HUGE Off-type Defence Bug

Yes, I've noticed this too. Sometimes the wrong defense is stronger than the right defense.

What I think is happening is that the value that is degrading is the theoretical value, not the square rooted value. In that sense, then, the defense won't appear to be degrading between hits, meaning that EACH WEAPON is rolling against that same square root value. If you're using, say, Harpoons against a ship with 16 beam defense, you've got a problem: each of those 3 damage projectiles needs to roll against 3-4 defense. At such low defense values, it's not unsurmountable, but once the defense gets high there will be no degrading of effectiveness. This is a serious flaw...

The obvious fix seems to be to degrade the square root value and then figure out how much of the theoretical value should be left after the hit... basically, off-type weapons will degrade defenses faster. The thing to watch for, though, is that players can't just stick a single off-type weapon on a ship to degrade the defense while the rest of the guns do the actual damage. The damage from the on-type weapons from ALL ships in the fleet would need to be applied first... or some similar scheme.

The way I think it *should* work, and the way I understood it to work, was that each ship calculates an 'effective' defence value in each of the 3 categories. So for a ship with 100 beam defence, it's effective defences are: 100/10/10. If it also has 16 missile defence, then it's effective defences are: 104/26/14. In combat, attacks degrade the effective values independently as if they were defences of the appropriate type.

Obviously, it does not work as intended.

I thought I noticed this in the beta with a pirate mega-event, but soon my ships outclassed their's and I forgot about it... Kryo have we got any comment from our coding geniuses yet? i'm beginning to think this is a real bug and not all in our heads

that is some high math, he he
I sugest that shield capacity to absorb missile damage is still sqrt value, and the missile damage reduces general shielding by the amount of damage absorbed AND sqrt value, so if ship have 16 shields, it have capacity to absorb 4 missile and 4 point defense damage, if it receives 5 missile damage, it absorbs 4, 1 damage goes to the hull and shield new capacity to absorb damage is 12 beam, 0 missile and 3 mass driver damage


one "quick" example (let just say that maximum values apply, for easier understanding):

defense: beam-missile-massdriver / missile-beam-massdriver / massdriver-beam-missile

Ship1 have defenses 16-4-4/16-4-4/16-4-4, 50HP
Ship2 have attack 3x9/3x9/3x9

beam A: 9 damage; beam defense reduced; new defense is 7-2-2/16-4-4/16-4-4
---notice that shield capacity is sqrt value---
beam B: 9 damage; beam defense destroyed; new defense is 0/16-4-4/16-4-4; 2 damage pass through shields, it is absorbed by missile defense, so defense is 0/14-2-3/16-4-4
---notice that missile defense is reduced by 2, missile capacity to absorb beams is reduced by 2, and missile defense capacity to absorb mass drivers is not reduced by 2, it is sqrt(14)=3---
beam C: 9 damage; as missile defense already absorbed 2 damage, it can only absorb 2 more damage, armor absorb 4 more damage, defense is 0/12-0-3/12-0-3 and 3 damage pass to the hull, HP reduced to 47...
--notice that missile defense is not capable to absorb any more beam damage, but it could absorb 12 missile damage or even sqrt(12)=3 mass driver damage!---

missile A: 9 damage; missile defense reduced; new defense is 0/3-0-1/12-0-3
missile B: 9 damage; missile defense destroyed, armor is reduced for 3 damage, and new defense is 0(0-0)/0(0-0)/9(0-0), defense is bypassed and 3 damage pass to the hull, HP reduced to 44
--notice that armor is penetrated to all beam and missile damage, but it still can absorb 9 mass driver damage!--
missile C: 9 damage; no defense so HP reduced to 35

mass driver A: 9 damage; armor is now destroyed, new defense 0/0/0
mass driver B: 9 damage; HP reduced to 26
mass driver C: 9 damage; HP reduced to 17

next round RELOAD and REPAIR, so ship2 receives 3 beam damage then 12 missile damage and first mass driver destroys the ship, two mass drivers left to attack the next one!

so, what do you say?

Oh it is absolutely real. If the system were working in any reasonable way, the beam defence ship could expect to win fewer than one in a million of those fights. Instead, it won 8 out of 9.

It may work as you describe. It's still a bug because it has the effect of making defences almost as powerful against off-type as on-type attacks. Or at least, Brad doesn't expect it to work that way, so it's a bug.

ok... i hae no idea wtf you just said, could you inlighten us non-mathmatical geniuses?

Goldie, it looks as though your proposed system is more complicated than I had realized. It may be good, but I think it's a little too much book-keeping to fit in the theme of GC2 combat.

More powerful.

So because defences don't degrade when they block an off-type weapon e.g. shields against mass drivers, when they are massed they give a guaranteed dice-roll advantage from the first to the last shot.

My opinion is that off-type defences should stop helping to repel damage when the on-type defence is zero to begin with, or reaches zero at some point during the combat round. At that point, a mass driver shot by the attacker is likely to sail through a hole in the armour and impact the hull, even if it is somehow off-course thanks to shield deflection or ECM against the attack.

Hi,

Is there a consenus that defence works not as specified, that it is better than stated in the manual?

Have fun

Well, there's no doubt in my mind that it's not working as advertised. I think that anyone who saw the same test case would be convinced as well.

Simply watching my own battles is proof enough for me. When I switched weapon types and ended up with ships that were MORE powerful than previous, I was expecting to be more effective, not less.

What kind of priority does this have compared to some people having problems with crashes, slowdowns, etc? Are the devs aware that there is actually a problem here?

They're probably paying more attention to tech trading right now--if not focusing on a different money-maker game altogether (as opposed to a free patch).

Stardock is really good about supporting GalCiv 2, so don't rush to any judgements unless they tell you otherwise.

Yeah, but this latest patch was, well.....

I'll send CodeCritter this way tomorrow if he hasn't seen this thread already.

Having re-read the example posted for the DA combat system, I noticed that weapons are targeted to cause optimal damage - that is, the weapon which can roll against the lowest possible defence value goes first. However, since this targeting method breaks down when a large off-type defence is present, I suggest an alternative.

Considering the same ship using the targeting formula, all ships should fire on-type weapons at the strongest defence if it exceeds the square of the second-best defence.

If there are shots remaining after this particular defence reaches zero, they are held in reserve.

They should then fire on-type weapons for the defence which was second best at the start of the combat round. Because overall defence has hopefully been reduced with the saturation attack, it's more likely that a little damage may be inflicted on the hull.

If there are weapons capable of degrading the defence which was worst at the start of the combat round, those are fired now, followed by any shots held in reserve because zero defence was reached.

Thanks Entorpy Avatar for letting me take a look at your saved game.

Yeah, this is definetly broken. And it is not because high attack values can be reached with weapons that all roll 0-1. His example the weapons being used are much higher on the tech tree. Maybe the off-type defense isn't degrading at all? If you play on full combat viewer you just get way too many 0s for damage done against a ship that should have squat armor against you.

I like the defense module changes, although it would be easier to see if they have hit a more appropriate game balance if this issue didn't exist. That being said, under the 1.5x patch, large non-matching defense has become a cheap and easy way to make invincible ships, and that is not fun or balanced. And it is not how it is supposed to work. This is yet another way to ruin the rock/paper/scissors strategy... for some reason rock no longer beats scissors, only scissors beats scissors.

I wouldn't worry too much about whether or not SD will fix this- given SD's History, there will be at least two more major DA patches- so we're still early in the cycle. That said, this is the first patch that I feel is a downgrade from before- causing more significant bugs then fixing of bugs. I see where SD was going with this, the things just need to be implemented better.

That said, I do trust SD to fix the problem in some form- I just wonder when it will be. Hopefully this month. I don't even feel a need to make some emo post about losing a customer ...

Suggested fixes.

On defenses- just changing the cost/tech requirements of the defenses and reverting it to the old system should work. Alternate suggested idea.

Sqrt of all defenses= a minimum defense so a ship with 10 shield and 2 md def will never go below 3 def. That might stop the bug.

On the tech trading- Hard to say what's needed here. This is the toughest balance issue SD has. I'm used to being able to tech lead easily, which is wrong, but the current system is wrong as well. Wish I had a better solution, other then the AI should be just as anal about AI-AI trade as AI-player.

Easier fix: have the off type defenses degrade by their square root value. Say you're attacking a ship with 26 armour with a beam weapon. You're attacking a square root of 5... if you do two damage to that, the square root value should descend to 3... meaning that the ship only has 9 armour left.

There's a certain logic to that. If you're attacking off-type defence, it shouldn't resist your attack as well as the on-type defence. However, I think that squaring the value of the attack is a little extreme. I would go for a degradation rate of three times the on-type rate, assuming that the on-type defence is unable to absorb any more attacks. That gives sufficient motivation for AI and human alike to put the proper defence in place.