What's more efficient, upgrade your units, or make new ones and replace them?

[Dis]

does anyone have a table showing all the unit upgrade costs?

it's weird how some cost so little, but some cost so much.

[Modo44]

I have no table. It is a set multiplier of the Shield difference, one you can change in the editor. In C3C it is 3. Basically 30 Gold for every 10 Shields.

[Terra Nullius]

Originally posted by Modo44
Why do you say Knights are useless?

i like knights. i just don't think they should be used in battle.

in the early middle ages, i'm sure to have a few warriors, swords or archers hanging about. these can be quickly and cheaply upgraded and they have the same combat value as a knight.
pike + med inf/longbow = 4.3.1
knight = 4.3.2
shield cost: 70 (same)
ok, the extra movement is nice, but it doesn't give me a crucial advantage.

second, archers and swords are reaching the end of their useful cycle. if i'm going to use them, it has to be now.

third, spears and pikes are reaching the end of an important cycle. so there is often a good reason to throw them into battle, then slowly build muskets to replace them.

fourth, foot troops keep pace with the artillery. this means that med inf/longbows can actually take a lot of targets with lower casualties than knights would.

fifth, it takes a long time (or a lot of cash) to build a decent squadron of knights. by the time i've got them built/upgraded, i'm half-way to the industrial. i'd rather hold on to them for the ludicrously cheap cavalry upgrade.

sixth, it's mostly about managing unit costs. the middle ages is the perfect time to switch from having lots of cheap, versatile troops to having expensive, specialised troops. this is a two-stage (but concurrent) process.
i. sacrifice/ disband the old
ii. build the new troops and wait for upgrades
because knights and muskets are so costly, it actually gives you an advantage. you build troops more slowly, so you have fewer of them. hence lower unit support.
at the same time, your old troops are only costing you money, so they might as well go hurt someone else instead.

basically, it's about looking for the best way to get some leverage. why spend 70 shields building a knight/musket, then throw him into battle and lose? i say, wait a few turns more and, for a mere 30 gold. you can have a 50% improvement in combat stats.

i wouldn't harp on about this, it's just that there is such a clear distinction between the value of the different upgrades.
to me it seems a crime to use knights/muskets in battle given that
(a) alternative, equally-powerful, cheaper troops are available
(b) the upgrade is extremely cost-effective
(c) there is a way to work in an advantage (lower overall troops numbers as aa troops die off and the replacements arrive more gradually)

I can say the same about Muskets ... if I have cities in peril, I will definately want them to have Muskets for defenders

sometimes, yes (although i'd still prefer two pikes versus one musket)
there will certainly be exceptions. the difference is, if my pikes can do the job (even with heavy casualties), they get the mission. muskets are intended to guard the castle (core) and look pretty until they can be turned into something worthwhile (rifleman).
it just isn't efficient to spend so many shields on a single unit, then lose it the next turn to a javelin thrower.

Originally posted by joncnunn
Cat -> Treb -> Cannon -> Art. A No brainer to upgrade

specifically, no
cat -> treb -> cannon = yes
cannon -> artillery = no

a 10 shield upgrade is almost always worth it
a 20 shield upgrade is normally worth doing
a 30 shield upgrade is sometimes worth doing
a 40 shield upgrade is hardly ever worth it
this is my cut-off point.

i hate having to disband my cannons, but the fact is, it usually takes more time and resources to upgrade cannon than to simply build new arties.

Archer -> Longbow. Only seems worthwhile if you don't have horses to me. (Or instead if your the Vickings and therefore able to upgrade to berskers.)

Horsemen -> Knight. An extremely worthwhile upgrade. Attack of Longbow + Def of Pike + a second MP.

strongly disagree

horseman -> knight is one that i have simply taken off my list. to such an extent that i hardly ever even build horsemen these days. it just doesn't add up for the investment.

i. requires 'horseback riding' - not a useful tech. by the time i have this tech i am about two clicks away from the middle ages.

ii. requires horses - ok, if they fall into my hands i won't complain. and by the time 'military tradition' comes around i must have horses. but i'm not going to distort my rex for the sake of a fast, 30-shield archer.

iii. requires 'chivalry' - this really bugs me. i hate civ's with knight replacement uu's for this very reason. the ai won't sell this tech easily. so i have to spend all that extra time researching an unneccessary technology.
this could be due to my own inexperience, but nothing ruins my middle ages timetable like having to divert time to researching chivalry.

iv. requires 'chivalry' - par contre, 'feudalism' gives - same attack strength, same defence strength, and a government tech.

v. i don't spend gold on a unit unless it will be used very soon. so i spend my shields on building knights, then spend gold to get cavalry and go out an' whoop me some! i will only use knights in battle when i have no other choice, so it would just be a waste of cash to upgrade horses.

vi. there's really no need to have iron and horses in the ancient age. one or the other will do the job. i think the iron-based units are simply better value for money.

vii. -> berserkers
150 gold pieces
you're actually prepared to spend that much on one unit at such an early point in the game!!!
(ok, berserkers can do some damage, but that is just obscene).

vii. archer -> longbow
archers can be extremely useful in the aa. it is situational. so i don't go after this every time. but if the opportunity is there, i will certainly use it.

overall, this tiny period (aa - middle changeover) is crucial to the game and has a whole bunch of alternatives in terms of tech, resources and upgrade paths. a scientific civ will want a completely different approach from a militaristic one, and so on.

of course each situation calls for a unique approach, but my rule of thumb lately has been pretty clear - any upgrade over 40 shields is just not worth it.

[Dominae]

Upgrading is always more efficient than building up a new army from scratch. Gold (Commerce) is much more plentiful than Shields. For example:

1. All tiles produce at least 1 Commerce if you put a Road on them; some (Flood Plains, Coast, etc.) produce no Shields. In Republic, all tiles produce at least 2 Commerce (with a Road).

2. The first Shield-boosting improvement (Factory) is available in the Industrial era; the first Commerce-boosting improvements (Library, Marketplace), in the Ancient era.

3. You can trade for Commerce in the form of techs or Gold with AI, but you cannot in any way get Shields from other civs. Trading one tech around will sometimes net you 500+ Gold; there is no analogous windfall for Shields (well, Great Leaders, but they're random).

Upgrading is also faster than building from scratch. At 10 Shields per turn it takes 3 turns to build 3 Warriors, or 3 turns to build 1 Swordsman. But, with enough Gold (180 to be exact), those 3 Warriors become 3 Swordsmen, and 3>1.

Furthermore, with upgrades, you can start building your army before you obtain the relevant tech or Strategic resource; for instance, building up Horsemen long before Military Tradition or even Gunpowder. Fielding 50 Cavalry the turn you discover MT is something the AI has a tough time dealing with.

The problem is that upgrading costs Gold (Commerce), which you also need for research. So the key question you need to be able to answer is: when can I afford to upgrade my troops at the expense of slowing down my rate of tech advancement?

A related question is: when will I strike? A mass-upgrade scheme requires planning to execute: you need the right level of technology relative to your opponent, an appropriate amount of Gold, enough cities to keep the unit upkeep down, etc.

The two best mass-upgrade schemes in the game are: Warrior->Swordsmen, and Horsemen->Cavalry. Usually one or the other is enough to gain a significant lead in any given game.

The reason for skipping Knights is that Knights almost always go up against Pikemen, while a Cavalry beeline sometimes catches civs with Pikemen to defend instead of Musketmen. If you're certain that your Knights will be facing Spearmen exclusively, then by all means research Chivalry. But also note that Cavalry does not require researching Theology or Chivalry.

Mass-upgrade schemes do not always work, of course, usually because the tech pace is too fast. In this case a mix of upgrading and building units from scratch is necessary. I guess a rule of the thumb therefore is: upgrade when feasible, taking care not to fall behind technologically (or fall out of the tech trading loop).

[vmxa1]

That is about as good of an answer as one could hope to get.

[Dis]

gold wasn't more plentiful than shields in my game . I was industrious too, so that helps in getting all my improvements built. And I had nothing left to build.

[Terra Nullius]

Originally posted by Dominae
Upgrading is always more efficient than building up a new army from scratch.

wha the ?!!

hey! get back to your beta testing and leave my lame-o posts in peace!

Originally posted by vmxa1
That is about as good of an answer as one could hope to get.

yeah, now tell me something new.

oh well, here we go again

dwr dwr dwr

[Russian_King]

ok in general:

if u have tons of gold: upgrade the damn things, u get the speed and if u have **** production it wont make a hole in ur building

if u have tons of shields, produce the new units, and move the old ones into cities with really bad production to help them build other stuff

if u have both, well i guess its up to u... get the cities that are maxed out to produce the units and the rest of em can produce things like nukes, which u can NEVER have enough of.. even if u have 5 per city

i personally like to build them cos from the early stages until the modern ages my economy is usually not as good as my production. im an industrialist. i keep my economy JUST good enough so i can produce about the same amount of science as rival civs at a high rate and only have minimal coin go to my treasury.

[realpolitic]

I do believe in using units aggressively before ASAP to circumvent the problem. When I'm upgrading, I still have scientists on the very currupt cities. Even if my science slider is at 0%, the scientists keep advences at 20% or so of normal. Then I can move science to it's normal rate and beat up my opponents. just upgraded 150 mounted warriors to cavalry this way and now am brutalizing my opponents. During the upgrades, I've been building marketplaces and banks, and after I completed this, I've built libraries and Universities. Although I eventaully had to make peace, my upgrades have kept the AI at bay until I was able to get back to unit production. After my first war, after the Rennanasance builds, I'll be able low enough in units/city to switch to Republic, and use it indefinately as a wartime government.

[UnOrthOdOx]

Originally posted by lebensraum
vii. -> berserkers
150 gold pieces
you're actually prepared to spend that much on one unit at such an early point in the game!!!
(ok, berserkers can do some damage, but that is just obscene).

You better believe I am prepared to spend that as the vikings!

My hell, what other unit has such a LONG lifespan!? It starts in the early Medieval, and is on par with all offensive units through mid industrial, not to mention amphibious.

Use of berserkers should be as use of cavalry: the game ending move in the hands of a human vs AI.

[Terra Nullius]

ummm, yes.

let me re-phrase that... berkos make for a very expensive upgrade if you're not a homicidal maniac trying to produce an entire star-fleet of the little bearded buggers.

i'd also point out that producing nothing but archers and then researching at zero for a thousand years might have different effects when you're not on an isolated land mass.

the other thing is, maybe i should change the way i do things, but:
i tend to aim at producing the bare minimum of units. i'll build what i need to get me through the war, then allow 'natural' attrition to thin the ranks. if things go to plan, by the end of a war, i'll have just defenders and a handful of elites left over.

typically, my wars involve taking out 3 or 4 ai cities and throwing their economy off balance. then i use the reduced unit numbers to allow fast research to the next offensive unit. i was forgetting that that just happens to be what works for me.

and in the spirit of accepting different ways of looking at things,..

If you're not a part of the solution, there is good money to be made in prolonging the problem.

where do i sign?

[joncnunn]

If your not interested in conquering the world with Berserkers, why are you playing the Vikings?

That's like playing the Germans and not using Panzers or playing the Ottomans and not using Saphi.

[Terra Nullius]

random games

i'm mostly interested in conquering the world (or at least getting rich by helping someone else do it). if the berkos want to come along for the ride, tha's fine bah me.

berkos, sipahi and panzers are all uu's that just don't impress me. berks are too costly, sipahi's are ok, i guess and panzers are just way too late.

i've never seen the need for a tank with extra movement - you've already got railroads by then, throw in a few combat settlers and there's just no need for panzers.

still, every map is different. it depends,..

[Addled Platypus]

Originally posted by lebensraum

throw in a few combat settlers and there's just no need for panzers.

just when I thought i heard it all

combat settlers?

I quess there is some Noob in me still

[Terra Nullius]

you've never seen jon step out of a covered wagon with his m-60 at the ready?