I have a specific question regarding siege weapons, but I also had hoped to spark a discussion on what is really the best promotions for them.

First, my question: After having been engaged in a few long wars, one of my city raider cannons had lived long enough to get a 4th promotion. It already had CR3 at this point. So, for a cannon used to soften city defenders, which would be better, drill 1 or barrage? I think drill would help it survive longer, but barrage might do more damage.

Usually I have one group of CR siege with my main city taking army, and a few barrage siege to handle enemy stacks (or more if I haven't taken out the AI's main SoD yet). I hardly ever use accuracy, as I prefer to have all my siege available to attack and therefore expendable. Maybe someone here could convince me to ;). Finally, does anyone ever use drill for siege weapons?

Barrage and Accuracy are great for earlier wars when the AI has put up walls and castles everywhere and it would take a ridiculous amount of siege and/or turns to bring cultural down to 0. If you're short of EP and don't want to take away from your science slider, I suggest using some barrage/accuracy siege for a quick defense reduction.

Also, I've used both drill and barrage after getting siege to CRIII. Drill is really only necessary if you plan on further warring at tech parity (IE cannons vs. rifles instead of cannons vs. longbows/grenadiers/muskets). Generally, though, I use barrage because the Drill line is really only useful once you hit Drill 3 and 4 and most siege won't see that many promotions.

Sometimes I'll even go with combat promotions. Particularly when speaking of artillery or mobile artillery (later units) as a single combat promotion adds a respectable amount of strength to base strength. If you can squeeze in combat 2 (which you can sometimes) you're looking pretty.

I think that the consensus on Drill is that it is only worthwhile if you're going to promote up to the Drill III+ level, which seems unlikely given the XP that would require. And even then, its main benefit isn't in increasing survivability even-footing fights-- it's in decreasing the damage the unit takes when dispatching inferior opponents (e.g., stack defenders lasting through a wave of counter-attacks).

I'd take barrage.

I take barrage? whichever one that is going to cause the maximum amount of collateral damage to the defenders. Loosing the seige is a crapshoot so when you suicide them you might as well be damaging the most things the best you can. IMHO at least.

Agree on the Drill 3+ only being worthwhile. I'd consider combat for some and barrage for others.

Loosing the seige is a crapshoot so when you suicide them you might as well be damaging the most things the best you can. IMHO at least.

At the time of cannons, I'd say that suicided units only really happens with tech parity opponents. I don't lose too many CR3 cannons vs. muskets or longbows. If the 1st one loses, you can be pretty sure the second and subsequent will survive.

Due to the fact that CR3 siege can still sometimes lose to inferior units, combat may help in that one. I've had siege with CR3/barrage 2, so CR3/combat 2 shouldn't be too bad. IIRC, that makes cannons have a battle strength of 14.4 to accompany the CR3 which reduces defender bonuses by 75%

Drill 2 and drill 3 both average 1 extra first strike.

I'd take drill, cr3 siege weapons don't do any more collateral damage than un-upgraded ones.

I'd take drill, cr3 siege weapons don't do any more collateral damage than un-upgraded ones.
I don't think you understood his question. Barrage definitely adds to collateral damage done and to # of units dealt damage to. CR3 itself does no extra damage, but he was asking if he should seek the extra damage Barrage promotions or Drill.

Drill 2 gives 1 first strike to add to Drill 1's 1 first strike chance. Drill 3 adds 2 more first strike chances on top of everything else. Drill 4 adds 2 guaranteed first strikes on top of all that. So really, Drill 3/4 are what make the drill line worth it.

I think it goes something like this. If not, it gets better at 3 and 4
Drill 1: 1 first strike chance
Drill 1/2: 1 - 2 first strike chances
Drill 1/2/3: 1 - 4 first strike chances
Drill 1/2/3/4: 3 - 4 first strike chances

and since siege units innately do not receive siege based collateral damage, those bonuses don't really matter.

Agree on the Drill 3+ only being worthwhile. I'd consider combat for some and barrage for others.

...

At the time of cannons, I'd say that suicided units only really happens with tech parity opponents. I don't lose too many CR3 cannons vs. muskets or longbows. If the 1st one loses, you can be pretty sure the second and subsequent will survive.

Due to the fact that CR3 siege can still sometimes lose to inferior units, combat may help in that one. I've had siege with CR3/barrage 2, so CR3/combat 2 shouldn't be too bad. IIRC, that makes cannons have a battle strength of 14.4 to accompany the CR3 which reduces defender bonuses by 75%
If my memory and this (http://www.civfanatics.com/civ4/info/promotions) is accurate, siege units can't take Combat promotions.

I don't think you understood his question. Barrage definitely adds to collateral damage done and to # of units dealt damage to. CR3 itself does no extra damage, but he was asking if he should seek the extra damage Barrage promotions or Drill.

Drill 2 gives 1 first strike to add to Drill 1's 1 first strike chance. Drill 3 adds 2 more first strike chances on top of everything else. Drill 4 adds 2 guaranteed first strikes on top of all that. So really, Drill 3/4 are what make the drill line worth it.

I think it goes something like this. If not, it gets better at 3 and 4
Drill 1: 1 first strike chance
Drill 1/2: 1 - 2 first strike chances
Drill 1/2/3: 1 - 4 first strike chances
Drill 1/2/3/4: 3 - 4 first strike chances

and since siege units innately do not receive siege based collateral damage, those bonuses don't really matter.

I don't think you understand my answer. In reverse order, I'm pretty sure first strikes aren't normally distributed, and a first strike chance has a 50% chance of happening whereas a first strike has a 100% chance. Therefore 2 extra first strike chances have a 25% chance of 0 first strikes, 50% chance of 1 first strike, 25% chance of 2 first strikes. If you like the drill 3 bonus, you like the drill 2 bonus.

Potentially losing a 17 exp unit for the same collateral damage as a 3 exp unit doesn't make sense to me unless you're doing additional collateral damage. Going city raider in the first place means foregoing barrage, so you promoted the unit to survive better (and to do more damage), which also implies you tend to bring more than enough siege. See countless threads on barrage vs city raider.
Now you attack a strong defender, and you expect the cr3 siege to win or you don't. If you expect it to win and have a few other cr2 or cr3 siege, then there's no need for barrage. If you expect it to win, they have other strong defenders, and you have no other level 2 siege, then barrage might be a good idea, but this seems like an odd scenario.
If you don't expect your cr3 siege pult to survive, you should probably soften it up with a suicide pult, and there's your barrage 1 damage + 50%. Then since you didn't expect your initial attack to survive you'd be better off with drill 1 to slightly increase your odds of cleaning up.

Other arguments: cr3 siege will do a lot of attacking, so you want to reduce damage to heal faster.

Thanks for explaining your answer with regards to drill.

However, a 3xp unit only gets 1 promotion (in this case CR1)
A 17xp unit gets 4 promotions (in this case CR3 and perhaps Barrage)

Which means 1) much higher survival potential and 2) more units dealt collateral damage from barrage promotion.

I'm not sure what you mean by "if I like drill 2, i like drill 3". Not so much as there are many more chances at having first strikes hit with Drill 3 (since you only get 1-2 with drill 2) and Drill 4 only adds to this. I know for a fact that Drill1/2 gives you 1-2 first strikes as per the unit roll over. I can't remember the exact #s for 3 and 4 but they are much better

If my memory and this is accurate, siege units can't take Combat promotions.
Ah, my memory must be serving me poorly. Took a 3 month hiatus from Civ, just getting back into it.

I think vic's point is that Barrage is a promotion for siege weapons that you expect to lose. The CR3 doesn't improve the collateral damage you get from Barrage-- it's the same you would get if you were using a Barrage I unit straight out of the Barracks. The only difference is that the CR3 unit is more likely to withdraw and survive the battle. Thus, if you want the Barrage effect, you're better off using the green unit, rather than risk the loss of the 17 XP unit. Adding Barrage to the 17 XP unit does nothing to protect that unit's survivability, while Drill I does. As such, it's a better pairing of promotions with the CR3.

It's a clever argument. I'm not entirely convinced because the benefit of D1 is so tiny, and because even if you only use your CR3 cannon in battles you expect to win, the extra barrage is still a nice bonus to reduce the chance of other defenders getting a pesky victory.

Drill 2 helps more in the open field where you can't rely on CR promotions to gain the upper hand, and you want your Cannons to survive to fight the next battle.

This especially becomes significant against mid-strength units like Macemen and Musketmen (although the first strikes are useless against Knights).

"Specced to lose" never struck me as a good way to promote. So Drill after CR for me. Having siege that survives is more collateral damage per unit in the long run than certain suicides with better collateral. Also Drill helps when defending, so in Steel beelines I usually have some dedicated Drill cannons mixed in with all the CR ones.

For the OP, I promote siege weapons on Drill frequently. Drill promoted siege constitutes a bigger part of my artillery corps than City Raiders. Here's my reasoning...

In the case of city assaults, my first preference is using a Drill siege unit. If my best Drill unit is getting poor odds, then I check with my CR arty. (With each level of promotion I have a higher threshold of acceptable odd, due to the XP investment.) If my CR is getting all bad odds, I promote the greenest one with Barrage and he becomes an hero. Rinse and repeat until all of the city defenders are redlined or all of my siege is damaged/moves expended.

For battles out in the field, the process is simpler because CR doesn't help (except the 10% vs gunpowder at CRIII). Whether to use a Drill or Barrage suicide depends upon the odds available for each attack. In both cases, 'good' and 'poor' odds is situation dependent and the strategic big picture dictates how crucial the battle is.

That's not to say I only use Barrage for one way missions. If I'm fighting a technologically inferior army and the enemy goes for a zerg rush (e.g.: Monty and with a stack of over 9,000 Jaguars against your Riflemen), Barrage III provides double the collateral damage making it quite attractive. If the odds are good and I know the siege units involved will have plenty of time to heal, Barrage provides more bang, literally. Conversely, the damage reducing effects of Drill are more valuable when those siege units will need to fight again very soon (e.g.: epic conquest campaign). On that note, siege units built at the Red Cross city can immediately get March which is perfect frequent combat.

specc'd to lose? I assume that is directed at me and my usual, (but not always) method of siege promotion. Yes one or two die but the barrage specialty has smashed everything to smithereens after that. Even when I did take city raider or drill chances are I would lose one or two just from bad rolls.

@blitzkrieg.

Yes, if I have such a huge tech lead that I am fighting cannon v longbow I will promote city raider for clean kills with minimal losses instead of barrage. This does not happen much for me on immortal/diety though. Many fights occur at near tech parity. (i still am != to winner on diety at all)

I don't think you understand my answer. In reverse order, I'm pretty sure first strikes aren't normally distributed, and a first strike chance has a 50% chance of happening whereas a first strike has a 100% chance. Therefore 2 extra first strike chances have a 25% chance of 0 first strikes, 50% chance of 1 first strike, 25% chance of 2 first strikes.

No. The conversion of first strike chance to first strikes is uniformly distributed (in your example, a 33% chance of 1 first strike). Your binomial distribution would be correct if each first strike chance were implemented independently, but that isn't how the calculation is implemented: see CvUnit::setCombatUnit.

Barrage definitely adds to collateral damage done and to # of units dealt damage to.

Oh yeah? Says who?


I'm not too tempted by Drill I, because the parlay really sucks

The FirstStrikes die roll goes against you, and you don't get nuthin'
You get the first strike, but then win the first combat round. Your first strike counter is decremented, and the net result is just like you had no first strikes.
You get the first strike, and lose the first combat round, but inflict no damage between the round that would have killed your unit, and the round that did kill your unit. The outcome is unchanged - just the order that the damage was done (no effect - all the odds are based on strengths of the units before the combat begins), and using up more random numbers, to speed the heat death of the universe.



In thinking about Barrage, it may help to consider the effect that promotion has on the collateral damage caps.

Wow I said aren't normally distributed then I gave a binomial distribution, so smart.

In any case, it ends up averaging the same with 2 first strike chances: probability of 1/3 for 0, 1/3 for 1/ 1/3 for 2. Averages 1 first strike. 1 extra first strike, averages 1 first strike.

Drill 2 +1 First Strike
Drill 3 +2 First Strike Chance
It should average out the same.

In any case, it ends up averaging the same with 2 first strike chances: probability of 1/3 for 0, 1/3 for 1/ 1/3 for 2. Averages 1 first strike. 1 extra first strike, averages 1 first strike.

Ah, I see what you are going for. Yes, exactly right.

On average....
Drill I adds 0.5 first strikes
Drill II adds 1 first strike
Drill III adds 1 first strike
Drill IV adds 2 first strikes

Ergo, if Drill III is useful (one more first strike on average), then Drill II should also be useful (one more first strike).

Drill is really useless except for the promos it unlocks. Once in a while I'll want to use a catapult as a medic. Or I'll have a 5 xp catapult that I want to use in the open field, where I can predict which unit will defend. Drill -> Shock or Drill -> Cover might be appropriate. Combat used to be a good choice for open field use, until it was disallowed as an option.

Siege units almost all get CR, until I've got a large stack and I know I'll be knocking down several cities' walls. Then it's worth promoting some to Accuracy. I don't generally use Barrage unless I'm really throwing them away on hopeless fights and the defending stack is numerous enough that I'll get full use of the collateral. In closer-but-still-unfavorable fights I prefer CR to do greater damage to the lead defender.

It's a clever argument. I'm not entirely convinced because the benefit of D1 is so tiny, and because even if you only use your CR3 cannon in battles you expect to win, the extra barrage is still a nice bonus to reduce the chance of other defenders getting a pesky victory.
That's my point. The CR3 siege is expected to survive. Drill I is barely noticeable with respect to survivability, so why not utilize the barrage to add damage.

Oh yeah? Says who?
Says seeing 6 units affected by collateral damage with Barrage promotions and only 5 without. Memory may not serve correctly, but I know I've seen the same siege unit do different # of units of collateral on a huge defending stack. Just last night I had a game where my CR3 cannon did collateral to 5 units and my Barrage 1 cannon did collateral to 6. If there is some other more hidden coding that causes this, then I'm wrong about it. I'm just not privy to the reasons in that case.

But as to your question, it's just from my observations.

lol a lot of time would have been saved if some of you bothered to read First Strikes (http://forums.civfanatics.com/showthread.php?p=7720659#post7720659). ;)

blitzkrieg,
For all intents and purposes Drill 3 is the same as Drill 2. IMO Drill 3 is the problem with the drill line because it's less than half as good as Drill 4 and only as good as Drill 2. In PIG I increased it to 3 first strike chances instead of 2.



Anyway, as for that lovely CR3 cannon, drill will improve its survivability, if that's what you value. If you are expecting it to die at any moment, I'd go with barrage.

^ barrage does't work like that.
if you are after collateral damage: the effect w/ cr3 cannon w/ full health is the same as
no-promo cannon w/ 0.1HP left, save to the top defender.
The amount of units depends on the type of the collateral attacker only, the extra damage may occasionally help (vs drill for example) to increase the number of affected units.
It is applied only in the bold part.



iTheirStrength = pBestUnit->baseCombatStr();

iStrengthFactor = ((iCollateralStrength + iTheirStrength + 1) / 2);

iCollateralDamage = (GC.getDefineINT("COLLATERAL_COMBAT_DAMAGE") * (iCollateralStrength + iStrengthFactor)) / (iTheirStrength + iStrengthFactor);


iCollateralDamage *= std::max(0, 100 - pBestUnit->getCollateralDamageProtection());
iCollateralDamage *= 100 + getExtraCollateralDamage();
iCollateralDamage /= 100;

if (pCity != NULL)
{
iCollateralDamage *= 100 + pCity->getAirModifier();
iCollateralDamage /= 100;
}

iCollateralDamage /= 100;

iCollateralDamage = std::max(0, iCollateralDamage);

int iMaxDamage = std::min(collateralDamageLimit(), (collateralDamageLimit() * (iCollateralStrength + iStrengthFactor)) / (iTheirStrength + iStrengthFactor));
iUnitDamage = std::max(pBestUnit->getDamage(), std::min(pBestUnit->getDamage() + iCollateralDamage, iMaxDamage));

if (pBestUnit->getDamage() != iUnitDamage)
{
pBestUnit->setDamage(iUnitDamage, getOwnerINLINE());
iDamageCount++;
}

blitzkrieg,
For all intents and purposes Drill 3 is the same as Drill 2. IMO Drill 3 is the problem with the drill line because it's less than half as good as Drill 4 and only as good as Drill 2. In PIG I increased it to 3 first strike chances instead of 2.
So the unit roll-over is misleading, then? Not that it would surprise me, a few of the rollovers and descriptions are misleading in Civ (not the least being the Dun's Guerrilla bonus).

^ barrage does't work like that.
if you are after collateral damage: the effect w/ cr3 cannon w/ full health is the same as
no-promo cannon w/ 0.1HP left, save to the top defender.
The amount of units depends on the type of the collateral attacker only, the extra damage may occasionally help (vs drill for example) to increase the number of affected units.
It is applied only in the bold part.

I think we all know that, but that's not the point. The point is that Barrage and CR have zero synergy. So if you have surviving siege might as well maximize their survival. Fresh troops can be suicided with Barrage as picking that line never gives a very good survival rate.

With Cannon wars (at least beelined) you usually get alot of relatively high level siege so you don't need to rely on the collateral aspect as much as the whack-the-top-defender aspect. No need to have the best guys do more collateral as well since after the first 2-3 attacks usually everyone gets favorable odds. If your siege numbers are very limited then CR->Barrage might be more viable.

With Cannon wars (at least beelined) you usually get alot of relatively high level siege so you don't need to rely on the collateral aspect as much as the whack-the-top-defender aspect. No need to have the best guys do more collateral as well since after the first 2-3 attacks usually everyone gets favorable odds. If your siege numbers are very limited then CR->Barrage might be more viable.
Going up against massive stacks (which is desirable since once the massive stack is gone, you generally have very little resistance to face) would dictate that surviving barrage promoted units is desirable. I see no reason to lose units with barrage-only suicide attack patterns. Perhaps with catapults in the very early eras, but once you're into Trebuchets and cannons, I'd rather have a stack of 10 CR3 Barrage 1 units than 20 mixed CR3 Drill and Barrage 1 suiciders.

It comes down to taste in the end as each person obviously is able to win utilizing their methods otherwise we wouldn't be defending them.l

That's the point about barrage: if you intend to suicide the unit, it's the best to take, otherwise it just plain sucks.
My point was only about the amount of units affected by collateral damage, in some circumstances the barrage promotions can increase that figure but usually they would not.
edit:
blitzkrieg1980, the effect of barrage is not so ground-shaking huge to truly matter. The max cap of damage is still there.

I'm used to seeing huge stacks now. So I'd rather have survivor siege doing extra collateral. Again, it's all flava.

So the unit roll-over is misleading, then? Not that it would surprise me, a few of the rollovers and descriptions are misleading in Civ (not the least being the Dun's Guerrilla bonus).

It's not misleading.

Drill 2 gives 1 first strike.
Drill 3 gives 2 first strike chances.

You can basically imagine 2 first strike chances = 1 first strike.

AH! I see. I read your Drill thread and didn't really come out of it with that view. But simplified, it makes more sense. Drill 2 is just about the same as Drill 3.

Still, Drill really only has a desirable effect at Drill 3 and 4. Although Drill 2 and 3 are relatively the same, at Drill 3, I there are enough 1st strikes guaranteed AND chances to be effective IMO. But it's only IMO ;)

On that note, siege units built at the Red Cross city can immediately get March which is perfect frequent combat.

Now there's an interesting idea I have never thought of. I guess it could be argued that it is a very niche strategy, as by the time I have the Red Cross I'm using bombers.

But I can see that a stack of medic I, march mobile artillery and mech infantry might be better at end-game mass conquest than the hordes of modern armor that I have been using.

Modern armor is good for rapid end game wins. They're fast, can attack multiple times, and unless your enemy has Advanced Flight, they really have no counter. A 5 turn war to end the game makes modern armor great.

Your strat sounds good too. Modern Armor just makes it happen faster since no counter.

wow - Drill and Barrage, really?

I read in old threads that CR is best for inflicting max col. damage. Even if the unit dies, the damage inflicted is higher and thus you need to use less siege attacks to inflict the max allowed damage.

Barrage I thought was posted that it's bugged and doesn't work as intended. I recall a an old thread of Barrage vs CR comparison with testing and CR came out ahead.

So if you're Spain and have 10xp trebs coming out, facing longbow defenses in most cities, you'd all promote barrage 3 or Drill 3?

So if you're Spain and have 10xp trebs coming out, facing longbow defenses in most cities, you'd all promote barrage 3 or Drill 3?

OP asked that once his siege has City raider 1/2/3, where should he go from there (barrage or drill). Pretty sure barrage works just fine (just not for tanks/modern armor lol)

OP asked that once his siege has City raider 1/2/3, where should he go from there (barrage or drill). Pretty sure barrage works just fine (just not for tanks/modern armor lol)

Thanks - apologies for not reading the op carefully. :sad:

I read in old threads that CR is best for inflicting max col. damage. Even if the unit dies, the damage inflicted is higher and thus you need to use less siege attacks to inflict the max allowed damage.

Barrage I thought was posted that it's bugged and doesn't work as intended. I recall a an old thread of Barrage vs CR comparison with testing and CR came out ahead.

That used to be true, but the world changed.

CR was never best for inflicting max collateral damage - except in the "live to fight another day" sense. CR changes the expected damage done to best defender only (for city defenders only, of course).

Hmm, after cr3 i dont care too much but i might go for accuracy. bombard w/ healthy cannons (need less) and the rest gets healed by the medic (do nothing). That means slow push and/or splitting the forces to bombard one turn, take on the next.