Is Barrage broken?

[Ibian]

How about using the trebs too? Better to lose a treb than a knight at least as far as hammer cost goes, and they should have good chances after the cats are done. Also, id use a little more siege and a little less normal troops if it was me.

I think cats can get shock too? Not sure, its rare i actually fight in the field but if so thats the proper promo to compare.

 

[CivCorpse]

How about using the trebs too? Better to lose a treb than a knight at least as far as hammer cost goes, and they should have good chances after the cats are done. Also, id use a little more siege and a little less normal troops if it was me.

I really try to avoid using trebs in the open field. In the scenario i used it was a stack coming to attack my city. I rarely keep trebs in the homeland for defense. Only if they have been recently produced and just happen to be in the area. However I do try and keep stack of 4-5 cats within 3 turns march of my border cities just in case. 3 turns is about what you need if you have a full BFC since their siege moves slowly.
As for the trebs having decent chances, they would get crushed. 2/3 of the time all the cats died and there were still a good number of troops with stregnths in the 6-7's. Those numbers are easy pickings for knights then maces, but they would smash the trebs. And trebs are 80 hammers. They are not reliable enough for collateral in the open field to risk losing one and still lose a night. Depends on what i have available at the time.
But regardless of if it's a treb or a cat...it gets barrage2 before going into the mix.

 

[Fezzik]

As soon as the AI has large stacks of longbows, and especially if they are CG ones, I wait until I have trebuchets and maybe macemen to crack the best defenders. So I would be interested in seeing your figures for trebuchets.

I only did one trial with Trebs. The first CR2 Treb had 55.2% chance of living. There's no way any combination of Barrage 2 and CR2 can improve on that in this test.

1) CR2 55.2%  B2 21.1%
2) CR2 72.1%  B2 21.1%
3) CR2 86.7%  B2 46.2%
4) CR2 93.7%  B2 69.2%
5) CR2 99.3% CR2 99.5%
6) CR2 99.5% CR2 99.5%

So you lose 3 Trebs instead of 0. I stopped testing there...

 

[CivCorpse]

How about using the trebs too? Better to lose a treb than a knight at least as far as hammer cost goes, and they should have good chances after the cats are done. Also, id use a little more siege and a little less normal troops if it was me.

I think cats can get shock too? Not sure, its rare i actually fight in the field but if so thats the proper promo to compare.

cats can get shock. but it is not the promo to compare. Neither is Drill. You could argue that giving them drill2 increases their survival and guarantees at least causing a little damage from1st strikes.

Use barrage. It scales nicely. If you have a unit get barrage 3 (izzy, or charasmatic) barrage 3 is like 1+2 combined. you really shred through some troops.

 

[Ibian]

You wanted to get rid of the pikes. Shock does that better than C2.

 

[CivCorpse]

You wanted to get rid of the pikes. Shock does that better than C2.

But you end up delivering less collateral damage. IIRC collateral damage is a reflection of the damage you do to the opponent you fight directly. That means the drill2 x-bows rotate to you rather than the maces, And now you have neither the barrage 2 for extra collateral or c2 for extra base damage

Barrage is the way to go in the open field

 

[Fezzik]

But you end up delivering less collateral damage. IIRC collateral damage is a reflection of the damage you do to the opponent you fight directly.

It's independent. i.e vs. longbows, it is about 0.5 damage per defender per attack, regardless of whether the primary defender is left completely undamaged or damaged severely to the 4.5 maximum damage. No matter how the primary battle goes, collateral just chips away at everyone else slowly and separately.

EDIT: also, the collateral damage reduction of the Drill line for defenders applies only to the receiver of collateral damage, it does not transfer from the primary defender's promotions to the rest of the stack.

CR2 Cat vs.
CG3 Drill4 LB (primary defender)
CG2 LB 5.5
CG2 LB 5.5
CG2 LB 5.5
CG2 LB 5.5
CG2 LB 5.5

CG3 LB (primary defender)
CG2 Drill4 LB 5.8
CG2 Drill4 LB 5.8
CG2 Drill4 LB 5.8
CG2 Drill4 LB 5.8
CG2 Drill4 LB 5.8

 

[DMOC]

I personally never promote my siege up the barrage line--it's always city raider for me.

 

[Ibian]

If its true that collateral damage is based on the fight against the defender then barrage just might be useful against mixed stacks. Anyone who can confirm this?

 

[VoiceOfUnreason]

IIRC collateral damage is a reflection of the damage you do to the opponent you fight directly.

I've never found anything in the SDK that would justify this claim. Collateral combat is calculated and delivered before the combat resolution begins, and the damage is recalculated for each unit that gets caught in the shrapnel. The calculation uses baseCombatStr, and as such is independent of current hitpoints, attacking / defensive bonuses, etc.

 

[VoiceOfUnreason]

I think barrage 2 versus catapults should be 11 damage. Civ 4 keeps damage out of 100, it's just that due to how it displays unit hp, it looks the same as barrage 1.

Judging from your chart, you meant B II vs longbows?

Trying the math by hand, I get
10 * ( 3 * 7 + 6 ) / ( 3 * 6 + 7)
= 10 (27) / 25
= 270 / 25
= 10.8
= 10.

Keep in mind that the reason that B I alone is absolutely worthless for Trebs is that 4 * 1.2 = 4.

 

[VoiceOfUnreason]

Combat2 average pike strengths are 5.25 for the healthiest and 4.59 for the weakest.

Can you be clearer here about how you are measuring this strength? There's a different thing that looks odd to me, but when I tried to reproduce it using a slightly different test, my numbers ended up being very different from yours. I understand your version of the experiment (I believe), but the "two digits of precision" numbers that I find look nothing like yours.

 

[vicawoo]

where'd you get 7? it's +20% +30%

 

[OTAKUjbski]

where'd you get 7? it's +20% +30%

Barrage I (20%) + Barrage II (30%) = +50%

5 * 1.5 = 7.5 ... floored to ...

7.

 

[OTAKUjbski]

I think barrage 2 versus catapults should be 11 damage. Civ 4 keeps damage out of 100, it's just that due to how it displays unit hp, it looks the same as barrage 1.

It's also worth mentioning ...

( 90 / 100 ) * 6.0 = 5.4
( 89 / 100 ) * 6.0 = 5.34

Civ4 would display them as 5.4 and 5.3 respectively.

 

[vicawoo]

Why would you floor so early? If you look at combat odds, its preserved to tenths (or maybe hundredths), so it's unlikely they apply a floor function before entering them into the formulas. The floor function occurs at the damage calculation, not before. Also, regardless of the strength shown, hp is retained to hundredths.

 

[Roxlimn]

The question here is attacking a mixed stack with enough defensive Catapults in order to weaken Pikemen enough to render the stack completely useless by then attacking it with enough Knights.


Personally I think it's kinda weird to be doing that. If the stack only has two Pikemen, your best course of action is to use Shock Knights to attack the stack. Your Knights may or may not die (they have retreat odds, too), but they WILL draw the Pikemen up to defend and they WILL most likely weaken them to the point where they'll be susceptible to further Knight attack.

The next unit that'll probably be up for stack defense are the Knights themselves, and in that case, it'll be nice to have Formation Knights or even better, Shock Elephants to draw up the Knights up the defensive priority in order to kill them at best odds.

At that point, the stack will have little to no stack defense (since it's carrying an abnormal amount of CR units) and it'll be trivial to mop it up with Knights or whatever. Past a certain point, the siege will die en masse, but really, once the stack defenses are done, the stack is more a walking pile of XP than a real threat.

I find that messing with a stack's defense priority is a more cost-effective manner of killing it than sacrificing large numbers of siege units. The only time the reverse is true is when you're up against a unit that can't be defeated using defense manipulation, such as a Protective Fortified Longbow on a hill, at which point it's best to just level the best CR siege against it and hope the collateral damages enough of its buddies to render defense manipulation viable again.

 

[CivCorpse]

Can you be clearer here about how you are measuring this strength? There's a different thing that looks odd to me, but when I tried to reproduce it using a slightly different test, my numbers ended up being very different from yours. I understand your version of the experiment (I believe), but the "two digits of precision" numbers that I find look nothing like yours.

After I attacked each stack with a stack of catapults. I hovered my mouse ove the stack and wrote down the remainging strength of the pike men. i added all the stregths of the strongest of each pair and divided by 20 to obtain an average..

 

[OTAKUjbski]

Why would you floor so early?

Because that's how the code does it ... from CvUnit::collateralCombat:

int iCollateralStrength;

...

iCollateralStrength = ((((getDomainType() == DOMAIN_AIR) ? airBaseCombatStr() : baseCombatStr()) * collateralDamage()) / 100);

If the developers would drop the silly "/ 100" and reserve that until the end of the formula, then Barrage wouldn't be losing as much damage.

 

[CivCorpse]

I've never found anything in the SDK that would justify this claim. Collateral combat is calculated and delivered before the combat resolution begins, and the damage is recalculated for each unit that gets caught in the shrapnel. The calculation uses baseCombatStr, and as such is independent of current hitpoints, attacking / defensive bonuses, etc.

Thank You. I have never been entirely sure exactly how collateral damage was figured. i did know that barrage increases it somewhat, but never sure what the base amount was. So basically anything besides barrage mearly affects the siege units ability to survive the battle.
I now consider myself enlightened.

So the question becomes, is the siege unit's life more valuable than the damage I need done?