What's a good army composition?

[Blitzzer]

This is it - the army composition to beat them all. I love it. Maximum firepower, ladies and gentlemen. (Pst! I'm talking about Portugal's army. I don't know what this player fool is trying to accomplish with his; infantry? who needs cannon-fodder when you can have cannons!)

 

[TenshiN]

In all my games, it usually all comes down to 15/6/6 stacks, especially after 1.8, since the losses seem to replenish much slower than they used to. And i always have a few cannon-heavy squads just for the quicker sieges of some provinces i might need. Works like a charm!

 

[zephyr7913]

5/3 is not bad but you are severely lacking guns.

Much to learn.. much to learn..more guns means more effective sieges too. I think I'll love that. And 2/1/2 is so perfectly splitable than my 5/3/2. Way to go, 2/1/2!

 

[AlphaSonic]

I like 6/2/8

 

[Gustav91]

Techgroup
Techlevel
Money
Manpower

That decides the compostion.

 

[hitchens]

When fighting large wars I always end up merging armies so I often end up with stacks that never were planned. At one point its all about defeating the enemy doomstack and merging armies.

 

[FrosT37]

Do you want an ideal or practical army composition?

The ideal composition looks something like:
Infantry = full combat width minus cavalry
Cavalry = 2 to 6 (depends of tech level and tech group)
Canons = full combat width

The ideal composition is not practical because you have to take into account supply limit, maintenance cost, tech level, etc.

A more practical composition is to insure you do not have insufficient support (depends of your tech group), you do not have more canons than infantry+cavalry, your stacks can move on most of your provinces without suffering attrition, and you can actually pay for the maintenance of your units.

Follow-up to my previous post.

In practice, I like to to have full stacks (first row + back row) equal to my combat width because it usually fits the supply limit in most of my provinces.

I also build armies around multiples of four, not only for the total army size but also for the number of infantry, cavalry and canons in each army). The reason why is because I'm lazy and I use the split button A LOT, especially when I begin carpet sieging. This allows me to split an army four times and still have the same amount of each unit type in each splitted stack.

Cannons are a specific case. I try to have at least 4 cannons per stack at the beginning so that when I split my stacks for carpet sieging, I can have the sieging bonus in all four provinces.

As soon as I reach 32 combat width, I increase the number of cannons to 16, so the stacks looks something like : 12 infantry, 4 cavalry, 16 cannons. When I split this stack in four armies, I have 4 cannons in each, allowing for a +2 bonus to siege in each province (with small forts).

So in practice (assuming Western tech group), my stacks looks something like this

16 infantry 4 cavalry (CW 20)
16 infantry 4 cavalry 4 cannons (CW 24)
20 infantry 4 cavalry 4 cannons (CW 28)
12 infantry 4 cavalry 16 cannons (CW 32)
16 infantry 4 cavalry 16 cannons (CW 36)
16 infantry 4 cavalry 20 cannons (CW 40)

 

[Morik]

I keep reading that it isn't effective to have more than 6 cav in a stack. Why is that?
Is this due to how deployment works, where only 6 cav get deployed if you have enough infantry to fill the rest of the width?
What about reinforcements for cav that dies?

If I'm mega-rich Ottomans, for instance, is it still the most effective comp per force limit to have stacks like 16/6/14?
8/4/8? Etc?

 

[DevastatingTech]

Changes for each technology group.

10/6/10 for western.

 

[nicechinos]

I keep reading that it isn't effective to have more than 6 cav in a stack. Why is that?
Is this due to how deployment works, where only 6 cav get deployed if you have enough infantry to fill the rest of the width?
What about reinforcements for cav that dies?

If I'm mega-rich Ottomans, for instance, is it still the most effective comp per force limit to have stacks like 16/6/14?
8/4/8? Etc?

It's only valid for early game when maneuver of cavalry is limited to 3. 3 cavalry on both sides guarantee effective flanking. As military tech progresses maneuver increases up to +150% from default which requires more cavalry for effective flaking. Besides I really love cavalry for its shock multiplier. I play mostly as Eastern tech.

 

[FrosT37]

I keep reading that it isn't effective to have more than 6 cav in a stack. Why is that?
Is this due to how deployment works, where only 6 cav get deployed if you have enough infantry to fill the rest of the width?
What about reinforcements for cav that dies?

If I'm mega-rich Ottomans, for instance, is it still the most effective comp per force limit to have stacks like 16/6/14?
8/4/8? Etc?

1. Cavalry is costly
2. You want to avoid insufficient support at all cost, even after a fight where you infantry took a lot of casualties
3. Fire phase becomes more important in the end game

The exception to this is the Eastern tech group and the Hordes.

 

[artemis667]

Morik, I think it's just that cavalry aren't very effective for their cost. You can use them to flank along the line, and that helps to a degree, but you wouldn't build a line with them, unless you were a nomad in early game or something. Whereas artillery sit behind your line, and pound away. They're expensive, but you don't have to replenish them (unless a battle goes really badly) and they'll really tip the balance of battles in your favour - they'll tear the enemy lines to your shreds, minimising the losses to your own troops in the process.

 

[nouli]

I keep reading that it isn't effective to have more than 6 cav in a stack. Why is that?
Is this due to how deployment works, where only 6 cav get deployed if you have enough infantry to fill the rest of the width?
What about reinforcements for cav that dies?

If I'm mega-rich Ottomans, for instance, is it still the most effective comp per force limit to have stacks like 16/6/14?
8/4/8? Etc?

Your statement about deployment is wrong (unless something changed in 1.8). If it's possible 50% of front line will be filled with cavalry.
The reason why people use specifically 2/4/6 cavalry is because of how maneuver and flanking works (these numbers allow for maximum flanking benefit).
Other than that, cav is simply bad. You don't get enough value for the price. But even ignoring price, inf usually is better (read this thread if you want details).

 

[Morik]

Ah I was basing my deployment statement on the wiki, which must be outdated. http://www.eu4wiki.com/Land_warfare#Unit_deployment
Specifically: If there is enough infantry to fill the entire first row:

Deploy all infantry in first row, except the outer positions in the amount of cavalry available but not more than 6[[this amount is made up]] (3 left and 3 right).
Deploy up to 6 cavalry in first row at the sides of the infantry.
Deploy all artillery in the second row.
If space left in the second row, deploy all remaining infantry in the second row besides the artillery, but leave the outer positions in the amount of cavalry available but not more than 6[[this amount is made up]] (3 left and 3 right).
Deploy all remaining cavalry in the available space in the second row, beginning with the positions behind the cavalry in the first row and then going inwards to the infantry.


EDIT: I looked over that thread (briefly, first & last page) and didn't see any commentary that cav is worse per regiment than inf, just that it is less economical per ducat.

What I'm asking is whether, per regiment, cavalry are superior to infantry (so long as you don't have insufficient support).
E.g., with combat width 30 and all modifiers equal, which army wins?

24/16/30

34/6/30

(Does the answer change depending on tech level? If so, what's the inflection point?
If the answer is the inf heavy one wins because it causes enough casualties to lead to insufficient support, double the # of inf & cav in each army til that no longer happens.)

And nouli--it sounds like you are saying that for deployment, it would be 15 infantry and 15 cavalry deployed in the front line of the first army, not 24 + 6?


EDIT 2: I read more of the thread, I see some tests in there .

 

[nouli]

And nouli--it sounds like you are saying that for deployment, it would be 15 infantry and 15 cavalry deployed in the front line of the first army, not 24 + 6?

Yes. (50% is hard limit, doesn't depend on your tech group cav limit)

EDIT 2: I read more of the thread, I see some tests in there .

I'm pretty sure that's the only hard data that we have. In theory you could calculate some stuff, but it pretty much means writing combat simulator (and figuring out how exactly new deployment algorithm works).

 

[Osman Pasha]

What does cavalry do, its so useless late game. Anyone answer me!

 

[artemis667]

What does cavalry do, its so useless late game. Anyone answer me!

OK this is a little vague, because I don't understand everything about the game. But it works a little like this.

The key is about numerical engagement, regiment on regiment. If you have cavalry superiority, you'll be able to flank the enemy infantry, meaning that while they're engaging your infantry in line combat, you have an extra regiment hitting them from the sides. So you're attacking one regiment with two. Doing this drops the enemy regiment strength, and when regiments are understrength fighting against near-full strength regiments, they fail and rout very quickly.

Artillery is, in a way, another sort of flanking. You're pounding the enemy line while infantry are engaging. And you can have almost a whole line of artillery behind, making your fire phases very destructive. In practice I don't think most players would use the whole line - it helps to keep a bit of a reserve of infantry just so if your line weakens you have reserves to stop your artillery getting engaged directly. But if you're 'theorycrafting', yeah, 50% artillery composition is great. Good luck micromanaging for the different battle widths and attrition in every province. In the real world, I wouldn't have more than, say, 30% artillery, super effective but flexible.

There are variations too - if you're playing with high manpower but less money, you definitely want more infantry and less artillery. If you're playing Poland or any other country with elite cavalry, you want enough cavalry to flank at full strength and probably a reserve as well.

 

[nicechinos]

OK this is a little vague, because I don't understand everything about the game. But it works a little like this.

The key is about numerical engagement, regiment on regiment. If you have cavalry superiority, you'll be able to flank the enemy infantry, meaning that while they're engaging your infantry in line combat, you have an extra regiment hitting them from the sides. So you're attacking one regiment with two. Doing this drops the enemy regiment strength, and when regiments are understrength fighting against near-full strength regiments, they fail and rout very quickly.

Artillery is, in a way, another sort of flanking. You're pounding the enemy line while infantry are engaging. And you can have almost a whole line of artillery behind, making your fire phases very destructive. In practice I don't think most players would use the whole line - it helps to keep a bit of a reserve of infantry just so if your line weakens you have reserves to stop your artillery getting engaged directly. But if you're 'theorycrafting', yeah, 50% artillery composition is great. Good luck micromanaging for the different battle widths and attrition in every province. In the real world, I wouldn't have more than, say, 30% artillery, super effective but flexible.

There are variations too - if you're playing with high manpower but less money, you definitely want more infantry and less artillery. If you're playing Poland or any other country with elite cavalry, you want enough cavalry to flank at full strength and probably a reserve as well.

That's closer to the truth. The misconception that you need to build a full battle width of infantry + 6 cav doesn't take into account maneuver and the enemy's line. If enemy has twice narrower line most of the troops will be in reserve because they won't be able to engage the enemy due to low maneuver.
Cavalry with a maneuver of 7 (late game) is deadly as hell. It has shock multiplier of 5 vs infantry's 2.15. Even with same shock pips cav would do more twice as damage in shock phase. It would also attack 7 (!) enemies to its right and left. The regiment which faces 7 cav (much less would do that) against it in a shock phase is wiped after the first day of shock phase. The opponent melts away from cavalry late game.

 

[nicechinos]

What does cavalry do, its so useless late game. Anyone answer me!

It's nowhere near useless. The misconception comes from the fact that people don't pay attention to increased maneuver and shock/fire multipliers.

 

[Vedinu]

I don't know if it's good but I usually go with 5-2-3 for a nice 10k army