How Sacrifice REALLY works.

Arran

To all those like myself who thought the Aeon sacrifice system deposited 100% of the units mass into the build project, prepare to have your preconceptions shattered.
Here are the results of my meticulous sandbox testing. What you do with this knowledge is up to you.

General disclaimer. All test were conducted 2 or more times. Most only two times.

Tests using unupgraded SACUs (except for sacrifice). Note: Mass of sacrifice upgrade (150 mass) not included in calculations. Also HP values rounded down to the nearest hundred.
GC (27500 mass): 90000/99999 HP for 27300 mass of SACU (#14)
Paragon (250200 mass): 2100/5000 HP for 251550 mass of SACU (#129)
CZAR (45000 mass): 30700/40000 HP for 44850 mass of SACU (#23)
Tempest (22000 mass): 46200/60000 HP for 21450 mass of SACU (#11)
Salvation (202500 mass): 4700/10000 HP for 202800 mass of SACU (#104)

Tests for RAS SACUs (not preset) with sacrifice. Note: Mass of sacrifice upgrade (150 mass) not included in calculations. Also HP values rounded down to the nearest hundred.
GC (27500 mass): 25500/99999 HP for 25800 mass of SACU (#4)
Paragon (250200 mass): 600/5000 HP for 251550 mass of SACU (#39)
CZAR (45000 mass): 9300/40000 HP for 45150 mass of SACU (#7)
Tempest (22000 mass): 12500/60000 HP for 19350 mass of SACU (#3)
Salvation (202500 mass): 1400/10000 HP for 206400 mass of SACU (#32)

Tests for RAS preset SACUs with sacrifice. Note: Mass of sacrifice upgrade (150 mass) not included in calculations. Also HP values rounded down to the nearest hundred.
GC (27500 mass): 84400/99999 HP & 99999/99999 HP for 25800 mass & 32250 mass of SACUs respectively (#4 & #5)
Paragon (250200 mass): 2700/5000 HP for 251550 mass of SACU (#39)
CZAR (45000 mass): 36100/40000 HP for 45150 mass of SACU (#7)
Tempest (22000 mass): 47500/60000 HP & 60000/60000 HP for 19350 mass & 25800 mass of SACUs respectively (#3 & #4)
Salvation (202500 mass): 6200/10000 HP for 206400 mass of SACU (#32)

Tests using T3 engineers as sacrifices. Note: HP values rounded down to the nearest hundred.
GC (27500 mass): 24200/99999 HP for 27768 mass of T3 engineers (#89)
Paragon (250200 mass): 500/5000 HP for 250224 mass of T3 engineers (#802) = 202680 mass worth of reclaim from wrecks alone.
CZAR (45000 mass): 7400/40000 HP for 45240 mass of T3 engineers (#145)
Tempest (22000 mass): 11300/60000 HP for 22152 mass of T3 engineers (#71)
Salvation (202500 mass): /10000 HP for *** mass of T3 engineers (#) <- Test not conducted. Got bored.

My general conclusion. NEVER use the sacrifice system unless you literally have no other choice. Its better to ctrl+k and reclaim the wrecks.
Have fun out there.

angelofd347h

Sacrifice works on a ratio of mass/energy.

This table from ZLO explores the efficiency of using RAS boys.
For the most part RAS sacus are around 90% efficient.
Engineers (from memory) are usually around 60%.

ZLO

Engineers have 0.6 sacrifice efficiency multiplyer
sacus have 0.9
that means that thier maximum theoretical efficiency will be 0.6 and 0.9

However that is not everything...
idk how to explain here is simple example:
Assume sacu cost 100 mass and 100 power
if you sacrifice it on something that costs 1000 mass and 1000 power it will donate 90% of its resources
but if you sacrifice it on something that has different power / mass ratio you will have less efficiency.
E.G. for something that costs 90 mass and 180 power you will need 2 sacus witch means you lose about 45% of mass you have invested

Edit:
However if you sacrifice on something that costs 180 mass and 90 power then you will have 90% mass efficiency and 45% power efficiency witch is kinda acceptable.

So sacrificing RAS presets on GC or CZAR or even tempest is not that bad and can be done in emergency

Valki

Wouldn't it be more intuitive if sacrifice added buildpower to the engineer? Results would be similar but it relies on the players regular eco skill.

Let the engineer add their own build time to buildpower, suggestion: When sacrifice is activated the units buildpower is increased by 1/10th of its buildtime for 10 seconds.

Example: (T3 engineer costs 2100 buildtime.) A T3 engineer goes from 15 to 225 buildpower for 10 seconds. Then the engineer is killed and leaves a wreck.

arma473

If we did it that way @Valki you would get zero benefit from sacrificing while you are mass stalling or e stalling.

Even a tele-GC would take 10 seconds to build up AND you would need to have enough mass saved in the bank and enough energy (which would actually be a LOT of energy if you're trying to spend it all in 10 seconds)

Valki

@arma473 but it is much more useful throughout the game T1 to T4. If you overbuild engineers you can 'reclaim' the used buildpower (and later the mass through real reclaim)

Tele-GC is gone, but now a T3 engineer can spam up 10 T1 land factories in seconds for a proxy-LAB attack. It opens the door for many interesting proxy and localized rush options. Don't do that if that wrecks balance come to think of it.

arma473

@valki Not if there's 2 trees in the way or if the engie has to walk

Arran

The sacrifice system could use a splash (or flood) of simplification so new (or experienced players who aren't math geniuses) can figure out what is happening.

Katharsas

@zlo

The current mechanic makes it very hard to balance this for most use-cases i think. I suggest using fixed mass/power conversion rates to convert all mass/power of builder unit into target unit.

So lets just say we use (example values)

Definitly worse than T3 mass fab for power -> mass = 150:1
We use a sixth of that for other direction, so mass -> power = 1:25
(At this rate energy cost of yolona is twice as expensive as its mass cost, but of course this might still be too efficient)

So lets imagine a builder unit that costs 400 mass and no power.
And you want to sacrifice to build a unit that costs 400 mass und 10000 power.
So we need to convert some of the builder units mass worth into power worth.

Conversion rate for that is 1:25, so 200 mass = 5000 power. This means that instead being worth 400 mass, builder is worth 400-200=200 mass und 5000 power for this sacrifice, so you would need to sacrifice 2 builder units to satisfy both its mass and its energy cost.

With the old formula, the builder would have had 0% efficiency because it did not cost energy so it would not contribute a single point of mass or energy during sacrifice.

With my idea we could balance both conversion rates to make it usefull to sacrifice builders into both mass and energy expensive units/buildings. Of course the conversion rates need to be so bad that sacrificing does not replace actual economic buildings like mass fabs or power generators.

Arran

Even if conversion rate was highly efficient, I don't think people could spare the APM to manually sacrifice all game if they, for instance, wanted to not build power generators. Aeon is already micro intensive enough that it wouldn't be an issue.
That said, your idea @Katharsas is already far superior (IMO) to the current mechanic.

Arran

In FAF there are already balanced sources of Mass to Energy conversion (Mass Fabricators) and Energy to Mass conversion (Mass Extractors). Sacrifice could use these ratios from those structures (at say the T2 stage) to balance resource conversion per say.