Re: Stealth and Scattering Armor bonus
 

Is your damage dealt supposed to be the effective hitpoints of the ship, rather than just any number?

Re: Stealth and Scattering Armor bonus
 

<font size="2" face="Verdana, Helvetica, sans-serif">Actually I was. http://forum.shrapnelgames.com/images/icons/icon7.gif That number takes into account the change in hitpoints, and the ECM bonus of 15%.</font><hr /></blockquote><font size="2" face="Verdana, Helvetica, sans-serif">Great. http://forum.shrapnelgames.com/images/icons/icon12.gif Now I'M getting confused by the whole thing. And I'm an engineer - I should be used to arcane cost-benefit analysis formulas!

Re: Stealth and Scattering Armor bonus
 

<font size="2" face="Verdana, Helvetica, sans-serif">If I was as petty as Geo, I would respond (with more than this not-such-a-response) to that. As I am not, all you get is http://forum.shrapnelgames.com/image...s/rolleyes.gif

Re: Stealth and Scattering Armor bonus
 

Suicide Junkie:
That depends on what you want to calculate. The damage dealt is the damage your enemy deals you. I intenionally left it there so that you can also include enemy's firepower(that also an answer to your question). This can give you more accurate view. Yes, the formula works with amplifying and decreasing basic damage. At first I thought two things that later turned out to be false: That I can't use "virtual hit points" instead of to hit(ot to defend) %. With the second I will not bother anyone until someone finds that possibility(It has something to do with very big or low chances of hitting). I hope I answered your question well enough.

[ March 20, 2003, 19:03: Message edited by: Ward ]

Re: Stealth and Scattering Armor bonus
 

If you are trying to maximize the defense strength of your ship (using a given amount of space, of course), then the amount of damage the enemy fires at you per round dosen't matter for shields and stealth/scattering armor.

And whenever the to-hit chances get locked at 1% or 99%, then our formulae will certainly be off.

Re: Stealth and Scattering Armor bonus
 

These are good formulae if you know enough math to appreciate what they are saying as a general principle. In practice though, you almost never have a reliable figure for enemy's chance to hit you, and even if you had full information, it's not a single number, but a probablility distribution, and a different one for each enemy design, experience level, weapon, etc.

So, this illustrates the theory, and would be a reasonable system for an improved AI to use, and if you're really confused and think this will help your designs, it'll give you something to give you advice, but there isn't really a valid formula that will tell you what the best thing to do is.

PvK

Re: Stealth and Scattering Armor bonus
 

Suicide junkie:

<font size="2" face="Verdana, Helvetica, sans-serif">Yes, you are right. But you can eliminate it if you wish. Just make an equation you know is right and cut the "damage dealt" off. The formula was meant to be as universal as possible. It could certainly use much refining but I thought this should suffice for your comparisons.

<font size="2" face="Verdana, Helvetica, sans-serif">Again you are right. But if you begin to count the formula for such extreme circumstances, you should also count how many times enemy fires at you (and the math of many chances - two chances of 5% aren't one chance of 10% - is much more difficult and I think it's useless in this example).

You showed some interest in that so tomorrow I'll post here a complete Version taking everithing into consideration (I'll look at it at home. Now am at work and I don't have much time).

P.S.: Excuse my English, please. :-))

Re: Stealth and Scattering Armor bonus
 

True, although hulls and most weapons have no accuracy bonuses, and designs will typically include the maximum attack bonuses possible with the available tech.

Given the range your ships fight at, and some knowledge of the enemy's tech level, you can get a pretty accurate baseline.

The true break-even point will be a fuzzy region around the calculated point, but if you are well on one side or the other, you can be fairly certain which of the two options (ECM-armor or shields) is best.

There are also missiles, talismans and rammers to worry about, as well as one-shot-one-kill weapons.

Given the complexity of the game, formulae like this will be a guide, rather than a law.
Before I went through the formula, I had no idea when it would be worth putting Stealth armor on a ship http://forum.shrapnelgames.com/images/icons/icon7.gif

Re: Stealth and Scattering Armor bonus
 

So, in short you can micro manage ship design or do as Unknown_Enemy suggested and put 1 of each on each ship worthy of it.
Or, I guess if you did it enough you would get an relyable gestimate for when it was worth it.

I'm too lazy, I'm going with Unknown_Enemy.
Enjoy

Re: Stealth and Scattering Armor bonus
 

Here comes the post I promised yesterday.

This formula works with statistics. That means it doesn't matter if you hit 1% of time with stength 100 or 100% of time with 1 damage(with little refinement you can make it matter). I will not give you a complete formula, but rather only pieces that hold together. I have also counted with tactical and strategical elements(different magnitude of weapons depending on range, movement, strategy chosen), effectivity of separate components and component's resource cost but then I decided not to include them here(I got to a point where it would be easyer to ask programmers for source code http://forum.shrapnelgames.com/images/icons/icon7.gif ). If someone is interested I can post it too(but the post will be thrice this long). If you find a mistake, please check if it is a mistake of principle or just hitting the wrong key(I wrote this late at night) before you accuse me of something wrong.

Abbreviations:
!: "faktorial"; I don't know the right word here. 5! means 5+4+3+2+1, 1! is 1, 10! is 10+9+8...+2+1. (3!)! is (3)+(2)+(1). Just substitute the number with the Last one minus 1.
AAW: "All active weapons" Sum of all weapons that can fire this turn. Used to calculate how much you actually hit after substracting efficiency.
ATH: "Attacker's to hit" Sum of all attacker's positive attack modifiers (experience, weapons) esxcept CAM.
CAM: "Constant attack mofifier" The bonus(penalty) you recieve from modifiers that don't change(ship size, ?).
CDM: "Constant defense modifier" The bonus(penalty) you recieve from modifiers that don't change(ship size, event predictor).
DD: DP*HE; "Damage dealt" HE-CDM has can be 0-1.
DP: "Damage potential" Sum of all weapons' damage. Do not apply anything except bonuses for mounts.
DRRA: "Damage resistance of remaining armor" Sum of all. The difference between DRRA and DRRS can be important if you implement weapons that bypass certain defenses.
DRRC: "Damage resistance of remaining components" Sum of all components' damage resistance. Do not include any armor.
DRRS: "Damage resistance of remaining shields" Sum of all. The difference between DRRA and DRRS can be important if you implement weapons that bypass certain defenses.
DTD: "Defenders to defend" Sum of all defenders' positive defense modifiers (experience, components) except CDM.
E: RSDR/OSDR; "Effectivity" This represents the proportion of your functionality.
ED: DRRC+DRRA+DRRS; "Effective defense"
HE: ((TH sqared with AAW)!)/(100 sqared with AAW!))*AWW +CAM -CDM + E; "Hit efficiency" This value can be 0-1. First you determine pure chances of attack. Then you add all constant modifiers and how damaged you are. This is the value that modifies the damage you deal.
OSDR: "Original ship damage resistance."
RESULT: ED-DD
RSDR: "Remaining ship damage resistance" DRRA+DRRC+DRRS
TH: ATH-DTD; "To hit"

This is a core from which you can derive anything about effectivity about whole ships. Now if you search for a break-even point just give wanted values to each side and give there an equation(of course leave some values as unknown). One equation is for 1 situation. For optimization purposes it's best to make a clumsy design that includes the type of components you want to include. Then Count higest and lowest numbers and find a golden heart.

Just a rough example(If this works well): A hi-end battleship +20 kt comuter core +50 kt engines(100kt) +240 kt of shields(375 shields) + 10 multiplex +10 senzors(65%) +10 ECM(60%) +30 Stealth(100 armor,15%) +50 Scattering(150 armor,15%) +20 kt Quantum +20 Sail +40 PD +300 Energy(90,135dmg/1round). It has overall +65% to attack and +90 to defense. It's components' damage reduction is 770, 2250 are shields, 250 armor. It can cause maximum damage of 5*135=675 kTs. It's overall defense is 3270 pionts. If this value drops to 770 the ship is beginning to take serious damage. 3270-770=2500/675=3.7 This means that it can take 3 full hits until it takes damage to internal systems. Let's optimize this ship. We just want to switch combat parts. ECM, Senzors and Multiplex are left on the ship. We have 800-420=380 space left. We want to have RESULT as high as possible. Now the effectivity of that weapon(on that range) is 3,75. Effectivity of shield is 10,375. That indicates Shields are much more efficient in this example. For simplicity let's assume HE is 0.2; TH=0.2(I know that this break a little bit this example, but it't O.K. for illustrating). DD=DP*0.2; ED=? Result "is bigger than" 0 (intead of zero I should have inserted the RESULT value of previous desigh, againg simplifing) DP*0.2 -ED "is bigger than" x; DP*0.2"is smaller than"x- ED; Maximum ED is 380(free space)-60(At least one weapon)/40 (space used by maximu shields, round down)*(value of each shield)=3000 Minimum is 0. Maximum DP is 380/60=6*135=810. Now we search a value that best fits these requrements. DP*0.2"is smaller than"3000; DP "is smaller than"15000 The best value that approaches this apparent no-sense is 810. This no-sense indicates also when it would be more effective to add weapons instead of armor. No surprise; the component effectivity values indicated that. Using these guidelines you can count any combination. Adding components with % is somewhat more complicated. You count the component effectivity and add the percent to overall ED or HE. This requires a bit math or an hour with Excel. If someone is willing to make that small program, you may know your answer to what universal combination is best.

WARD