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

[jlehtone]

I saw a bit baffling OOS fight. My patrol Hydra (6 BHEPT, 2 BPBE, and AHEPT on Tur.M.Def) against swarm of KM5. One KM5 had shield damage. It was recharging, but occasionally it was hit a bit. At the same time, other KM5's were vanishing from the list, one by one. Almost like the main guns of the Hydra would have done one-hit-kills while the turret was tickling one KM5. But that turret is AHEPT. One would expect to see average hit more than what the KM5 can take. I guess that there were local police, although I did not see their icon on the stack.

[wyvern11]

yesterday i ran the script to check if player capital-fleet affects OOS-fights

i used the stupid-looking-argon-freigher pilot starting with a freighter and a disco - did nothing significant to player ranks, flight time was 57minutes
then

i did the following : 30 replays of colossus vs. colossus (falcon and P took too long because of pathetic damage-to-shield-ratio), then script three player-K, rerun...etc

these are the results:

Code: Select all

# K      player : xenon
  0        14   :   16
  3        10   :   20
  6         6   :   24
  9         5   :   25
 12         9   :   21
 15         7   :   23
 18         7   :   23
 21         5   :   25

from this i would draw the conclusion that having more than say a handful of capital ships definitely makes xenon angry and i looks like it does not make a large difference if its 10 or 20

[jlehtone]

from this i would draw the conclusion that having more than say a handful of capital ships definitely makes xenon angry and i looks like it does not make a large difference if its 10 or 20

Indeed. I wonder what 1, 2, 4, and 5 would yield, particularly 4 and 5?

So 6 M2 + 1 M1 is already "much". Maphys (IIRC) wailed in another thread that did he make a mistake by buying fleet of M2's and if M7 fleet would work better? So is it just M2 that triggers the enemy "boost"? One repeat with 21 M1 or M7 should say something about that.

[wyvern11]

so you agree with me on the fact that there is some singularity around 4-6 M2 in player-ownership?

lets see this in details
now that i am this far, it is no problem to test this for m7 / m6 as well

do you perchance know how to script fight / traderank (to test influences of these?)

so these are some of the results we think we found so far from my point of view
(to be continued or disagreed if necessary )

- OOS combat comes in two slightly different flavors - watchin sector-map
and not-watching sectormap. if sectormatch is watched, then combat timestep is 5 seconds. if not, it is 30 seconds

- although it looked otherwise on first notion, it makes no difference to combat outcome if one watches via map or not. the same combat with same damage-rates seems to take 6 times longer

- independently every 3 seconds shield recharge is calculated from shield generator strength after known rules and added to overall shield

- missile usage and damage is taken into account. for capital ships only torpedoes might affect the outcome, but in 1:1 situations are easily overpowered by heavy guns

- player possession of larger numbers of capital ships seem to affect strength of hostile races. it looks as if scale is tipped such that losses occur almost twice as often. how this is achieved by game remains unclear yet

[BrigandPhantos]

jlehtone


(snipped and removed)


I didn't realize how much there was to this thread, I spoke too quick. Sorry.

[zazie]

do you perchance know how to script fight / traderank (to test influences of these?)

I have never used it but it may be helpful: King Luis Trainer

[googlemeier]

player possession of larger numbers of capital ships seem to affect strength of hostile races. it looks as if scale is tipped such that losses occur almost twice as often. how this is achieved by game remains unclear yet

Most likely. My whole fleet was shred to bits in about three to four hours.
I even noticed that one X-K (though heavily damaged after the first one) took out two of my Ks including a P, some LX and Ls.
Great. More than 40 hours spent on the fleet and there it goes.

[wyvern11]

an addition to result-list:

- in my tests i could *not* confirm that carriers with docked fighters are OOS superior to same carrier empty - it looks like firepower of fighters is not added to strength

i did quite an amount of testing yesterday how many capital ships the AI tends to ignore but these gave such strange results, that i intend to do a complete overhaul of my testing script (results differed greatly depending on which ship i used as combatant for both player and xenon, player-dragon vs. xenon-dragon 38:22 - player-K vs. Xenon-K 10:40, although every other parameter was the same)

[zazie]

in my tests i could *not* confirm that carriers with docked fighters are OOS superior to same carrier empty - it looks like firepower of fighters is not added to strength

Good to know. Therefore the AI is acting on a 'realistic' basis:
- docked fighters are not pushing fighting power
- unmounted weapons/shields (in docking bay) are not pushing fighting power

[Carlo the Curious]

I saw a bit baffling OOS fight. My patrol Hydra (6 BHEPT, 2 BPBE, and AHEPT on Tur.M.Def) against swarm of KM5. One KM5 had shield damage. It was recharging, but occasionally it was hit a bit. At the same time, other KM5's were vanishing from the list, one by one. Almost like the main guns of the Hydra would have done one-hit-kills while the turret was tickling one KM5. But that turret is AHEPT. One would expect to see average hit more than what the KM5 can take. I guess that there were local police, although I did not see their icon on the stack.

In itself that's interesting. I was under the impression that OOS one ship would only attack one target, regardless of the number of turrets.

[wyvern11]

after overhauling:

player owns next to nothing: rate of combat success:
50% (n=650)
player owns 1 destroyer (xenon K)
48% (n=230)
player owns 2 destroyers
46% (n=220)

further values will follow

[Trehek]

This entire thread has been very interesting to read, but I would have a request now:

Since the thread has become very long, could you guys please edit the OP and add your key results and speculations there? It's getting quite hard to follow your chain of tests.

Something on the line of "Current theory" would be appreciated.

[wyvern11]

@ trehek - i tried to start this - see page 7 of thread and below:

- stepping up the amount of capital ships seems to worsen player win rate - but the results tend to blend into one another - having 20 M2 resulted into at least a 10% drop. i tested if a larger amount of Hyperion (30+) leads to a similar drop but it didn't look like that - so it really seems that some corvette-only-players' (zazie f.e.) experience of "easier" OOS-fights might have something to it

[googlemeier]

One Xenon K has just destroyed 4 of my Ks...a joke.

[newdean]

This entire thread has been very interesting to read, but I would have a request now:

Since the thread has become very long, could you guys please edit the OP and add your key results and speculations there? It's getting quite hard to follow your chain of tests.

The nuggets I've carried away, their implications, and my request for clarification and further testing:

1. As you acquire more M1s and M2s, OOS combat gets harder, and combinations that would have been easily defeated at an earlier time when you had fewer capital ships become steadlily more deadly to your OOS gatekeepers. This does not appear to happen to the same extent when you load up on M6s, M7s and TLs. Implication: (A) Don't add M1s and M2s unless you want to have the OOS landscape continue to become steadlily deadlier; (B) there is a point it might be wiser to stop at, before adding further capital ships; and (C) this enormously complicates cross-game comparisons of ideal OOS gate keeping forces: My Elephant plus Dragon plus LXs and Pirate Falcon Vanguards may be very successful holding forces in my game, but because you have more capital ships, might be a suicide squad in yours, gate keeping the exact same sector.

Need for further testing: Where during that critical point between 3 and 6 capital ships do the odds get worse, and by how much. (Kind of an important question right now, Wyvern11, since I just added my fourth M1/M2, but I still have a saved game before that add.

2. Hangared fighters appear to have no value OOS. Implications: (A) Don't waste your money on stuffing the hangar of your OOS Raptor or elephant (unless you deploy the fighters once it's at the sector its patrolling/protecting); (B) M2s are much more cost effective OOS killers than M1s, if you can deal with their sluggish speed.

Need for clarification: Wyvern11 and jhetone, when you tested this did you have the hangared fighters set so that the mother ship they were hangared in was their home base? IS that's the only setting that causes the fighters to protect the mother ship.

3. Carried (cargo) but not installed weapons are not taken into account in OOS fighting, and have no value OOS unless then installed. Implication: Your raptor you have optimized to take out swarms of fighters IS with flaks probably ought to have those weapons swapped out with GPPCs before you send it to guard a sector OOS.

4. Missiles actually may have some utiltiy OOS, but only the most damage-causing ones.

5. OOS TL set on "attack all enemies" can do well against a popped cluster, but set on "attack [cluster]" flubs spectacularly, as it does not select a second target once the cluster pops. (This effect caused me to misunderstand the mechanics of what happens when you watch OOS combat--I thought it was watching that had nerfed my TL, but it was the attack command I'd given it.)

6. Watching or not watching has some effect on the timing of combat rounds, which has some impact on the durability of the losers (takes longer to die, if you're jumping in to save it.) Might have some impact on matchups of beams vs. shield regeneration but uncertain, and for the most part OOS is OOS, very similar whether you watch or not.

[newdean]

Another thing that would be nice to see tested: The utility of drones, OOS. We do know that UTs (and other AI freighters) deploy them OOS, because you can see them left over. And the massive drone dump (OOS) was a technique advocated on this forum by some to deal with things like Khaak invasions in sectors with lag problems. And still others have pointed out that in what resembles a turn-based combat where a ship picks one target at a time, a multitude of small ships, like drones, can (A) whittle away at larger ships because they are only killed one at a time, and (B) draw fire, causing potentially deadly strikes that could have taken out a large ship in one hit to be wasted on a drone instead.

[wyvern11]

@googlemeyer: did the enemy K have a fighter-patrol as they usually have ?
then chances are high that your K attacked a fighter giving the Xenon perfect time to destroy your fleet one after the other. How much capitals do you regularly own?

@newdean
impressive compilation of the theories so far - thanks a lot
- to drones : i used discoverers as drone-replacement. it looked like the xenon making more often the mistake of attacking the scouts (your option B) instead of the destroyers - but it was not really a success to my op - the xenon were still terribly insistent on attacking the destroyers - as a rule my patrols ran on patrol-command - maybe there are better options?

- to an unknown m1/2-threshold - with my scripts somewhat advanced i have not been able to reproduce the sheer drop in success rate between 3 and 6 destroyers any more - newer results indicate a slow decrease of rate - so no sweat for your buying a 4th m2 - having lots of corvettes did not decrease probability

- to hangared fighters : my test went with a colossus , 40 M3+, fairly maxed out (as a rule full shields 8ahept) , docked at and homebased to the colossus. *if* they were added to the m1-stats they would have thus almost doubled the shielding and tripled the guns. loss rate was almost the same docked as empty and much higher than with all fighters undocked and ordered to kill all

and now some advice: do not take my experiments all too serious - they stem from testing series which i did my best to do as reproduceable as possible. there is no evidence based on hard facts for these theories - it might all be "bad luck" only

[newdean]

and now some advice: do not take my experiments all too serious - they stem from testing series which i did my best to do as reproduceable as possible. there is no evidence based on hard facts for these theories - it might all be "bad luck" only

You're too modest, wyvern11--This has been great stuff, and you have used large enough samples to provide a higher degree of certainty than the usual "I quickly got two captures" school of recharacterizing random noise.

And your doing the experiments in ways that can be repeated means that they can be verified (or corrected by counter example) later.

Given how much impact a full hangar of M3s should have had, your experiment is strong proof that they aren't included in damage calculations.

[googlemeier]

@googlemeyer: did the enemy K have a fighter-patrol as they usually have ?

Yes they had as they always have.

then chances are high that your K attacked a fighter giving the Xenon perfect time to destroy your fleet one after the other. How much capitals do you regularly own?

I didn't check but whenever I did they were attacking the X-capitals.

I OWNED 22 K, because after the tins had destroyed most of them and I was down to 2 Ks I installed the XTM Mod and restarted the game.

After nearly 80 days of playing since Nov 2005 the last challenge to be met was clearing 347 and 472.
With some effort I gathered 23 Ks and put 22 of them on patrol (I wouldn't let my first K be cannonfodder).

Now I own a mercury and 9500 cr, I'm strolling through Herron's Nebula looking for some lousy Meatstaks to be sold to a Military Academy.

And you know what? I seem to like it. Funny, eh?

[wyvern11]

yeah - i know, pushing twenty destroyers up and down the board can become boring after a while - i see myself in an xtm-scout soon, too

to my tests depending on the strength of xenon in 101 you can either clear 347 *or* 472 (my having cleared 101 and 472 there remain precisely 5 capitals for 347 - and i have not enough hardware to keep them in constant respawn

some further numbers maybe tomorrow