OOS Combat Resolution (Was: OOS Missile firing test.)

[wyvern11]

i'm pretty sure, that for capitals max loadout *is* at least near the default loadout

but i see your point that it is necessary to set up an exact copy of the ship as opponent for starters

[ttl]

I would like to focus on the fact that the M2 inflicted serious damage with just one BFAA installed. In my mind this means that the the OOS damage is largely determined by weapon generator, or just hardcoded by ship type, which is really stupid when Ego claims to support modding. Would be nice to know quantitatively what happens when guns are increased. The problem is likely that the M2 can pretty much wipe out the pirate base in one "OOS shot" with just the BFAA.

[jlehtone]

I would like to focus on the fact that the M2 inflicted serious damage with just one BFAA installed.

I would not put too much weight on bit that without repeating the test. I'm not sure any more how long it took to beat the shield, it just felt quick. Wyvern's lack of two BPPC appeared to make a slight difference too and that would contradict the previous.

I had once Hydra with 8 BPBE (and 1 AHEPT) and for some reason it took a long time to cut the hull of Pirate Nova OOS, while shield remained low easily. 6 BHEPT + 2 BPE performs much better. Too much randomness.

[wyvern11]

@ttl

the BFAA - was it the only weapon *installed* or was it the only weapon *on board*

i thought that there was a discussion in german forum about that it sometimes looks in OOS fights that only one weapon is installed but others are used as well (might have been in correspondence with X2 and AEGIS/MEFOS ?!?!)

[jlehtone]

the BFAA - was it the only weapon *installed* or was it the only weapon *on board*

It was the only weapon installed and on board. Just having it on board was not enough to get the "Attack .." command running, it had to be installed, but once the "Attack .." passed its initial tests, it was possible to uninstall the gun without interrupting the command. AFAIK, uninstalled weapon on board did no damage, but maybe I did not wait long enough.

[wyvern11]

yesterday i ran a homebrewed script to check the outcome of N "identical"
battles of "identical" opponents - startet it as a bar brawl all ships in one location

ten tries - looked much like 50:50 (4 colossus each)

thirty tries - like 40:60 in xenon favor (6 colossus each)

I need to clear up some messy parts and fix some stuff but it looks like you cannot really state outcomes with less than 50 replays. to do this, i have to set up npc-race-ships, because being told, that some 300 carriers of yours are unter attack is really annoying

man, i lost more than 300 carriers yesterday - but it was fun to see

dunno why test with docked vs. no-docked fighters had such a different outcome !!

[jlehtone]

For reason or other, Pirate Base in New Income went hostile. Kind of sad, because I did like its position and I still want to hack the Anarchy Port. Well, another test.

Stripped all guns from K and installed only one AIRE on it. All turrets to 'None'. The Base has four Lasetowers, so told the K to attack one. Neither side makes permanent shield damage. One might see occasionally a small dent, quickly recharged, while watching the OOS, or right after starting to watch OOS (after minute of not watching). Thus, the weapon does indeed matter. The damage rate of weapons must exceed the shield recharge rate of the target.

For next Base, I'll keep the AIRE installed and fill the hold with other weapons. The NPC does not keep more than one weapon installed in turret if no enemies are near. But that is in sector. If the turret scripts are frozen OOS, then NPC ships do not install/uninstall weapons OOS. If they don't and each turret has only minimal guns by default, then the hold contents does count.

Added 2,000 Wasps into the K. No real difference. It seems that there can be several seconds between player ship "making a hit" (while watching OOS), and AIRE or Wasp does less damage than Lasertower/Base shields recharge. And Lasertower does less than 6*1 GJ recharges.

The '!fight.attack.target' script (which I assume to run on 'Attack ..') does indeed check at start if the ship has "laserenergy". Ship has "laserenergy", if it has a gun installed). The same script makes the attack run (at least on Capital ships) by either firing a missile or by firing guns.

Then I replaced the single AIRE with 8 BHEPT on K and targeted the Base. Missed first hit, but had to wait some seconds (watching) before another hit removed 59% of shield (1,8 GJ) and 20% of hull (300k points). Next observable hit (again several seconds later) wiped the minimal recharged shield and about 600k hull points. Third hit killed the Base. 8 BHEPT on M2 ship against a station.

[Edit] That missed first BHEPT hit. It took 1,2 GJ out. By specs, 8 BHEPTS can do that in 10 seconds. [/Edit]

According to specs, 8 BHEPTs could do 1,8 GJ of shield damage or 300k of hull damage in 15 seconds. If the "hit" would occur only once per 30 seconds, then the observed damages would "match". The downside is that almost anything should then kill nearly anything.

Hyperion was moved to same coordinates as hostile Lasertower. Turrets on "Attack all enemies". It looked like the lasertower attacked the M7. Changed the M7 to "Standby" form "None". No reaction. Only the change to "Attack .." activated the M7 and then the LT died quickly.


On guns and missiles the damage values are different. Could it be that their "chance to hit" would differ too?

[newdean]

OOS combat report:

Other observations:
* Only small fraction of Xenon ships has Missile firing probability other than 0%.
* Strong patrol ships are known to die, but not often.
* Xenon Patrols and Pirates can carry Hammerheads.
* Xenon Patrols and Pirates can quickly destroy even 5--7 GJ of shields.


All that seems suddenly very logical. Missiles.

Thanks for your test jhetone. It's great to get some more visibility into the murky workings of OOS combat. But I think there are some other explanations for why OOS patrol ships sometimes (but don't usually) die beyond hammerheads:

1. Ps can carry a wide variety of weapons, up to GPPCs. What they carry does seem randomized. So sometimes they're much tougher OOS than others.

2. There appears to be some randomness in what targets are attacked by the various ships in the patrol. If they gang up on even a TL (or M7) but the M7 starts knocking off the weaker ships first, the M7 or TL can get wiped out, because if took out the weaker rather than stronger ships in turn-based combat.

3. The combat resolution table has some randomness to it.

4. If you wait long enough, you can have simultaneous attacks by multiple patrols or P patrols plus pirates, or successive attacks, the second occurring while your shields are low from the first encounter.

The missile loadout may also be a big randomizer, but there are also other things at work leading to ocassional OOS patrol losses.

[wyvern11]

I did 50 tries with 4 xenon colossi against 4 argon colossi - was near 50:50 - ran quite good offscreen

will add some heavy missiles and change firing probability next - any reasonable-to-hack proposions`?

[jlehtone]

The 'fight.attack.target' calls a 'acp.*' subroutine with all ships that are not M1, M2, or TL. Could that lead to difference in OOS for Capital ships? Elephant did beat Hyperion when halo*** tested OOS.

The 8 BHEPT M2 (K) seemed to inflict 30 seconds worth of damage twice in a row on Pirate Base. Average Xenon P does not have full weapon compliment, but if NPC shoots first, then lone TL patrols should drop like flies. Could the rate of fire be different on different ships? Colossus with 12*GPPC+8*BPPC might make 22.2 GJ of shield damage in 30 seconds, if the weapon specs are really used for OOS. One would have to target a Dock to proof that.

[Cith]

My hypothesis about the OOS combat resolution is that it resembles D&D combat:
* Roll dice to see if you use a missile
* Roll dice to see if you hit
* Roll dice to see how much damage that hit did
* Recharge your own shields

I've hear many players say that, including myself. Dice are the way to go for OOS.

[misioooo]

And dice = save/load dice

[wyvern11]

If a colossus scores an OOS-hit to another colossus i noticed three outcomes more frequent than others :

1. instant kill
2. shields down to about a half
3. shields down to approx 5-8%

will test one-on-one-situations maybe this evening

EDIT: is orientation taken into account?

[ttl]

From these reports, my hypothesis is :

Damage per "OOS shot" is a random number from 0 to maximum damage dealt in 30s (OOS time step). This gives typical values as reported by jlehtone.

However, three questions remain :

1) What is the role of ship weapon energy and weapon generator?
2) Is there a hardcoded ship type factor?
3) Is there a difference between RoF and damage/shot? Since there have been reports of decreasing RoF, and correspondingly increasing damage causing OOS unbalance.

Missiles are very much a secondary issue, imo, compared to guns.

[misioooo]

I can only speculate on pt. 3.

If in OOS is taken damage/30s and damage is between 0 and 100%... Then if u reduce RoF and increase damage the formula will give different results. Or maybe it is taken like that:
1. Damage of one shot - taken from 0 to 100%.
2. AFTER 1 is calculated damage/30s...

From what i have read ship speed and energy does not matter in OOS. Only big guns really matter (even when you equip all GPPCs agains bunch of M5/M4/M3... IS u will loose (cant hit them with GPPC) in OOS - they will do better than weaker but faster weapons.

[wyvern11]

for clearance sake my observations stem from watching sector map
i take it that your observations stem from watching properties list?

btw:
where do the update-rate-information for
sector map
properties list
originally come from?

@misioooo :

concerning your part 1: imho it does not look that way in sector map. one almost always sees same fractions of shield/hull vanish - like it is calculated which battery scores a hit and then the full damage of the battery is dealt to the opponent - the more batteries score the better

@ttl : point 3: you speak about PPC and XTM, i presume ?
Did not really test XTM until now, but i heard that they lowered PPC firing rate to better performance an increased damage - i understand that from this OOS-issues arose?

[ttl]

@misioo: You are right that this sequence is not the same for a general probability distribtuion. It is for a uniform distribution, however. That is 1/a*f(a*x)=f(x) for a uniform prob. distribution f.

@wyvern11: PPCs with XTM definedly, but I recall seeing statements like that sometime before at these forums, too. And you're right, it would actually make sense to do the logic turret-by-turret, since every turret runs independent scripts on their own tasks. And maybe the math is just that the full power of the turret is applied or then nothing (a miss). Don't know, but experimenting with ships with only main guns would tell (they should always deal constant damage then).

In terms of modding the lasers to make for better performance and gameplay, the three guestions above are still the ones I would like to know. The details of the randomization procedure are not something we can influcence, and almost impossible to prove by experimentation anyway.

Oh, and the only difference between sector map and property list is likely the time step length. (5s when sector map open, 30s otherwise)

[zazie]

(...) But if command of the ship is ignored, then any Freighter with installed turret gun will cause damage on enemies OOS, even if player forgets to set any turret command.

Very interesting !
(slightly of topic:) I have written it many times in the German section (and of course it has been discussed in the English section too) that player should refrain from putting missiles onto their UniTraders for missile defence because there is a risk that the UTs start fighting/defending themselves instead of jumping away. But my advice was based on observation (and some very basical hand written statistics about death-ratio of UTs with and without missiles in the same sector).
It looks like you have found a objective explanation (or the hypothesis of an obj.expl.)

On topic:
What do you think about this reflection ?
In my (vanilla-)games I have never bought a ship of M2 or M1-class though I had by far the ranks I needed. My sector patrols were in 97% (or more) able to stop Xenon Patrols without loss, even after the LX appeared (we talk about OOS here). They had different set-ups (mainly Hydras, but Elephants, Centaurs and Nemesis as well; different number and type of wingmen).
But wyvern11 and others gave several examples that THEIR comparable patrols were no longer able to resist the LX-patrols. They lost more than I did in my game. One of the explanations on the German forums were: missiles.
But I had a theory that the overall power of the Xenon patrols increases not (only) corresponding to the player's rank but to the type of potential enemy, i. e. M2 or M1. The lack of this ship-class in my games seem to have an effect on NPC-behavior.
Comments on this theory ?

[wyvern11]

@ttl:
the 5s and the 30s i already heard - do you perchance know a source?

[ttl]

@wyvern11: The globals.txt file and http://forum.egosoft.com/viewtopic.php?t=178881 . Of course there is no way to tell that is the only difference, but I would guess so.