Wang (King) Kon

[Robber Baron]

I've been playing the Koreans, lately, trying to get a handle on workable strategies for playing what appears, on the face of things, to be one of the most challenging PTW civs. Some observations follow. (Note: I've been playing at Monarchy level, but much of the analysis should hold at higher levels.)

First, let's review the basic civ set up. You are scientific and commercial. Kind of like the Greeks, I suppose. Except.... no hoplites! Instead, your UU is the Hwach'a -- a bombardment unit with the firepower of artillery but the range (and cost) of a cannon.

The Big Picture
Your civ traits lend themselves well to projection of power. The commercial trait in particular should pay off handsomely in the later stages of the game, when you will wring more productivity than most other civs out of an extended empire. And the scientific trait gives you a boost both in building culture (key early in the game) and science capacity (which comes into play later in the game).
But some aspects of growth will be tricky to manage. In particular, you will depend on a combination of Great Wonders to trigger your Golden Age (since your unit, unmodded, will not). And you will have to figure out how to grow your empire to the point that your civ qualities will start paying out discernable advantages. In the later stages of the game, a well constructed Korean empire can be extremely formidable. But you have to get there without depending on some of the advantages other civs enjoy.
Initially I assumed the Koreans were built for building. But I've found (in my small and admittedly unscientific sample of some half dozen games) that in fact, I have better results via military expansion. This civ becomes most potent and most powerful when it attains critical mass both geographically (continent-sized at least) and chronologically (late-middle ages on).

The Early Game
You start out with the Alphabet and Bronze Working. There is your first (and for a long time your only) advantage over other civs: these are both very potent starting techs. Bronze Working allows you to build spearmen and/or the Colossus from the beginning of the game -- and puts you one tech from Iron Working. Alphabet is even better. You can hoard it, research writing, and run for either mapmaking or literature -- meaning you can target either the Lighthouse (powerful, on certain maps) or the Great Library (even better). Skillful pre-building should land you one.
Or, if you tend to avoid building the ancient wonders, you can broker Alphabet, which usually fetches a fair price (unless you start out near other scientific civs, of course). You can use it, for example, to trade for warrior code and take out a neighboring civ in an early archer rush. (Better trading the Alphabet, of course, than the Bronze Working. No sense in setting up future conquest victims with spears.)
Very early culture will be an expensive proposition, on the other hand, particularly if you hold onto both alphabet in the hopes of landing the Great Lib and bronze working for military purposes. It will be awhile (literature) before you can start building anything cultural (unless you can broker something for Ceremonial Burial).

The Colossus might be a better build, for Golden Age purposes.
Which brings me to ....

Your Golden Age
Your UU will not trigger a Golden Age. Korean players need to manage wonder building especially carefully. You have to build a commercial and a scientific wonder to get golden.
Um, sort of. Actually, to tell the truth, wonder GA triggers remain somewhat mysterious to me. In one game, I built both the Great Library and Colossus and found myself catapulted into a very early (and despotic -- ) GA. (Made sense, though: GL is sci; Colossus is commercial.) In another game, I built the Colossus, then captured the Great Library, then launched a golden age when I built Sistine's Chapel (using a great leader). (Maybe because I built a commercial wonder, then captured, not built, a scientific wonder; which set up a trigger whereby the game was ready to dispense a GA as soon as I built any other wonder?)
Scientific wonders are relatively easy to come by (though remember, Copernicus is not scientific). Commercial wonders can be trickier. Colossus is early -- takes a focussed strategy and some luck. If you miss that, you have to wait for Adam Smith (I think) -- which frequently comes in the midst of AI cascades, and (for me, in many games), when I'm still hurrying to catch up to the AIs techwise.

The Hwach'a Gotcha
This UU is a bit of a tweener. 40 shields to build, bombard strength of 12, range of 1. In other words, same cost and range as cannon (though you don't need iron to make 'em, just saltpeter), but the firepower of artillery. Actually, that's a pretty potent, if limited, unit. It is noticably more effective at bombarding cities than cannons (which I find to be too costly to be worth the investment, except as un-upgraded artys). By implication, the Korean player should be able to assemble effective combined arms city seiges at a point in the game that otherwise tends to bog down into defensive warfare. (Note, too, the Hwach'a is a wheeled unit, needing roads to travel through rough terrain.)
Of course, combined arms invasions take a lot of resources and a bit of time -- another reason this civ is best played big.

The Korean Sweet Spot
The Koreans may struggle a bit to keep up through most of the middle ages. But if you can get them into the later stages of the game and you have acheived critical mass, you should find yourself in position to pull away from the AI civs. Universities are cheap, and those slingshot bonus scientific advances create powerful brokering opportunities (especially if you land nationalism). You can manage a large number of cities better than most other civs (because you're commercial). And it all comes together at that point in the game in which I tend to find myself taking control anyway (for example, I start to put together an economy and tech infrastructure capable of outresearching the AIs).

Other observations?
What wonders have triggered GAs for you? Anyone have better luck than I did playing Koreans as builders?

[joncnunn]

Re: Wang (King) Kon

Correct, the way the peaceful GA code works is:

1. If you've just built any GW & haven't had a GA yet then:

2. If you empire now posesions any wonder of combo of wonders that represents both your traights then your GA is trigered.

If you build the Colossus, but the AI takes over the city for a turn, and you then complete the Great Libary, this won't trigure your GA, and you'll have the recapture the Colosusus & and build another wonder to trigure your GA.

Originally posted by Robber Baron
In another game, I built the Colossus, then captured the Great Library, then launched a golden age when I built Sistine's Chapel (using a great leader). (Maybe because I built a commercial wonder, then captured, not built, a scientific wonder; which set up a trigger whereby the game was ready to dispense a GA as soon as I built any other wonder?)

[Robber Baron]

Thanks for the clarification, jonc!

The results, when everything comes together, can be very satisfying:

[GusSmed]

Re: Wang (King) Kon

Originally posted by Robber Baron
By implication, the Korean player should be able to assemble effective combined arms city seiges at a point in the game that otherwise tends to bog down into defensive warfare.

What are you talking about? When the H'wacha shows up, it's Knights vs. Musketeers, which isn't a period of defensive warfare, and before Riflemen show up, Cavalry is available.

Both of these periods are exactly like the ancients era: Horsemen vs. Spearmen. Attack strength is the same as defense strength. The only real difference is that cities are now common, so the adjusted odds are 2:3, or 40% (+25% for fortified, +25% for city or walls) instead of 2:2.5 (44%). Actual casualty rates tend to be around 15%-20%, due to retreats.

In fact, the advent of Cavalry makes this period the most aggressive until Modern Armor, since Cavalry can move 2 space through enemy territory and attack.

The only era of defensive warfare I know of is the period between the discovery of Infantry, and the discovery of Tanks. During this period Cavalry has a hard time (typical odds are 6:15, or 28%, and 40% casualty rates). However, that's also when Artillery shows up, making the H'wacha obsolete.

I can't imagine any reason why I would build even 1 H'wacha. Instead of a fast striking force of Cavalry which typically attacks from safely within friendly territory, I'd have a slow force of Riflemen / H'wacha / Cavalry which typically gives up 2 free turns of counterattacks before it can attack an enemy city.

Otherwise, I agree with your assessment. It's nice getting a clarification of how Golden Ages get triggered by wonders, by the way.

- Gus

[joncnunn]

When Calvary first appears, if you've beelined for it, you can take a few cities before Nationalism.

Nationalism is the first of the techs that tilts things backs towards the defender:
Mostly by increasing the ciy's unhappiness when your in a threatening postion (2 squares are closer). If you instead stay at 3 squares so the AI doesn't think the city is threatened, you don't have retreat with your Calvary.

The second is Steam Power. The AI can now relocate troops raipdlly to the front lines and can replace the units if don't take out all their defenders in a single turn.

The third is Sanitation. Once Hospitals are built, Metroplexes start appearing, and the odds significantly favor the fortited riflemen in the metroplex. If that city has a Hill, you will need bombardment if you don't enjoy watching your Calvary die.

This is where the Korea's UU shines.

The forth of these is the Replaceable Parts with the show-stopping Infentry. (Or else a massive multi-turn Artillery bombardment to reduce the population.)

[Robber Baron]

Gus: yes, I realize cavalry extend the window for mobile warfare. I was thinking more along the lines of that period of early nationalism, when riflemen pop up in enemy civs (almost always, seemingly, just as I'm trying to finish off a last round of expansion before a modern era building spurt). Sending in waves of cavalry against cities defended by riflemen is prohibitively costly. The Hwach'a enables another, judicious round of pruning along the borders. You're not going to mow down a major superpower. But you will be able, under certain circumstances, to extend expansion at a time when few other civs will be able to do so (assuming rough tech parity).
Think of the situation as gaining access to an attack power nealry comparable to the infantry/arty stack of doom (with cavalry for pillaging expeditions and to mop up against damaged enemy units).

[GusSmed]

You guys must be playing some other game than I am. I've never had any real trouble with Riflemen if I've got Cavalry.

Sending a wave of Cavalry against Riflemen isn't significantly more costly then sending a wave of Horsemen against Spearmen. It's not even as bad as putting Horsemen against Hoplites, and I don't flinch from that.

It's only when you're up against Infantry that the casualties start getting nasty, and as I pointed out earlier, at that point the H'wacha is obsolete.

It's true that Railroad means that you pretty much fight the defender's entire mobile force if you don't take a city, but I prefer to go in with sufficient force. I never take more than one turn to take a city in any era, anyway - it's bad tactics, and greatly increases the expense of an attack.

I include multi-turn bombardments in that, by the way. Even if the target doesn't have a barracks, you lose most or all of the damage you've done to enemy HP by waiting a turn. An efficient, cost effective attack with Cavalry / Infantry / Artillery mix takes place in a single turn. Cavalry is only part of that mix so you can attack while the main force is 2 tiles away, the range of artillery, instead of waiting for the Infantry to get adjacent.

Before Infantry, you're losing a lot of the value of Cavalry if you're not using their full 3 square attack range. If you attack with enough force, instead of piddling your strength away, you can move defensive units (i.e. Riflemen) up to defend the units left outside the target city after you win.

If you consistently treat Cavalry as if they were Knights, i.e. move into enemy territory before attacking, it's no wonder the slow speed of a H'wacha attack doesn't bother you.

It only makes sense to move adjacent to an enemy city before attacking once you've got blitz units, tanks and modern armor, since they can get a second attack with the extra movement. Before then, the best defensive terrain in striking range is preferable.

The very best defensive terrain, of course, is one of your own cities, inside your cultural boundries. Cavalry can frequently move 2 squares along road, and 2 squares inside enemy territory, striking from a square 5 tiles away.

I don't understand the remark about Nationalism increasing a city's unhappiness. Are you referring to some rule I'm unaware of?

- Gus

[Zero]

Gus what difficulty are you playing at?

[GusSmed]

I almost always play at Emperor. That's the highest difficulty that I can win reliably, i.e. barring a start in the middle of the jungle or the the desert.

I can beat Deity, but it's just too frustrating to play at that level regularly.

- Gus

[Charis]

Ah

> I never take more than one turn to take a city in any
> era, anyway - it's bad tactics, and greatly increases
> the expense of an attack.

Really? I would say it's time to take a step up in difficulty, but you've indicated deity is too frustrating.

Given your preference for blitzreig tactics and the assumption of having plenty of cav around compared to the number of defenders, it would be frustrating.

Not using combined arms *increases* the cost of attack? Not really - unless of course you're only using cav, in which case the counteratttack when you take too long would be brutal.

What's the cost of Vet cav vs regular rifles? Let's asssume a size 7 city with three defenders, standard fare. Cav win 32%, retreat 32%, lose 36%. You'll need to bring six cav to take the town in one round, and you'll lose two of them. Losing two cav per town is not an insignificant expense.

Bombard those rifles down to 1hp, and heck, double the number of defenders - The cav win 81%, lose 19%, you only need seven to take out six defenders, and expect to lose one. (Numbers from a combat simulator)

You inflict twice the casualties on the enemy and take half for yourself! So where is the increased cost to attack? Counterattack? Let's say they have cav (probably longbows, but we'll give them a strong counterattack). You have rifles in your slow stack, and let's say several cav counterattack, and you have a few rifles on defense. You're going slow, so avoiding open ground, right? (Good!) The AI prefers regular units, while we prefer vets. Togethter, that makes the most likely outcome: half the cav die, you have 1/3 as many losses as they do! Plus your units are cheaper. Counterattack? We fear no counterattack!! Those numbers don't even take into account bombardment defense -- add some Hwatch'a to the stack and 2hp cavs counterattacking lose 62% of their units, and their losses are ten times yours!

Maybe this is slow and frustrating - I'll admit that stylistically this approach isn't for everyone, but combined arms attacks are quite efficient, minimize losses, and inflict massive pain on the AI.

It gets better! As you take the first few cities the AI will throw all the offense it has on you. Any cav they have will not sit back in the core, but come charging. You see strong counterattacks for less than a half-dozen rounds, often much less, and have FREE REIGN after that, with more offensive units coming up one at a time to promote your rifles, as they are built.

It's quite enjoyable, with the right mindset, to wage war with only a modest amount of troops, and take out a much larger and stronger civ. It's also the key to deity becoming a LOT less frustrating. In my last deity game I got into war with the world superpower and had about 1/10 the army he did. There was no question of "cranking out tons of cav", I was majorly outgunned. But after a few pure defensive rounds just staying alive, I used combined arms, slow advance warfare to take a half dozen cities including his capital

I think on higher diff (or in general, where you can't get away with simply overwhelming the foes with numbers), for a player who likes to see his opponent squirm (rather than just 'disappear'), the Koreans and the Hwatch'a are a formidable tool.

> I don't understand the remark about Nationalism
> increasing a city's unhappiness. Are you referring to
> some rule I'm unaware of?

Oh my! You *REALLY* do like cavalry don't you???
With Nationalism comes drafting. The AI is *totally* draft happy. If you come within two squares of his city he will draft it down mercilessly. So bad you often have little choice but to raze it or wait until Modern era for unhappiness to wear off (well, starving down to 1 and keeping it there is fine) You wouldn't notice this though, strictly using cav or MA attacks, as you're often 3 squares away when you start the round of attack.

Gus maybe you've tried all this and it's just not your style - if so that's fine. But if you've not, or other lurkers who haven't, try it for a change of pace, take the Koreans even, force yourself to build several DOZEN Hwatch'a, and give it a whirl.

Charis

PS The more masochistically minded should check out:
1) Using French Musketeers as the *offensive* force along with large artillery stacks...

http://forums.civfanatics.com/showthread.php?s=&threadid=14518

2) A game with no speed 2 units made, ever.

http://forums.civfanatics.com/showthread.php?s=&threadid=14286

[Arrian]

I haven't played the Koreans yet. They are, to me, Greece with a UU that is both better and worse (better because you don't have to worry about an wasted GA if some AI rushes you early on, worse because the Hoplite is pretty powerful). To be honest, I see the Com/Sci trait combo as extremely weak. But bravo to those who are playing them and figuring out how best to use them.

On the subject of bombardment & Cavalry...

Bombard units aren't particularly useful, IMO, unless you MUST fight a seriously entrenched foe, which to me means Cavalry vs. Infantry, or Cavalry vs. Riflemen in a large number of 7+ size cities on hills or something. In that case, I'd use arty. In just about any other situation, I'd rather have more attack troops.

By the way, it was my understanding that H'wacha's have leathal bombard, and can thus trigger a GA. Am I wrong on that?

-Arrian

[GusSmed]

Originally posted by Arrian
By the way, it was my understanding that H'wacha's have leathal bombard, and can thus trigger a GA. Am I wrong on that?

Yep, you're wrong. They'd be mildly interesting if they had lethal bombard, but they don't, so you cannot trigger a GA with one.

Originally posted by Charis
Really? I would say it's time to take a step up in difficulty, but you've indicated deity is too frustrating.

Very glib, but it's not the late medieval, Cavalry and Riflemen that are frustrating. It's the early game where you're locked out of trading, and by the time you're ready for a horseman rush everyone else is well into Feudalism and the like.

So do you play Diety all the time? If so, what is your early game like? Do you go for swordsmen, spearmen, and catapults?

Not using combined arms *increases* the cost of attack?

I think you meant to say "Using combined arms", instead of "not using", since as phrased it contradicts the rest of your argument.

In any case, I didn't say any such thing. I said that taking multiple turns to make an attack, i.e. multiple turn bombards, or attacking a city and not taking out all the defenders, allowing the remainder to heal, that greatly increases the cost of an attack.

However, using combined arms does increase the cost, in the sense that the force you need to take a city is much higher, though the casualties are lower.

This is the trade off for bombardment units. All of them, including artillery, are very weak. Particularly against cities, where they outright miss the defender 50% of the time. However, they cannot lose. Given sufficient force, bombardment minimizes casualties.

The catch is that the amount of bombardment required to pull this off is hideously expensive. On the higher difficulties, you don't have the luxury of that production advantage until you're already winning anyway.

Bombard those rifles down to 1hp, and heck, double the number of defenders - The cav win 81%, lose 19%, you only need seven to take out six defenders, and expect to lose one. (Numbers from a combat simulator)

Could you point me to the combat simulator? I've done stuff with spreadsheets, but it would be nice to see one that took all factors like automatic promotions into account. It's too tedious to do more than the initial odds by hand.

On the other hand, maybe your simulator doesn't do that. The 2 casualties per 3 defenders sounds like a straight multiplication of 6 x 36%. I'll assume for the moment, though, that it takes into account partial damage from failed attacks.

You inflict twice the casualties on the enemy and take half for yourself! So where is the increased cost to attack?

You're leaving out the cost of building all those H'wacha. Odds of actually inflicting 1 HP damage are 26%, at 12:10.5 odds, and taking into account the 50% miss rate. Sure, you need only 3 cavalry, but you need 22 H'wacha to bring down 3 defenders to 1 HP each!

One way to look at it is that H'wacha are an "investment". 22 H'wacha (cost 880 shields) saves you the loss of 1 Cavalry (80 shields) per town, and the cost of building 3 Cavalry (240 shields) more to take the town. Net cost is 640 shields, so you're getting 12% return on your shield investment per town taken.

Of course, there's also the opportunity cost of taking 3 times as long to take each city. That translates to increased AI production, and increased unhappiness from War Weariness.

Counterattack? Let's say they have cav (probably longbows, but we'll give them a strong counterattack).

Yeah, the AI is longbow happy, where I never build them, but why would you assume "probably" longbows? At Emperor and above, Cavalry counterattacks are common.

With Nationalism comes drafting. The AI is *totally* draft happy. If you come within two squares of his city he will draft it down mercilessly. So bad you often have little choice but to raze it or wait until Modern era for unhappiness to wear off (well, starving down to 1 and keeping it there is fine) You wouldn't notice this though, strictly using cav or MA attacks, as you're often 3 squares away when you start the round of attack.

So this is another drawback to a slow bombardment attack.

Yes, I've noticed the AI likes to Draft, but generally that only affects me with tanks.

I've never really noticed the unhappiness you mention, because I don't really care about cities I take during the early Industrial era and later. They're worthless due to corruption anyway.

- Gus

[Arrian]

I'll trade higher casualties for cutting the time it takes to conquer an empire by 1/2 or more.

The last major Cavalry vs. Rifleman war I fought was on Emperor (1.29) as Egypt vs. Japan. I did have some artillery (mabye 8 of them), which I used to bombard any counterattackers that showed up, along with their original attack force. That helped: bombard works pretty well in the open, and like I've said before, Samurai do bad things to Cavalry... unless they only have 1-2hp left. Once I cleared them out of the way, I simply massed my Cav and picked off one city at a time. No bombardment.

One more thing about casualties: I don't necessarily mind taking moderate to heavy casualties with my Cavalry. I usually have a fairly large standing army of knights which gets upgraded, and not too many Cav are actually built from scratch. Cav are a dead-end unit. Those that don't get killed off or become elite are going to end up being disbanded as I approach Tanks anyway.

-Arrian

[Charis]

Response

Nice response Gus...

it's not the late medieval, Cavalry and Riflemen that are frustrating. It's the early game where you're locked out of trading, and by the time you're ready for a horseman rush everyone else is well into Feudalism and the like... So do you play Diety all the time? If so, what is your early game like? Do you go for swordsmen, spearmen, and catapults?

Very true. I play about half-and-half Emp/Deity, wishing there was something in between. Emp is more 'fun' but too easy, deity - as you noted - often more just an increase in frustration not just challenge. I most commonly have a quiet opening, but if playing aggressive go for horsemen (if my UU is horse-based) or swords (otherwise). And no, I've never built a catapult on deity, or emperor for that matter.
Just too many other things to build at that time.
Where I usually make a ton of them, if I'm planning to make heavy use of bombardment, is from corrupt towns where I don't want the expense or delay of barracks (since it doesn't matter for cannon) - cats are very cheap and even 100% corrupt towns can do a decent job making them. The same cash that 'usually' gets used for a big knight or cav upgrade is used to upgrade them to cannon, or later, with Wall Street online and the economy flowing, to artillery. My most common 'large-stack' use is, like others, artillery due to its range of 2.

I'm about to start a Korean deity game, military emphasis, where I do take the atypical approach of a decent amount of cats.

Could you point me to the combat simulator? I've done stuff with spreadsheets, but it would be nice to see one that took all factors like automatic promotions into account. It's too tedious to do more than the initial odds by hand.

It's a VB program I'm writing. It currently accounts for partial damage, retreats, different simulations for blitz attack more (how many do you need to take a town in one turn) and 'siege' (tracking hp's between rounds). It was just meant for my own toying around, but if that would be of use to others I could try to polish it up over the holidays and release it (?) Next thing to add is promotions.

Odds of actually inflicting 1 HP damage are 26%, at 12:10.5 odds, and taking into account the 50% miss rate. Sure, you need only 3 cavalry, but you need 22 H'wacha to bring down 3 defenders to 1 HP each!

Good point, that is a lot. Actually, that sounds like a real good addition to the simulater, as I don't have a good 'quantitative' handle on how many artillery pieces are needed.

Regarding the longbows, I must have in mind several recent games where my main wars all started with resource denial on turn one. Cut off the horses, absorb whatever initial cav they can muster, and that's all you see. Yet longbows with no resource requirements kept coming, and coming. It was particuarly painful in a game with the Mongols, as the lowly longbows just chewed up their meager 2-defense hides!

Thanks for the response,
Charis

[GusSmed]

Re: Response

Originally posted by Charis
I most commonly have a quiet opening

So how do you avoid being a complete backwater when you use a quiet opening? I just fiddled around with a Korean Deity game, in which I started on the same land mass as the Persians. By the time I had 5 cities (closely packed, just 3 tiles apart), they had 14, and I had nowhere left to expand. My choices were either fight or stagnate.

This was made worse by the fact all the AI civs were now in contact. The AI production / research advantage would be bad enough, but they actively share technology. I think that's why there's such a sharp break between Emperor and Deity - if you don't get a tech lead, they trade with themselves to the point where they're a full era ahead of you. Your research disadvantage is 7:1 or more, not just 1.5:1 or whatever the bonus for a single civ is.

You can buy tech at a discount once everyone has it, but it's not enough to make up for this.

As an experiment, I tried amassing a catapult / swordsman force. I figured the minimum to take a city was 10 catapults + 5 swordsmen. This took until 270 BC to build, and when I did attack, I encountered Knights and Musketeers.

I knew everyone was in the mid-Medieval period because both Copernicus and J.S. Bach's where under construction.

Actually, that sounds like a real good addition to the simulater, as I don't have a good 'quantitative' handle on how many artillery pieces are needed.

If you do this, see if you can't get some hard numbers on the retreat percentage and the artillery miss percentage against towns. From experience, I believe it's 50%, but I haven't seen that verified anywhere, and I haven't had the patience to test it empirically.

Regarding the longbows, I must have in mind several recent games where my main wars all started with resource denial on turn one. Cut off the horses, absorb whatever initial cav they can muster, and that's all you see.

I find resource denial is often too difficult. I look for opportunities, but often the source is 3-4 cities in.

- Gus