordsmen, with adjustments for various UUs. I've never used chariots. Hoplites and legions are very hard to overcome.

In the medieval era, it's knights vs. pikes, or longbows if you can't find horses or iron. It quickly turns into knights vs. muskets. Late medieval turns into a cavalry-fest, right at the point where roads are still the main transport and riflemen haven't yet come out. That 3-square movement is the big factor.

In the industrial era, riflemen make cavalry more bearable. A quick beeline for infantry makes you downright comfortable for a long time, until tanks finally come out late in the era. The naval action also really heats up. Ironclads have a long period of dominance before destroyers and battleships finally arrive. Too soon to tell the effect of air power, since until the patch came out we were pretty much at the mercy of bombers.

Not much experience in the modern era yet. In the early part, it's almost all tanks, mech inf, and artillery. I have almost no use for marines and paratroopers are really hard to use if the enemy can instantly ship 30 tanks over to destroy any that come in.

Overall, resources haven't been a problem for me. I usually cover enough land that I'm bound to have whatever I need. In my first real game, I had no horses, which made ancient war very slowwww. Then I later had no oil, which made things touch-and-go when I faced tanks with nothing but cavalry and a few cannon and infantry (thank god I had rubber), but still doable. So resources are a welcome challenge IMO.

I try to avoid any combat until the industrial period, when I have rails. Rails make force deployment much simpler. It's all about movement, which is why the fast units are so dominant. They give you a 1-2 punch of better deployment options, and retreat. Retreat is proving to be very powerful; people use this to take virtually no casualties in a campaign.

I really like the right-of-passage feature, and not being able to use an enemy's roads and rails. Evens the playing field.

My biggest peeve with the AI is that it doesn't seem to recognize a major invasion force, and doesn't respond in a way I'd consider proper. Given the way wars are typically waged in the game, the AI shouldn't spread its defenses so evenly; or it should at least adapt.

I'd expect an AI to first destroy incoming roads and rails from the enemy civ ASAP, either with bombardment or with fast units. Destroy them on the enemy's squares, of course. Prioritize squares with rough terrain, and move extra defensive units into cities the enemy could move to in one turn, as far as it can tell. For example, a border city with enemy grasslands and plains nearby should have just as many defensive units as possible

More importantly, the AI really needs to use bombardment units more effectively. Right now, it seems to put one catapult, cannon, etc. in each city, to auto-bombard any incoming unit. This is almost completely uneffective vs. large invasion forces, particularly fast units. The AI should instead recognize such waves for what they are, and deliver as much bombardment as it can on top of that force, before moving in with conventional units.

Offensive bombardment should work similarly. The AI should lean more toward concentrating bombardment on a target before attacking it with its conventional units. Right now, if I have a civ with only cavalry, artillery, and riflemen, I can defend very successfully against an AI with tanks and mech infantry. I like that I can do this - it's still a bit tricky - but the AI really ought not to roll over when it sees this being done.

It may be that the AI is more adept at war in the road era, and simply isn't very good at it in the rail era. Maybe this is what could be improved.

One other thing the AI is lax about: it doesn't upgrade its old units nearly as expeditiously as it should. Was this to keep it from dominating the game if it has a tech lead? If so, it should keep enough money on hand to upgrade pronto if a superior civ attacks it.

[vmxa1]

I agree that marines are of no real value as it is now. I have never made them. The idea that a well trained marine would have any trouble with a knight in battle is ludicrous. The marine would have a field day with a target the size of a horse loaded down with armour. Now you have a dismounted and slow knight with a sword or pike verse an automatic weapon, hum.
I would like to see a filter to allow me to remove obsolete items from the build list or if you make them go away after a period of time that is fine. I hate seeing warriors in my list while I am making tanks.
I am thinking that bombers have too much value now. This is due to the fact that thy can be relocated to any city and have a huge range. In civ2 you have to fly them from place to place and it took time, not to mention the out of gas part. If at least it took an extra turn before they could be used it would reduce their effectiveness.

[dac]

Ok, I'll dissent. I like Marines. They're the only unit that can invade a city right off the transport. I've flanked many an AI by having the Marines hit what I deemed an important coastal city a turn or two after I've invaded from another direction.

Helicopters would be more useful if you could use them to "pick-up" units in the field and return them to a city. Paratroopers drop in, pillage the roads to a strategic resource, hold out a turn, then helicopters swoop in and evacuate the paratroopers, mission accomplished. Or use the helicopters with infantry both for entry and egress.

I agree with slowing cavalry down, and lowering their attack a bit.

[David Weldon]

I'd like to second the idea that coastal fortresses get a range of 2. They might almost be worth it then. Good suggestion!

I'd also like to slip in my thoughts on units and resources (sorry for those that have heard the argument in other threads):

Obsolete units being competetive vs. modern units was a design decision to prevent people without resources from being completely screwed. I understand the sentiment but think that it was a poor implementation. Instead...

Create "non-resource" versions of all units that simply cost more. You can balance the extra cost so that resources are just as important as they are now, and then you can introduce a more powerful obsolecense mechanic that will prevent all the whining about frigate v. sub, or spearman v. tank, etc.

Vel:
I'll call your Impi wager and raise it one Mounted Warrior rush. Since both units cost 1 pop to create, and mine has a 3 attack (vs. your pathetic and shriveled little '1'), I'm pretty confident.

[Illyrien]

What I REALLY want is to declare units obsolete so that they dont appear in my build ques.

Having Longbows, immortal etc on my buildques is not only pointless, but actually quite annoying. Especially when the my governors start building these units.

So make an option in the Military Adviser to declare (and undeclare) units Obsolete.

[Velociryx]

Aack! Deliver me from Impis! Hehe....it's Aztec Jaguars, and you're on (erm...eventually...lol...if/when there's MP). I won't deny that you've picked a fine choice (#2 UU in the game, behind the Jag, IMO), but I think that, in such a game, if we started relatively close to each other, then before you could even research Horseback Riding, you'd find yourself facing several score Jags.....that's why my money would be firmly on the Aztec. Iroquois have a great unit, but the Aztec one is available from 4000 BC.....OUCH!

-=Vel=-

[Seth]

Originally posted by Jeffrey Morris FIRAXIS
Any upgrade holes?

I haven't had time to finish reading this thread, so my apologies if I'm repeating someone else's post....

I like to play as the Persians--their civ attributes agree with my style of play and their UU kicks. But it is frustrating not to be able to upgrade a warrior. You end up just disbanding them. I'd appreciate it if I could upgrade them to swordsman.

TIA

Peace

[Arrian]

I personally think the problem with Cavalry supposedly being overpowered is primarly due to the AI's inability to defend itself properly, not the unit itself. However, if the AI isn't gonna get any better at fighting in the industrial/modern eras, Cav should probably be weakened in some way (lowered a/d/m or higher cost, OR perhaps no knight->cav upgrade, or even putting it in the industrial era, attached to the same tech as riflemen). As it stands now, I punch out a bunch of knights and as soon as I get cav "poof" I have a large, very powerful force. No need for footsoldiers at all. I annihilated 2 entire civilizations (1 large, 1 medium sized) last night with roughly 20 cavalry. I think increasing either 1) the shield cost or 2) the MAINTANENCE COST of horse units might be a good idea. What if they cost 2 gold a piece?

You could argue the Mounted Warrior is a bit too powerful, but generally only if you poprush. I don't. The Immortal is nasty, but it's slow, so leave it be. The Musketeer.. ha, it's a joke. Who attacks with a defensive unit? Either leave it be but allow upgrading, or increase it to 3/5/1. The Cossack doesn't seem to offer much of an advantage, but if Cavalry is weakened, and Cossacks are left alone, then it's a solid UU.

Ironclads, IMHO, should be resticted to the coast. They were never exactly seaworthy vessels. Besides, the period of time in the game during which building frigates makes any sense is quite short. I think a minor tweak may be necessary to prevent Battleships losing to galleons and such (I don't mind occasionally taking damage from older, weaker units, but it does happen too much).

Someone mentioned the AI's tendency to ignore combined arms stacks approaching their cities. I brought that up in Vel's strategy thread, because I think it illustrates perfectly the AI's major fault - it cannot mount an effective, mobile defense. This ties in with the failure to mass artillery and use it effectively, along with it's inexlicable love of longbowmen and inability to use horse units the way we do to chip away at an advancing force. Also, the AI MUST UPGRADE IT'S UNITS! Why were my Cavalry fighting spearmen in 1/2 of Japan's cities last night? They had musketmen and Samurai (I saw and killed exactly 3 of those)!!

-Arrian

[Sirian]

Historically, sailing ships of the frigate and galleon class ruled the seas for centuries. When ironclads finally came along, they instantly made all wooden ships completely obsolete, but I have never even heard of any ironclad battles except that among the two prototypes. Naval warfare went from that very quickly to destroyers and battleships.

In Civ III, frigate class vessels sit too late in the tree, while ironclads come way too quickly in the industrial age. Result? The frigates never get used except by civs with no coal. Ironclads are often then dominating the seas in huge fleets for centuries. The ironclad ought to be something with a much shorter lifespan. Dole them out with The Corporation, that should give some life to wooden naval warfare and shorten the span of ironclads except for those who have no oil.

If the reason they are handed out with steam power is "historic realism" I urge you to look at the RESULTS of that decision, and tell me how realistic they are. Something there has gone awry.


Musketmen are too weak. Increase their defense to 5 and give them a 25% chance of disallowing retreat. Give pikemen a 10% chance to disallow retreat. Give riflemen a 30% chance of disallowing retreat, and increase their defense to 7. Give infantry and paratroops a 40% chance to disallow retreat, and give marines a full 50% chance to disallow retreat.

Alternatively, if you dislike the idea of building a chance to stop retreat into the unit itself, then do it with the artillery. Make any unit that subjects itself to bombardment upon attack then unable to mount retreat. Thus, a city defended by four catapults would be able to prevent the first four units that attacked it from being able to retreat. Any after that, could do so. You would then also have to increase the AI priority on building artillery, but it would allow combined arms to take some of the bite out of the dominant retreat option.


Move Carrier back units to Advanced Flight. Move paratroopers to Flight and drop the airport requirement, but increase their cost to 120 shields. Leave their range moderate, and require oil.

Increase the AI priority on using transports full of marines to attack coastal cities. They currently only bother to build marines at all if they don't have oil for tanks. Give marines an attack bonus when making an amphibious landing, except against fortified garrisons or cities with coastal forts. Double might be too much, but right now these units are just not important at all in most situations. Think about France and D-Day: the Germans had every mile of coast fortified. In Civ III, the AI doesn't even consider fortifying its coasts, and can be COMPLETELY stopped from invasion by merely posting a unit in each square along a coast. Get control of your landmass, become invincible. If the AI actually USED marines, that would put an end to that. If it fortified its coasts, or even SOME coasts near cities, with forts and garrisons, that would make taking them over from the sea a whole new ballgame.

Also, as someone mentioned, they tend to rush all their surplus units into battle, leaving behind only minimal city garrisons. What happened to the Civ2 principle of building forts around their land and garrisoning those? This AI has a lot going for it, but there is still plenty of room for improvement. Considering that forts DO offer ZOC, the AI doesn't have enough priority on building them and manning them. Once you smash their surplus forces, you just roll from city to city in a matter of a few turns, with knights, cavalry, tanks, or armor. Never anything to contend with except cities and an all-too-predictable form of attack with surplus units. Perhaps the AI's should be given some priority to build forts on resource squares, man them (even within their borders), and even station workers there, as someone pointed out, to rebuild roads. Perhaps have them station some of their captured workers, if they have any? Those don't cost any maintenance. Or even just one worker. Could there be some way to have them identify strategic locations for forts? Defensive terrain, proximity to cities or resources, identification of vulnerable coastline?

I liked the idea of allowing Helicopters to retrieve foot soldiers from the field, not just drop them off.

Could you give radar artillery ZOC everywhere within its range? Right now, there's very little incentive to upgrade to Radar Artillery. The standard howitzers are plenty strong enough, there is not much to recommend the radar version.

Could you give AEGIS cruisers a visibility range of 3? Perhaps still only 2 vs subs, though. Oh, and could you PLEASE allow nuclear subs to carry standard cruise missiles as well as tac-nukes? And perhaps even allow Aegis to carry a cruise with them, too? Right now, they are nothing more than weak battleships that can see subs. You haven't captured their historical persona in the gameplay.

Also, why are Stealth Fighters nothing more than cheap/weak versions of stealth bombers? Why can't they FIGHT? They ought to be able to execute anti-air-superiority missions, to clear out enemy fighters so the bombers can go in. That's what they are there for in real life: to fight more than to bomb, although they can bomb, too, with a JDAM, for instance. If they weren't meant to kill enemy aircraft, they'd be called bombers, wouldn't they?


- Sirian

[player1]

THE BEST SOLUTION TO CAVALRY PROBLEM

Localization of problem: 2+ movment units shouldn't be used as "mobile artillery"

Historical fact: You can't attack walled city (or uban city) with cavalry units effectivly.

Solution: Disable retreat ability for 2+ move units when attacking walled, 7+ pop. cities and forts.

Watered down verion of solution (just in case...): give them 50% chace of retreating in above case.

Exception: Blitz units (tanks) should keep full retreat ability.

Good points: It is simple. It is easy to progaram.

Bad points: Civ players need new strategies to consider (this applies to AI also).

[David Weldon]

Originally posted by Velociryx
I won't deny that you've picked a fine choice (#2 UU in the game, behind the Jag, IMO), but I think that, in such a game, if we started relatively close to each other, then before you could even research Horseback Riding, you'd find yourself facing several score Jags.....that's why my money would be firmly on the Aztec. Iroquois have a great unit, but the Aztec one is available from 4000 BC.....OUCH!

"... an African swallow, maybe. But no' a European swallow, that's my point..."

I didn't know you were planning to start off right next to me! Geez, there you go changing everything just 'cause you're Impi isn't as big as mine. (OK, I know it's Jags now, but I just love typing "Impi")

On a more serious note, I really prefer MWs because they are effective long enough to take over the whole world/continent. They tear through knights and Cavalry with 'nary a problem and they're Sooooo much cheaper. Only fortified Riflemen finally slow them down.

This all raises a great point about units that maybe Jeff can answer:

When will Multi-Player units be available?

[pcasey]

I've done a lot of thinking about this game recently, a lot of it spurred by the a lot of the excellent information that came out in my anti-cavalry thread.

I've sort of modified my understanding of the problem.

I still think cavalry is too fast, but I think the reason its so utterly dominant isn't so much a feature of a broken unit (its moderately broken imho), as it is a feature of a game that doesn't give enough advantages to teh defender.

I hate to say this because it drove me crazy in civ ii where city walls could make a city essentially unconquerable for huge swaths of the game, but I think Civ III makes cities way too easy to take.

Consider the modern era because its the easiest to see the problem in action.

Lets say we have a defender, Alice. She's being attacked by the evil empire of Eve.

As a defender, Alice havsto garrison, say, 5 border cities heavily and 10 interior cities tolerably. Lets say she has 50 defensive units. That lets her put 2 in each interior city and 6 in each border city.

As an attacker with railroads, Eve can infinitaly concentrate her entire attack force on any one of Alice's cities. So, she takes her 20 modern armor and proceeds to start attacking one of Eve's cities. Since it has only 6 defenders, 10 or so modern armor attacks later, and zero casualties, Eve now has control of one of Alice's cities and 10 damaged modern armor.

Eve promptly razes the city to the ground and retreats all her damaged armor back into her own empire.

Eve still has 10 undamaged modern armor left though, so she goes ahead and sacks another of Alice's cities for good measure, again at zero loss to herself.

The issue comes down to this. CIV III lets the attacker infinitely concentrate his forces against only a subset of the defender's forces. This makes any defense where one attacker = one defender impractical.

Civ II dealt with this by giving defenders huge, huge bonus's. Just try digging mech infantry out of a city wall defended city with anything that doesn't ignore city walls and watch the butcher's bill rise.

Civ III doesn't give the defender these kind fo advantages, and hence its vulnerable to a concenration strategy.

Now this is going to be an issue with any turn bases stategy game. It was also a critical issue with any board war game (yes with the little carboard counters, I'm old school). Because of teh phases nature of combat, it made a stategy based on engaging and defeating a defending force in detail incredibly powerful.

Until somebody had a bright idea that pretty much fixed the problem.

STACKING

All we need to do to fix this problem is put a stacking limit in place in the game.

You can stack infinite units in a city, 5 units in a fortress, 3 units in friendly terrain, and only 2 per square in unfriendly terrain.

This way I can't rush 20 cavalry down one road at one of your cities. If I tried, they'd be strung out along the road in a column and vulnerable to a counterattack on your turn, as they should be.

[David Weldon]

Gotta say I love the "can't retreat if artillery bombarded you" idea.

I hate the stacking idea, but it makes a lot of sense, it would be effective, and it would make me change my strategies a lot. Can we somehow make sure that this never happens?

[Arrian]

The stacking thing might work... but it sure would be annoying (particularly if it applied to workers too). I'm not sure about it.

In your example, Alice's main problem is that she is attempting a static defense, just like the AI. Static defense is a great way to lose, as your example shows. Mobile defense, however, is another matter. If Alice also had 20 Modern Armor, she could 1) track down Eve's retreating, beat up MA or 2) go do the same thing to Eve's border cities.

Quick question (and this is gonna make me look stupid, 'cause I really should know this): Does having a mobile unit in a stack prevent retreat by an attacker, even if the attacker is hitting a 1-move unit in the stack? If so, the real issue here is getting the AI to put a few mobile units in its cities, instead of endlessly patrolling its land.

-Arrian

[Velociryx]

LOL @ David's post....yeah...good word that...."Impi"

I too, am a big fan of the Iroquois Mounted Warrior....hadn't really given the civ much consideration until I read Sauron's *excellent* piece on them, and then, the more I got to thinking, the more it dawned on me that he was dead on.

Early on too, I saw a few people heaping praise on the Jaguar, and I was thinking to myself...WHY? It's just a crappy little 1-1 unit!

Then I actually played a game with them, and Ohhhhh Lordy, what an eye-opener that was.

Those guys were *everywhere!*

Even better, because they're so cheap, I didn't have to pop-rush at all to make them....a size 2-3 city could crank 'em out in all of 2-3 turns, so once I got half a dozen cities set up, I had this continuous stream of new Jags.

And with so many hordes of 2-move little guys scampering all over the map, I had near-perfect knowledge of my starting continent in no time.

Once the "continental-mapping-project" had been finished, I simply selected the biggest, baddest civ on the continent with me, made use of my fast moving troopers to consolidate them, and charged.

Losses didn't matter. Damage per attacker didn't matter.

It was a beatdown.

The surviving Jags from the first war were, of course, all Elites, and those were combined with the latest crop, and turned their attention on the next biggest Civ....repeat till the continent was soaked in blood.

Of all the UU's in the game, the two I feel will either be outright banned or the most sought after will be the MW and the Jag. Amazing units, both!



-=Vel=-
(who has....Impi Envy?) LOL

[pcasey]

Arrian,

Having a mobile unit in the stack doesn't prevent retreat unless you're actually attacking that specific unit. Thus if my 3 horsemen attack your 2 spearmen and a knight, I can retreat out of the two spearmen fights, but the knight fight is to the death.

You're absolutely right about Alice trying for a static defense. The problem is though, since each side alternates moving, its the only defense possible. She can't react to eve's advancing horde by rushign reinforcements to the threatened city. After all, its not her turn.

In any turn based game, all you can do is line your army up in good terrain, ride out your opponents assault, and then react on your own turn. Mobile, reactive defense is impossible. Since civ III lets the attacker strike me anywhere they please, the effectively means defense is impossible.

[Arrian]

pcasey,

Ok, I see your point about the nature of turn-based combat. However, how about a rule change (similar to the no-retreat if bombarded) that disallows retreat if there is an enemy mobile unit in the square you're attacking? The only problem I see here is that a horseman could prevent a tank from retreating and that would be a bit silly.

Maybe a combination some of our suggestions would work best. A slight increase in defense modifiers for cities/walls. A chance of failed retreat. A slight increase in the cost of mobile units. Add it all up and it may help balance things better.

Toss in some tweaks to the AI - the suicidal charge through my army to get at a worker needs to stop, for instance - and warfare gets a lot more challenging.

EDIT: Perhaps just as important as anything else, the AI needs to fortify its borders! A couple of defensive units placed strategically on hills, forests and mountains near the border could really slow me down.

-Arrian

[Comrade Tribune]

Retreat ability makes Fast Units nearly immortal. Pop Rush with Fast Units kills everybody.

Longbowmen are somewhat weirdish: Same Attack as Rifles simply doesn´t make sense.

Muskets and Rifles seem extremely expensive, especially compared to Fast Units.

Catapults and Cannons are unimpressive -very cheap, but also very weak.

Walls are also unimpressive, for the same reason. This feels wrong: Ancient Walls were a formidable obstacle. This is also one more reason why you don´t need Catapults.

Special Bonuses are missing: Bombardment Units should ignore Walls (but Walls should be stronger). Spearmen/Pikes/Muskets/Rifles/MechInf should have a Defense Bonus against Fast Units.

[Seth]

I'm very interested in the discussion of the dominance of the Jaguar, having missed the original conversation on the strategy thread (too much posted here to read).

One question: don't you waste your golden age by building those units from the start?

TIA

Peace

[player1]

Originally posted by Comrade Tribune
Walls are also unimpressive, for the same reasons. This feels wrong: Ancient Walls were a formidable obstacle. This is also one more reason why you don´t need Catapults.

Look my solution to Cavalry units (several threads up).

[randomturn]

I agree with many of the comments posted already. IMO, though, the game has much bigger balance problems than that created by units. The difficulty levels are not properly balanced. The Wonders are not properly balanced. The civ traits are not properly balanced. I hope Firaxis keeps at it, though, since Civ3 is already very good in a lot of ways.

[pcasey]

Seth:

Sure, your golden age blows early on, but so what? A well orchestrated jaguar rush can win you the game before the ancient era is over.

Of course, you can do the same thing with a horseman rush, or a war charior rush, or an impi rush, or, well, any unit with movement two. It works like clockwork, its just boring.

[DaBear]

I like the idea of navel battles with the Man of War but as mentioned previously on this thread they don't stay viable that long. So I only build 1 and put it with a privateer then move them out in the open to get GA without war.
I would also like to have great leaders (Admirals) from water battles and use it join water vessels into a navel task force.

[David Weldon]

Seth, seth, seth, ...

You shouldn't get confused by Vel's ramblings. We all put up with him in these forums, but we all know he doesn't really know anything or contribute to the community. Jaguar Warriors only dominate in his little "Jag Warrior Dominance" mod. You can see that he doesn't really have much of an imagination either, judging from the name anyway.

To answer your question, any ancient UU causes a smaller GA, but there are a few of factors that make it not such a big deal:

- Since growth is exponential during the expansion phase (while there's still unlcaimed land), a few shields and trade early on can be the equivalent of a lot of shields and trade later in the game.

- Most of the ancient UU's, especially the Mounted Warrior (which is far superior to the JW, btw) would be worth using even if they forced you to forfeit your GA. The extra cities, tech (via peace negotiations), and gold they create for you, not to mention the sheer growth that's possible by using them to conquer cities, is worth far more than any small penalty you incur by having an early GA.

Oh, and I've come to my own conclusion through extensive testing that the JW might be a very strong unit if you start "relatively close" to your opponents. Or so I've heard...

[Velociryx]

I agree with the general consensus that AI border fortification would solve much of the problem, and I steadfastly hold to the conviction that every strategy can be beaten.

Example:
I have strongly agreed with the notion that mounted units are, with auto-disengage, overpowered in the context of the game, and yet, overpowered or no, they CAN be countered. Fortification of borders is a part of that.

Problem:
The nature of the turn-based game seems to leave a situation where fluid attack is possible, but fluid defense is impossible. This is because, as has been stated here, an enemy can mass his troops and strike with pinpiont precision, while the defenders troops are more-or-less evenly (or at least widely) dispersed, thus, forced to react on the following turn to unexpected aggression.

Solution:
The crux of the problem then, becomes not so much that fluid defense is impossible, it is that fluid defense AT THE CITY GATE is impossible, and the solution lies in simply forcing the issue to some other locale (ie - fortification and staffing of the border).

Now, it is still true that the attacker can mass his forces fluidly. It is still true that he can overwhelm selected points OF that line. That part of the equation has not changed.

What HAS changed though, is the potential gain/harm that will result because of that combat. Now, when the position is overrun, no city (production center) is lost, however, much of the attack force (assuming proper staffing of border forts) has lost its momentum. Worse, said force must end its turn exposed (even on favorable terrain, a defending unit of (wounded!) cav is no match for a fresh unit of attacking cav. Thus, the attacker has tipped his hand (attacked), exposed a large portion of his attack force (his turn ends), and must now bear the brunt of the defender's counter-attack (fluid defense).

This is actually a variant of "Doctrine-Defensive" from the SMAX guide, except instead of ringing your continent with sea bases and using them as "spotters," you're building fortifications at relevant borders and staffing them with an increasingly deadly array of nasty things.

If the attacker decides it's not worth the trouble to "punch through" since there would be no immediate gain (that is to say, no immediate city capture, nor a degradation of the defender's ability to crank out troops--raze a city), then the defender wins by default, having averted the attack before it began.

Or...no?


-=Vel=-

[Velociryx]

Heh! David, I like you more all the time! Anytime you're passing through Columbia, drinks are on me....I think we'd get along famously!



-=Vel=-

[Arrian]

See, just when I thought I'd explained something pretty well, Vel has to come along and explain it much better

I really think that the solution isn't to mess with the units, but with the AI. I think I could defend against a massed attack... I might take losses, even a city or two, but the counterattack would be vicious... and I think I'd end up winning. The AI is hopeless once the human has the initiative.

As for MP balance... that's a discussion that should only occur once (if) there is MP. Look at all of the theoretical discussions about things in Civ III that took place prior to the release of the game. How many, in retrospect, were accurate or worthwhile?

-Arrian

[sophist]

Does the AI know how to handle choke points in geography? I've seen some instances of what could have been that sort of behavior, but I've also seen the AI extend itself beyond an isthmus where cities would be more of a liability than an asset. If the AI knew:

1) how to expand better with regard to strategic geography

2) how to keep prospective attackers at arm's length so they couldn't just swoop in and grab 5 cities in one turn

3) use mobile units effectively

4) effectively attack after the ancient era, mainly with the infantry/mobile/artillery stack that's becoming all the rage

Then we wouldn't be having this discussion. The four strategies above put together would make the AI a much more formidable opponent militarily and it would become clear that the unit stats aren't the problem.

It would also help if the AI were able to manage its research better; for example, in a lot of situations, it makes a lot of sense for me to go straight for Military Tradition because then I can get a serious military advantage over the AIs and use that to catch up on the other techs I've missed. If the AI better understood the value of that tech in relation to, say, Printing Press, and

1) moved more aggressively towards it (knowing that a human player would be likely to do the same)

2) tried to get it from other players (AI or human) as soon as possible

3) did not share out the tech except for exorbitant cost

that would also go a long way.

[Sevorak]

Hmmm...I like the idea of defensive artillery preventing retreat.

Defensive artillery, despite being a last-minute throw-in, has a lot of potential as a game concept. At this point, however, it's a little too weak since it only gets to fire at one opponent per turn, and at a ROF of 1. Why shouldn't a defender have to use combined arms too?

Oh, and perhaps this is a bug, but as I've mentioned before, it's not only land. Ships act as defensive artillery for each other in a stack, and from what I can interpret from the readme that's not intended.

Specific comments.

Vel,

I agree regarding fluid attack and fluid defense, and I'm wondering if the solution isn't Civ2/SMAC style ZOC, granted only by fortresses. That would make border defense really border defense, in that every fort along the way would have to be engaged and defeated before the forces rip in, burn down the city, and leave.

One shot ZOC doesn't really cut it, not much of a threat considering its low chance of success. Fortresses seem to be kind of an afterthought, just tacked on - Soren once said that the AI generally doesn't build fortresses because their value is "debatable" or something like that. I mean, any unit passing a fortress has a flat 50% chance not to be hit at all. Then, the unit in the fortress uses its attack value (which is not always too impressive) to attempt its advantage shot. You're sort of stuck there in terms of fortress garrisoning - you either stick a unit in there that can defend against a direct attack, or you put one in that might actually hit with its opportunity fire. Or, you could stack more units in, but then you're just dispersing your defense more.

Without the Civ2 style ZOC lock, even fortresses are of minimal use, since to guard a border that way would be to occupy every tile on the border with a fortress and unit. Otherwise, they ignore your fortresses, their only risk being a single 20-30% chance at a single hit point off a single target.

Just a thought.

Sirian,

Totally agree regarding your wood hulled ships comments, their presence there is totally nonsensical. I'm actually rather annoyed at the way the ship stats are done overall, though I've beaten that particular issue to death elsewhere on the boards.

As for the 'stealth fighter' - since they're basing the technologies and units on real-world things (excepting the spaceship - note how even Fusion Power is gone from Civ2) they use the F-117A as their example, which is why it can't perform air superiority missions. There has never, in real life, been an operationally deployed 'stealth fighter' per se, just a 'small precise stealth bomber' and a 'large imprecise stealth bomber'. You can see they use the F-117A as its graphic, and the real-life F-117A carries no combat air-to-air radar, and most importantly, no air-to-air weaponry either - indeed, it cannot carry air-to-air missiles. So if you really think about it, the lack of air superiority from a stealth 'fighter' is completely realistic.

-Sev

[gus_smedstad]

I'm going to repeat the comment about fast units, but from a different perspective.

Artillery units generally aren't cost effective, at least not on the attack. They do add a bit to the effective defense strength of the defender.

If we're attacking a city with a barracks, we must kill all the defenders in a single round. Wounds don't count.

We also want to use a technique that generates no (or very rare) casualties for the attacker. That means either fast units or lots of artillery. Stacks of infantry generally take casualties unless they are much more advanced than the defender.

2 Horsemen (60 shields) per defending Spearman is generally enough to win.

You need 6 Catapults plus an Archer (140 shields) per Spearman to have a reasonable chance of winning without casualties. Each catapult has a 2/3rds chance of causing 1 HP damage, and about half the time they hit the town instead of the archer. We need to knock the Spearman down to 1 HP if we don't want to risk losing the Archer.

On the plus side, the Catapults / Archer require no special resource. However, they have a much higher upkeep (7 vs. 2), are slower so reinforcements take longer to arrive, cost more shields, and the catapults often miss and mess up the town instead.

Even if Catapults never hit the town, it's still 3 Catapults and 1 Archer per spearman. If the defenders are Vetrans, it's much worse - you need 50% more Catapults.

The next artillery piece is the Cannon, and the contemporary units are Cavalry and Riflemen.

2 Cavalry (120 shields) is generally enough per defending Rifleman.

Cannons cause 1 wound 4/7ths of the time on Riflemen, and again miss about half the time. You need 6 Cannons and a Longbowman (280 shields) per rifleman.

The next artillery is Artillery, against Infantry and Tanks.

You only really need about 1.5 Tanks (150 shields) per Infantry.

Artillery hits Infantry 6/11ths of the time, but shoots twice. You need 4 Artillery plus a Rifleman (400 shields) per Infantry. That probably wouldn't give you low-casualty results, either, since the infantry only wins 2/7ths of the rolls. Artillery + Cavalry is a better bet, provided you have saltpeter, but don't have rubber and oil.

The final one is Radar Artillery, vs. Mech Infantry and Modern Armor.

You need about 2 Modern Armor (240 shields) per Mech Infantry.

Radar Artillery hits Mech Infantry 8/17ths of the time, and fires 3 times. Including misses, you need 3 R.A. and a Tank (460 shields) per Mech Infantry. You can't really use any regular infantry against Mech Infantry, since all regular infantry types have attack values that are so low that they'll still lose significant numbers to Mech Infantry with 1 HP.

In short, all artillery needs to be about twice as good as it is now, since the value of the hits on the town instead of the unit are very low to the attacker. That value isn't zero, but it's at most elminating the 25% bonus for size 7+ cities.

- Gus