Some small tests: Testing on Elemental Damage Coefficient: Spoiler: Click Let’s say I do “A” damage on an attack with no element added to it whatsoever. This damage also has no piercing attack involved with it. Now I want to know what I can gain from elemental damage. I know that: But I don’t know R, or even C for that matter. I want to find these values for myself! I then give my weapon an element, and with my elemental damage of E1, attack the mob for an improvement of I1. I add on more element, E2, and now hit for an improvement of I2 more damage. Now, I have two equations, and two unknowns (R and C), and thus can solve for them: Since I can use an element-neutral attack (named.... Shadow Cut) for a baseline damage, then enchant with Shadow Element and Black Web, and vary that a bit with other Necromancer skills and free skill resets, it's pretty easy to show element stacking's linear effects, and solve for these values. I do this on a level 70 Reinforcement Machine in Training Mode. This seems to point towards the elemental coefficient being 222, rather than 200 from ye olde DFO testing. It also implies that the Shadow Resistance of the Reinforcement Machine is -11. This kind of sounds odd at first (I'd expect that it would be zero), but you can observe a similar effect by using either a Salamander Flask (or maybe a Jack Frost Flask) on a lowbie character and seeing the damage dealt change by a fair amount. Some Fiddling Around in Training Mode: Spoiler: Click So, I broke down Shadow Cut with no element, and compared it to what I'd expect on a 0 DEF mob. Looks like it has no defense or similar reductions. I started using Vallacre for testing, and compared the first basic swing with these two wands: So I started varying piercing, and later shadow elemental attack, on various targets in practice mode. Here were some findings: Some observations: There seems to be a 20% reduction of piercing damage, at least in Training Mode, compared to what would normally be expected just from multiplying piercing attack by percentage. Since these tests were going for breadth, rather than any real depth, the found shadow resistances are a bit inconsistent. This is because the small difference in improvement percentages (due to variance on these single instances) have a huge effect on calculated resistances.