Build barracks later - Most of the time

[rah]

Agreed, as long as you survive to create the 8 experienced axemen. In some of our MP games, if you wait, you're dead.

[Yosho]

Originally posted by rah
Agreed, as long as you survive to create the 8 experienced axemen. In some of our MP games, if you wait, you're dead.

Fair enough. It does depend though; if I have 6 experenced axemen and you attack me with 8 inexperenced axemen, I've got a pretty good chance of winning.

[snoopy369]

You're forgetting that cities grow, and military units cost more later on, while the barracks always costs the same.

Turns to barracks at size 3 / 4 hpt: 15 turns.
Axemen built in that time: 2.

Turns to barracks at size 8 / 20 hpt: 3 turns.
Macement built in that time: 1.

I'd much rather build the barracks at size 8 and have the extra axemen earlier on, when they're really important. Trading the 2 axemen for 1 macemen is not such a difficult trade.

Certainly if i'm making 8 axemen in one city early on, and know I can afford to wait for 2 of them, I'd build the barracks... but early on you might not have that one production city, and instead make 2 axemen each from 4 different cities. Building the barracks then would mean you have 1/2 the number of units that I do, not 3/4...

[Ingot]

The higher the difficulty or the more proficient your enemy, the more the need for earlier
troops and the lesser time to create early barraks. If you use your troops wisely,
you will get a promoted army in spite of having barraks.

Barraks are nice but, it could be that delaying their creation may be one key to increasing your play level.

[Yosho]

Originally posted by snoopy369
You're forgetting that cities grow, and military units cost more later on, while the barracks always costs the same.

Turns to barracks at size 3 / 4 hpt: 15 turns.
Axemen built in that time: 2.

Turns to barracks at size 8 / 20 hpt: 3 turns.
Macement built in that time: 1.

I'd much rather build the barracks at size 8 and have the extra axemen earlier on, when they're really important. Trading the 2 axemen for 1 macemen is not such a difficult trade.

Certainly if i'm making 8 axemen in one city early on, and know I can afford to wait for 2 of them, I'd build the barracks... but early on you might not have that one production city, and instead make 2 axemen each from 4 different cities. Building the barracks then would mean you have 1/2 the number of units that I do, not 3/4...

Well, you're not trading 2 axemen for a maceman, though. Your trading 2 axemen for a maceman and experence points in every unit you would build in the city between the axemen and the maceman.

(shrug) If you need the axemen right now, build them right now, of course. But if you don't, building a barraks first at least in your main production city or two is very helpful.

It's also worth mentioning that a barracks is much better with agreesive civs.

[joncnunn]

My own rules of thumb (SP only):

Scouts / Explorers : Never need a baracks
Units built as policemen: Probably don't need a baracks
Units built as defenders *:
If Warriors, Spears, Archers: Probably don't need a Baracks.
Axemen: Might or Might not need a Baracks. It depends.
Any other defensive unit: Probably needs a baracks
Units built as attackers: Almost always need a Baracks

* Note that I classify counter units engaging the enemy in your own territory as defenders for the above even if they are actually attacking.

[realpolitic]

Originally posted by VoiceOfUnreason
Broken math. Go back three squares.

Problem #2: CIV4's combat model is non linear. Therefore treating a 20% attack bonus as 20% more hammers is very suspect.

True, but if you can take a city with 3 unpromoted axmen, then you'll have 2 cities in which to build axmen. You'd be lucky to take that city with 2 level 2 axmen, which would take longer. Eventually cities become too hard to take that way, but barracks often come in useful before that, when finaces become an important issue, but when you only have a few cities, caputuring cities increases income, and the economic spoils of war are good, too.

If you know you have access to Copper or Iron, and can build only chariots or barracks and will have Bronze Working soon, building barracks might be a good idea, although a settler might be best.

[Blake]

Quite true.

Nominally, the strength of units increases by approximately the square of the strength ratio.
Ie a unit which has 20% more strength,is 44% more effective.

But also, army effectiveness increases with numerical superiority... this is a bit non-linear, but it's approximately squared for numbers between 1-1.5.

Ie if you are fighting 50% battles:
You have 100 units, your enemy has 100.
Turn 1:
You attack with 100 units, 50 live, 50 die.
Turn 2:
Both sides have 50 units, the defenders heal, the attackers promote (it ends up kind of even)
And then it fights out until 0 units (on average) remain.

So when you attack with equal unit count and equal unit strength, you can win the battle, but lose everything...

Now try it with 200 units vs 100 units.
100 units attack, killing 50 units, 50 die.
The next 50 units attack, killing 40 units (They're badly injured), 10 die.
11 units attack killing the last 10 units with 1 loss.

In total, you take out 100 units with 61 losses. In effect, the numerical superiority made your units 63% stronger.

But obviously also, the last 39 units didn't actually do anything.

So the idea is, you need to hit the sweet spot where you have enough units to win convincingly.

The sweet spot depends both on odds and on unit counts.

Sometimes a single promotion can bring a fight from ~30% odds to ~60% odds (50% odds rarely happens), this DOUBLES the chance the unit survives. In that case, losses in the first round of fighting are halved.

If you strongly suspect you can push units from 30% to 60%, then the promotion is well worth the investment in a barracks. Some cases like this are Axemen or War Chariots vs Archers in flatland cities, the promotion often pushes you over the threshold. In contrast, Chariots wont get pushed over the threshold, so you should just build more chariots...
And likewise if the enemy has cities on hills, axemen and war chariots wont be pushed over the threshold, so you'd be better off with more...

Over all it's pretty complicated :P.

On final sum, Quantity is the best kind of Quality, but individual unit quality shouldn't be discounted, in particular because of things like expenses - a higher quality army costs less to maintain. That favors quality. IN contrast, if you're building your units not to keep, but to throw away, you needn't care about the long run cost of keeping them around, so if quantity makes sense, quantity makes sense. This is precisely the case of a chariot rush, the idea is the chariots DIE to take the enemy capital. You end up with nearly no chariots, but 2 capitals, and one less opponent. So the chariot rush is a special case where barracks really don't make much sense (both for reasons of odds thresholds and reasons of upkeeps), it can be the same with non-aggressive Axe rushes too, depending on how quickly you rush.

[realpolitic]

If the odds are 30% for the first unit attacking a given unit, there's a good chance that an equally powerful second unit will have at least a 80% chance, which at the crux of my quantity arguement. Further, if the unit has a 30% chance, and survives, it will get more EPs, and the promotions get progressively more expensive

2-3-5-7 etc.

I agree it's complicated.
(OT)
Sometimes people want the absolute best at any price, in life, rampant consumerism,
and I suspect in Civ, too.

[Yosho]

Originally posted by Blake
Quite true.

Nominally, the strength of units increases by approximately the square of the strength ratio.
Ie a unit which has 20% more strength,is 44% more effective.

But also, army effectiveness increases with numerical superiority... this is a bit non-linear, but it's approximately squared for numbers between 1-1.5.

Ie if you are fighting 50% battles:
You have 100 units, your enemy has 100.
Turn 1:
You attack with 100 units, 50 live, 50 die.
Turn 2:
Both sides have 50 units, the defenders heal, the attackers promote (it ends up kind of even)
And then it fights out until 0 units (on average) remain.

So when you attack with equal unit count and equal unit strength, you can win the battle, but lose everything...

Now try it with 200 units vs 100 units.
100 units attack, killing 50 units, 50 die.
The next 50 units attack, killing 40 units (They're badly injured), 10 die.
11 units attack killing the last 10 units with 1 loss.

In total, you take out 100 units with 61 losses. In effect, the numerical superiority made your units 63% stronger.

That's true, and it's even more true when defending (it takes a much bigger force to take out 3 fortified axemen on a hill then 2 fortified axemen on a hill). But it depends on tactics as well; with proper tactics, a smaller force of more experenced axemen will get overwhelemed by a larger force if it's caught out in the open, but so long as you have a defendable place for them to heal you can sometimes wear down the larger force with low casualties because your guys can win one on one fights.

[couerdelion]

Originally posted by Blake

Over all it's pretty complicated :P.

On final sum, Quantity is the best kind of Quality, but individual unit quality shouldn't be discounted, in particular because of things like expenses - a higher quality army costs less to maintain. That favors quality. IN contrast, if you're building your units not to keep, but to throw away, you needn't care about the long run cost of keeping them around, so if quantity makes sense, quantity makes sense. This is precisely the case of a chariot rush, the idea is the chariots DIE to take the enemy capital. You end up with nearly no chariots, but 2 capitals, and one less opponent. So the chariot rush is a special case where barracks really don't make much sense (both for reasons of odds thresholds and reasons of upkeeps), it can be the same with non-aggressive Axe rushes too, depending on how quickly you rush.

I’m not entirely convinced with the quantity over quality argument in the case of early chariot rushes. I probably need to run a few tests on WorldBuilder to have a closer look but my scepticism is founded on the empirical evidence that this numbers game works very different when you are not massing stacks of 100 against your enemy on the open plain.

The thing with the early chariot rush is that it requires a little time to get in place. Your nearest rival will certainly have 2 cities and may even have three. With luck they will not have any metals but they will have at least two archers in each city and another 2 or 3 spare archers.

So somewhere in this rush you are going to have to hit at least four fortified archers in a city with some cultural defence. Your chariots are going to have something like 5-15% probability of winning (+10% retreat). With the odds you might get in this situation, I would think that at least 3 chariots are needed for each defender – or 12 in total. With a single promotion for each chariot, I think that 10 should be enough with the hammer cost of the barracks being saved by the smaller army needed – which still has to be paid for while you are building it.

Add to this, the fact that other cities need to be taken and I would believe that the barracks becomes very important.

I can see that the argument falls down a little if a significant part of your army is being built in more than one city. Here the cost is of two barracks or at least one set of unpromoted chariots. But with the need to take several cities I would probably still argue that the barracks is needed in the city building the larger part of the army.

Like I said, something that needs to be investigated more.

[Arrian]

In the case of an early Immortal rush (not that I do that often or anything ), I tend to build barracks. It's fairly easy to whack a barb or two on the way to the enemy, giving me the 2 added XP I need for a combat II (or combat 1/shock) immortal.

-Arrian

[VonSharma]

I don't know, I don't play offensively in the beginning, but I can build about 2-3 cities with only 1-2 warriors. Unless its barbs, other civs tend not to attack that soon. I think I was only attacked once by the Germans before 1000 BC.

[realpolitic]

Chariots get a 100% bonus against axmen, so it might help to have a few in reserve in case you need to fight them, as I do. That bonus makes defending with eigther archers exclusively, or some combination superior to defending with just axmen.

[wodan11]

Chariots only get their bonus when attacking, not defending

Wodan