So knowing already that SCV’s repair rates are unfortunately impacted by erroneous production time value, thus directly affecting their repair rate for Raynor. I decided to further test this with Mengsk’s Dominion Labourer.
Test 1 - DL repair rate.
As expected, the rate is exactly the same as Raynor’s SCV or MULE. It repaired a 2HP level 3 Pride of Augustgrad to full in 90sec (as is PoA’s production time displayed).
Since Level 3 PoA has 1400 HP, ~70DPS (varies again due to in-game mechanic), and 6 total armor. I was able to determine a few thing, but this in particular.
1400 HP / 90sec = 15.6 HP/s of repair rate<== becomes important later, > 12HP/s medic’s beam rate
Test 2 - PoA w/ SCV vs PoA w/ Medic
So this Level 3 PoA dealing ~70DPS to an 0-armor dummy while Double Edge is active should (by calculation) take 1400HP / 70DPS = ~20sec to die.
PoA w/ SCV: While being repaired, it took approximately 30sec before I had to jump out of there without PoA dying.
1400HP / 30sec = ~46.7DPS
Recall the 15.6HP/s repair rate from earlier, 15.6 + 46.7 + 8HP/s (DE healing) = ~70.3DPS<== we get the averaged DPS of PoA back.
PoA w/ Medic: While being healed instead, it took approximately 38sec before jumping out again without dying.
1400HP / 38sec = ~36.8DPS
Medic’s healing is only 12HP/s, again +DE’s 8HP/s
So it seems Medic's Stabilizer's 25% damage reduction does work against DE
36.8DPS / 0.75 + 12HP/s + 8HP/s = 49.1 + 12 + 8 = ~69.1DPS<=== again, getting back to that approximate 70DPS
I am no expert on the exact equation (deriving them backwards from test results). Though it seems to be that the 6 total armor played little to no role in terms of Double Edge’s self-damage. This damage seemed to be directly determined by the amount of damage dealt to target (and that value taken, then dealt back to unit regardless of said unit’s own armor values). This is different than the target having armor, which therefore affect the amount of damage dealt (thus affecting the self-damage after the fact).
Overall, interesting tests and results for me. As I was wondering if that 25% damage reduction does affect DE or not. As sc2coop's mutator details didn't show this particular situation while showing Conservator's field damage reduction (similar). If my math is off, please do provide some details and corrections.
If this is done at P3 (which I presume this isn’t), then at least in my testing the BC’s gas cost is correct to its P3 reduced cost (ie. 25% of 168gas, from 4 to 550HP).
These prices seem to match that of non-P3 (for example, the 240gas x 25% of 60). Actually on second look, seems like you did yours on P2, which doesn’t have the 20% cost reduction either.
Which is a little over reported but not much, could be observation error. Maybe need to let the PoA actually die. I’m assuming you had multiple medics to heal if one ran out of energy or come up with another solution?
Mmmm, thanks, but I’m pretty sure the 25% damage reduction (or 0.75 factor) is only applied to the damage (hence the observed DPS of 36.8). As the 12HP/s is a heal rate not affected by 25% damage reduction, similarly as is the 8HP/s of DE recompensed rate.
No, you don’t need more than one (as long as it has full energy). Remember that the rate is consumed at 12HP/s per ~2.5energy/s (tested). That meant in the 38sec x 2.5energy/s = ~95energy consumed before PoA dies. And NOT 1400HP / 12HP/s x 2.5energy/s = 292energy to fully heal PoA.
Tested it on MM (for P0 and P2, to see if it’s different to P3) a BC at 2 hp indeed cost 100/60 for P0, and 100/75 for P2. So, even sillier; while the build time increase doesn’t apply to SCVs, apparently the cost reduction abilities and prestige changes to cost do apply to repair cost.
Yes, I was lazy as I was simply working backwards. Reason being that it shows the “observed DPS” from 1400 / 38sec.
You mean that it does. As your data show P0, P2, and P3 with their corresponding repair cost, each at their appropriate 25% of unit cost (affects at various levels of cost reduction).
Like I said before, the production time value being wrong is clearly a bug that should be rectified (especially given how the cost side of it is working).
Maybe 4-8 OC if you are macroing well, if you are going for BC you might want extra mule drops for repairs for example. If you don’t macro well, you are going to be very inefficient on both income/army when it is all spent on OCs that have to pay themselves off. I only go for 3.
If you had so many OCs like 8+, you could use that extra production and minerals on workers, with a protoss ally maybe using robotics heavy unit comps, you could flood workers to soak damage and repair in front of your medics. It might actually make the map go faster but I feel it is better spent in army in any case.
Maybe missed some of the context depending on where you entered this conversation so here are some dot points:
conversation is about healing vehicles with medics or SCVs
medics take up front investment but pay themselves off over time
SCVs and Mules have less up front cost but incur ongoing usage costs
In my opinion both are useful. Mules will be good on shorter maps or if it’s difficult to keep Medics alive. Medics better on longer maps if there is a lot of attrition due to mutators (double edge etc). Building some Firebats or grounded Viking’s to tank is a strategy I plan to try the next time I get Double Edge as Raynor P3.
