FreeOrion

I think i figured out chaff theory for single bouts. And some the results are

• for adding chaff units to a fleet of combat vessels of the same type, only two input values matter: number of vessels in the fleet and combat vessel cost/chaff vessel cost ratio
• that means there is a double-bias against cheap ships; if you buy two ships for the half cost, you have more vessels in the fleet (-> chaff is less effective) and the cost ratio is halved (-> chaff is less effective)
• Looking at a single bout, comsats are at 6PP still crazy effective, but if i divide the added value by the number of bouts not so much anymore

So my question: anybody knows how to value chaff (single-hit-decoys) for a whole combat? Is it a third for three bouts, because it is effectively only soaking hits in a single bout?

bump - still looking for input

I need time to finish a full combat simulator and give empirical input.

Ophiuchus wrote:Is it a third for three bouts, because it is effectively only soaking hits in a single bout?

Probably not.

When there are ships dying in the first bout, then the enemy first bout damage is higher than the one in subsequent bouts. Subsequently, your chaffs block a higher portion of the damage than 1/N.

Morlic wrote: ↑Wed Nov 20, 2019 1:54 pm

Ophiuchus wrote:Is it a third for three bouts, because it is effectively only soaking hits in a single bout?

Probably not.

When there are ships dying in the first bout, then the enemy first bout damage is higher than the one in subsequent bouts. Subsequently, your chaffs block a higher portion of the damage than 1/N.

I thought actually the opposite way. Looking at a third bout with the state of a single bout you need the same amount of chaff. That means in the bouts before you have to have extra chaff to soak hits (so e.g. for each shot in each bout one piece of chaff). But going back e.g. two bouts that means you need a lot more chaff than in the third bout which in turn means the shots spread out much stronger which in turn means overkill is much more unlikely which means you need extra chaff to compensate for the "missing" overkill.

This may be obvious, but:

Currently, there's no guarantee a chaff ship will be killed in any of the three combat rounds. In the current multi-player game, against one maintenance ship i had:

2 MS Frigates
1 "Martyr" (empty BSH)
1 "Mini-Carrier" (BMH w/ 3 fighters & 1 SAP)

The Martyr lasted 4 turns. One of the frigates departed after the 2nd turn with 4 structure left, one left after the 3rd round with 1 structure left. The Martyr took no hits the first turn, one in the second.

So, targeting randomness means it's not so simple as "the chaff dies first round of combat". They do (sometimes) die the first time they're hit, but that can take round and rounds.

Jon

So, its theoretic annihilation game time. In this game are ships and chaff. All ships are Robo Hulls, with 2 Laser-4 and 2 Zortrium (that is 2-hit-damage and 5-hit-structure). Chaff is a nameless ship with not enough hitpoints to survive a hit with Laser-4 (that is 1-hit-structure).

One the red side we have two ships, on the blue side we have one ship and N chaff. At the end of the game there should be no ship alive - complete annihilation! If that is so we can say that the fleets are equal in one sense. The question is what is N? How much chaff we need to "double" the combat value of the blue ship?

For the game I assume that it is probably enough to annihilate the 2 ships by shooting for 5 bouts with the two Lasers (there could be at maximum in one bout a 25% chance of 1 overkill, which would save one of the ships, but i will ignore that now).

My combinatoric spreadsheet tells me that starting with nine chaff the blue ship should make it into bout five barely alive and with no chaff left. On this side i ignore that maybe one of the enemies is gone before bout 4 which basically ensure survival of the blue ship. With eleven chaff I am pretty confident that the blue ship would survive.

The battle itself would slowly ablate the chaff shield (count: 9-6-3-1-0) until there is only one chaff left at the end of bout three and then the blue ship starts suffering for real (hit-structure: 5-4.6-4.2-3.4-1.9-0).
Overkill goes up slowly (0.5-0.7-0.9-1.1-1.1) and in whole adds 4.3 "imaginary hit-structure" over the course of the five bouts.

Red team costs 224 PP, blue team with 9 chaff at 6 PP costs 166PP.

Now somebody only has to reenact that in a freeorion multiplayer game

Ophiuchus wrote: ↑Fri Nov 22, 2019 9:47 pm Now somebody only has to reenact that in a freeorion multiplayer game

It appears that battle RNG is seeded so reloading a save I got the same results three times in a row. I was able to get different results by sacrificing a scout or two but this was too tedious for a statistically significant test so I only did it a few times. And might not affect the RNG enough, no idea.

Anyway, here are my quick results:

== 9 chaff ==
- Red side wins with 14 hp remaining on one ship
- Draw

== 10 chaff ==
- Red side wins with 14 hp remaining on one ship
- Red side wins with 25 hp remaining on one ship

swaq wrote: ↑Fri Nov 22, 2019 11:18 pm It appears that battle RNG is seeded so reloading a save I got the same results three times in a row.

This is probably controlled by the game setting: Random Reseeding (Top option in the single player "General" tab.)
Also known as RULE_RESEED_PRNG_SERVER

alleryn wrote: ↑Sat Nov 23, 2019 2:35 am

swaq wrote: ↑Fri Nov 22, 2019 11:18 pm It appears that battle RNG is seeded so reloading a save I got the same results three times in a row.

This is probably controlled by the game setting: Random Reseeding (Top option in the single player "General" tab.)
Also known as RULE_RESEED_PRNG_SERVER

Is there not a way to toggle this from the multiplayer UI?

Minimum chaff required for one ship A to ensure annihilation of another two identical ships:

Shots per ship	Hits to die per ship	Bouts ship A needs to survive to ensure killing the others	Chaff
1	6	11	17
2	5	4-5	13
3	4	2-3	11
4	3	1	6

For Ophiuchus scenario of robos with 2x laser-4 and 2x zortrium (2 shots per ship, 5 hits to die), with 11 chaff ships you maximise the chances of mutual annihilation, although there are slim chances of one or the other prevailing due to overkill shots. With 10 chaff mutual annihilation is still the most probable outcome but A begins to have less chances of survival than the others. And vice versa with chaff 12.

So chaff requirements goes up a lot with low proportions of shots/hits per ship. In other words, if your enemies abuse chaff, you abuse armour.
And that is without considering PP costs.

Oberlus wrote: ↑Sat Nov 23, 2019 4:42 pm if your enemies abuse chaff, you abuse armour.

That conclusion is not supported by the data. Your data only shows that it is a bad idea for the chaff-player to mirror the low shots, high armor ship designs.

Morlic wrote: ↑Sat Nov 23, 2019 6:06 pm

Oberlus wrote: ↑Sat Nov 23, 2019 4:42 pm if your enemies abuse chaff, you abuse armour.

That conclusion is not supported by the data. Your data only shows that it is a bad idea for the chaff-player to mirror the low shots, high armor ship designs.

You are so right:

If chaff-fleet ship is 4 shots, 3 hits, against two ships of 1 shot, 6 hit, it needs to survive 3-4 rounds and needs 6 chaff. I get also 6 chaff needed vs 2 shots, 5 hits. And I haven't made the calculations but now I pressume the chaff requirements are constant for a fixed shots/hit ratio of the ships of the chaff fleet, and so the chaff player must spam weapons, and put armour only to ensure structure enough to survive 2 or 3 shots.

Another real-life (or at least from the 6th multi-player game) report:

Beating up on a maintenance ship (MD 3 damage), versus 6 of mine: 2 BSH w/ SAP, 1 BMH w/SAP and 3 fighters, and 3 robotic hulls (1 laser 6 damage, and 3 fighters). So, 2 chaff to 4 warships.

The eventual outcome has never been in doubt, but so far only 1 hit on any of the robos, and only 1 on the mini-carrier. Everything else has been soaked by the 2 empty BSH or fighters. One BSH is down to 1 structure left, and the maintenance ship is down from 425 to 45 in three turns.

Chaff which can withstand one hit is more effective than chaff which dies on the first hit. Armor is cheap. Maybe not as good price/damage ablated as comsats, but an armored but unarmed BSH still pretty cheap, and mobile. I find it a good mix with carriers, which are susceptible to being killed before they can launch all their fighters.

Caveat: i haven't had much chance to test this against Real Players

J.

swaq wrote: ↑Fri Nov 22, 2019 11:18 pm Anyway, here are my quick results:

== 9 chaff ==
- Red side wins with 14 hp remaining on one ship
- Draw

== 10 chaff ==
- Red side wins with 14 hp remaining on one ship
- Red side wins with 25 hp remaining on one ship

Oberlus wrote: ↑Sat Nov 23, 2019 4:42 pm For Ophiuchus scenario of robos with 2x laser-4 and 2x zortrium (2 shots per ship, 5 hits to die), with 11 chaff ships you maximise the chances of mutual annihilation,

So it looks like my spreadsheet is overvalueing the chaff. Anyway this looks like a good balance at this data point. The robo ship costs about 110 PP and and 11 small ships as chaff cost the same (but much more maintenance) - so the red team is still more cost effective which is sensible. 11 comsats cost only 66 PP but are a lot more restricted - i often scrap my comsats if the enemy does not attack as i predicted.

For the original question of this thread: Could you try just varying the number ships of the red team (e.g. 1 robos, 2 robos, 3 robos) and look how many bouts and chaff are needed in the blue team to see how the number of chaff scales the bouts.