On the Utility and Use of Armies

[Harovan]

One of the best ancient times army I ever had was one with 2 Horsemen and 1 Impi. That's a great mix. I used it in a lot of wars and it remained even useful after I had Knights. Later, the Impi and 1 Horseman died while the last Horseman survived, and I was able to add 2 Cavalries. This army served me well till I got Tanks.

I'm also pretty much into army usage. There's nothing better to protect a big stack of Artilleries than an army of 3 (or even 4) Infantries. To my offensive forces I usually add a Cavalry or Tank army as spearhead and nutcracker. Helps a lot against those $§"!%&*# entrenched vet Infantries, even a single Tank has a problem here. After I took the city (if I decide to take it, what rarely happens, but if the city has the Sixtine...), I place the Army there as defender. It's usually about 50% wounded, but still a though defender. Praying that the city doesn't flip. Sometimes it does, oh well, I have the MilAc. The remaining forces encircle the city and push back the counterattack. After a few turns, the Army is healed and I move on.

Armies can be very powerful, if used well. I can't understand the rants about them being useless.

[Dominae]

rpodos, early fast units (Jags and Impi) are simply amazing. I use them primarily to destroy my neighbours' infrastructure in anticipation of a future attack. The ability to move-and-pillage or pillage-and-move is great; once their roads start crumbling, their pesky Archers won't ever catch up to well-coordinated Jags. I've found that the AI has a Ancient era mobilization mode: if you threaten their cities early enough, they'll call in their Workers, call all their Warriors home, and start producing their best military unit. This obviously dampens their expansion significantly. I continue this sort of terrorism until a force of Horsemen or Swordsmen come along, at which point the AI is in no shape to defend itself.

Using this strategy, I can actually delay my GA until I have at least 5-6 cities; I simply let my Impi run away from their Archers and whatnot. Once I start taking over their cities though, I look to create Elites at every opportunity. I play on Standard-size maps exclusively, so an early GA usually translates into a superb advantage (hopefully from 1-3 dead civs by the time I run out of steam).

Because of their 2 movement, Jags and Impi are never useless until very late in the game. During the early Industrial era (when I'm usually starting a Cavalry offensive), I again use the little suckers to pillage the enemy landscape. The AI will actually send defensive units (Musketmen and Riflemen) out from their cities to destroy a nearby Jag, making the city much easier to conquer. Of course, I'll also usually try to get some Elite Jags or Impi to prove their worth in combat and become Great Leaders. I never upgrade my Jags because I think that 2 movement is better than 2 defense for resiliency; Jags will last longer than Swordsmen in the open field, meaning more chances for GLs to appear.

So, rpodos, I suggest you try fast early UUs (even War Chariots are good in my book): their resiliency and longevity really supports any GL/Army strategy.


Dominae

[Arrian]

I can vouch for the power of the war chariot, but the problem there is the lack of the militaristic trait. If your goal is to destroy a couple of neighbors, war chariots will do the trick, no problem. If your goals include destroying neighbors and generating GL's, then war chariots aren't nearly as effective as a Japanese horseman rush. I've tried - really, really, tried - to pull off early, GL-generating war as the Egyptians. I've failed. You need elites to get GL's and non-militaristic civs have trouble generating elites, especially early on when you don't have all that many units. Egypt might fare better on large or huge maps, due to sheer number of units.

-Arrian

[DrFell]

I've generated a couple GLs from early war chariots rushes (on small maps). The best thing about Egypt is that you have good builder stats in addition to the nice UU, which really pays off later on (quick change to govt, cheap temples, fast workers and development, and free shields in your bigger cities). You can develop up your core a bit faster than say, the Iroquois or Japs. I'm going to try out another Japan game soon, I think they do have a better timed GA and it will be nice to see horses from turn 1.

By the way, normally my goal is just 1 GL by 1AD for the FP. Any others are bonuses.

[alexman]

Army stats

Originally posted by rpodos

The Incredible Hulk gets an additional vet Musketman, which, as the 2 Immortals are now elite, takes the stats to 3-3-1/18, which is more than enough for even a 5-hp Musketman. I am still experimenting with this... I have gone so far as a 4 unit Army composed of Horseman, Samurai, Cavalry, and Tank!! That's still 7-4-2, which is better than the Cavalry lying around, and along the way all offensive Army units get promoted, so by the time an Army is full it will have 20 hitpoints!

Wow, is this really how army stats get calculated? Taking the average and rounding up? If so, it's interesting because, for example, an army with two legions and a warrior is the same as an army with three legions!

I always thought that each unit in an army fights (using its own stats) until it's left with one hit point, and then the next unit takes over.

[rpodos]

You are correct, each unit fights individually, with their own stats, which is much better for both the upgrade through add'l units, and for Armies that have both offensive and defensive units. Blended stats are just my own shorthand for total strength.

R

[Arrian]

There are typically only three types of armies I create:

The fast attack army: All fast mover attackers (horsemen/knights/cavalry/tanks/etc). Blitz really makes the later variations of this type useful.

The early 1-move attacker: 3 swordsmen. Oh little spearman...

The defender: infantry (or the equivalent). Yeah, come and get my artillery! I dare you.

The overall most balanced is the army of knights. Powerful on both offense and defense, with 2 moves. I've added Tanks to these later (pentagon) just for the hell of it. I don't recall actually attacking anything, though.

I really like the defender army that the AI is scared to death of. 12 or 16 hp of infantry is not to be trifled with. It's the perfect siege train escort. Actually, that's what I use armies for, mostly. Escorting a stack of artillery. I've tried using them as "nutcrackers" but after having a 12hp army of knights come within 1hp of dying attacking a veteran musketman defending a Persian city twice (I attacked, ran away, healed, attacked again, and had the exact same, heart-stoping 1hp result) I switched to the escort role.

-Arrian

[Dominae]

Anyone know the exact probabilities for generating a GL with and without the Heroic Epic? I'm under the impression that it's 1/8 with and 1/12 without: could someone confirm this?


Dominae

[Harovan]

According to Firaxis, it's 1/16 in general, and 1/12 with either the Heroic Epic or the militaristic trait. There's no probability given if you have both. I say it's 1/9 (1/16 * 4/3 * 4/3), others say it's 1/10 or 1/11.

[alexman]

Actually, Sir Ralph, the militaristic trait does not affect GL creation at all. It affects only unit promotions.

Edit: Added link to the post where Soren clarified this fact.

[Dominae]

A closer look at the Heroic Epic

I don't know if anyone else has written up a good analysis of the usefulness of the Heroic Epic, so I'll just post what I've found.

I think the real question is whether it is worth it to build an Army with your first Great Leader, or to rush build some Wonder. For those of you who are interested, using any civ you must win 22 battles with Elite units in order to be guaranteed a 75% chance of getting your first GL, and 47 such battles for the odds to rise to 95%. So, if don't get a GL after around 50 victories with Elites, consider yourself unlucky.

Once you have your GL, consider the first case where you do go ahead and form an Army and subsequently build the Heroic Epic, and the second case where you rush build, say, the Hanging Gardens (or whatever). Forget that the Army you'll have in the first case is useful in and of itself (as rpodos has already described), and that the Heroic Epic generates substantial culture: what we want to know is how fast the Heroic Epic allows you to get another GL, thereby offsetting the "loss" of the first one to the Heroic Epic. Here's a little table showing how many battles you need to win to get a good shot at another GL:

-----1-----2-----3-----

HE 16/35 32/56 47/74

HG 22/47 43/75 62/99

Here, HE stands for Heroic Epic, HG for Hanging Gardens, and the numbers across the top are the number of the GLs you can expect if you win the number of battles given in the entries (the first number is for a 75% chance, the second for almost a guarantee - 95% chance). For example, you need to win 56 battles to almost guarantee two more GLs with the Heroic Epic, as opposed to 75 battles without it.

Obviously, you should expect to do quite a bit of fighting with Elites to reasonably hope for a GL, with or without the Heroic Epic. What the table shows in general is that you'll get your GLs 12-20 victories sooner if you've built the Heroic Epic. Put another way, without the Heroic Epic you should expect GLs 2/3 as often as with it.

The Militaristic trait doesn't change these numbers. What it does do is create more Elites than non-Militaristic civs, meaning that you can battle with Elites more often, drastically increasing the rate at which you can expect GLs to appear. This advantage is a lot harder to quantify, but I think it's pretty obvious to see from just playing (personally, I've always received at least two GLs when playing a Militaristic civ).

This analysis assumes that the tests for Great Leader creation are independent. This means, roughly, that the chances of getting one early are just as good as getting one late, for any given Elite victory. I'm pretty sure this is the case.

So, I have 2 conclusions. First, the Heroic Epic isn't as effective as I intuitively thought; it does what it claims to do, but not particularly drastically. If you're planning to go to war a lot, especially wars with many skirmishes, I think the Heroic Epic is worth it. If you just want to perform surgical strikes on enemy cities, this isn't the Wonder for you (stick with the Hanging Gardens). Second, it is well worth it to "train" your units on Barbarians and other petty forces to increase your chances of getting Elites. This is why I think the Romans get GLs so frequently (in all my games, at least): the fact that Legions can do "double-duty" on offense and defense means that Elite Legions will enter (and win) more battles than other types of units, not to mention that Roman Warriors can "train" early and then upgrade to Veteran Legions.

I hope that this rather long post has (at least) given some of you food for thought.


Dominae

[Zachriel]

Re: A closer look at the Heroic Epic

Originally posted by Dominae

I think the real question is whether it is worth it to build an Army with your first Great Leader, or to rush build some Wonder. For those of you who are interested, using any civ you must win 22 battles with Elite units in order to be guaranteed a 75% chance of getting your first GL, and 47 such battles for the odds to rise to 95%. So, if don't get a GL after around 50 victories with Elites, consider yourself unlucky.

Dominae

Those are the chances of at least one Great Leader.

An easier number to remember is that if you fight eleven battles, you have a 50% chance of at least one Great Leader. If you are militaristic, you have a 50% chance of at least one Great Leader every eight battles.

[asleepathewheel]

this is a great thread, and has shown me how to more effectively use an army. Thankyou.

but...I thought there was no strategy to Civ3.
This seems like strategy to me. Is anyone accumulating this and other strategies, so they won't be lost in the mass of threads?

again, great thread

[Dominae]

Zachriel, your numbers are correct. So are mine. I was just trying to give the odds of getting one or more GLs with a high probability (75%) and near certainty (95%). Saying that you get a GL 50% of the time every eight victories with Elites doesn't mean much to me, to tell you the truth; if it helps you to think of those numbers, so be it. In any case, my post focuses on what you can expect from Heroic Epic, not how likely it is to generate your first GL.


Dominae

[Txurce]

Rpodos, I use my elite units the same way that you do, but I finally understand why you get so many more GLs than me: you have a lot more elite units! And that's because you have a lot more military, period. I build as many as I think I need to carve out a dominant position on Monarch, and that number is a fraction of yours. As Dominae spelled out, you have to fight a lot to justify going the HE/army route... but if you do, having armies is better than using those original GL's for wonders. (I still can't imagine not saving my first for a FP, though!)

Now that I'm starting to play on Emperor, I need all the help that I can get. Since I'll never pop out as many settlers as the AI, I will have to fight more - and better - to create enough lebensraum. This should lead to more GLs, and therefore to armies. Which means I'm using more of the game, making it more interesting. A good thing.

Using state-of-the-art units is certainly key, and so is having the mobility or smarts to gain tactical numerical superiority, even if your force is not necessarily the largest. (I don't enjoy having a huge numerical superiority all that much, because then the game is effectrively over.) Knowing how to mass units when you aren't necessarily numerically superior is the key to any war-game battle; in Civ3, it seems that the easiest way to gain it is by distracting the AI. The AI eventually responds to a thrust into its territory, regardless of whether it's a feint or not. But sometimes it's easier to make the AI move the wrong way forward rather than backward. You've pointed out that the AI invariably prefers to attack wounded units and workers; it also homes in on undefended cities. I will sometimes take a city garrison and move it out three spaces on a road. The AI goes right for it. As you said, the goal here - apart from taking a consequently less guarded enemy city - is to lure the enemy units into open ground, within range of (usually) mounted units. This gives you the sort of advantage that Dominae cited, helping you to win a war of attrition, even against a larger enemy.

Dominae, you mention my two favorite early units, the JW and the WC. Pillaging is a great feature of these units (as well as the Impi). The retreat factor is also huge in the very early game, since no one has that many units. In theory I prefer JW's to Impis, because being twice as cheap is more important than defending twice as well, given that you plan on taking the offensive most of the time.

Rpodos, with regard to both these units, you again mentioned an early GA as an argument against. Why? To me, the GA benefits don't last long enough to seem all that meaningful... and I might even prefer them at the start, when I can use all the help I can get. (I only play on standard maps as well.)

Zachriel, that image of your armored armies is as unforgettable as Aeson's perforated world map.