which is better for endgame?
the values of the pen affix on amulet seem too low to be better than the %inc spell damage. i think itās best paired with builds that have a lot of resistance shred. itās a pretty minimal increase, but if youāre capable of reaching the res shred cap (20 stacks of 2% vs bosses, 5% vs regular mobs) then pen is the only way of driving enemy resistance lower than that.
iām still not clear on how player penetration is calculated. with 0.7.10 and the resistance change, monster penetration was designed to be proportionate (where 1% of pen = 1% increased damage taken) as opposed to the way it works in a traditional resistance/pen design where 1% pen = 4% inc damage taken. if player pen works the same as monster pen, then itās a fairly weak stat. if ours is more like the traditional system, then itās probably the better choice for amulet.
as far as I can tell on a target dummy 9% pen does more dmg but thereās alot of variability in the base spell dmg so itās not fully clear. i also have a complicated build with several different dmg buffs (flame ward, closed circuit and teleport) all give me dmg buffs so itās nearly impossible to factor all that in without a big math project and better understanding of the game.
yeah i think thatās the state of a lot of things in the game currently, unfortunately. until we have some in-game tooltip DPS (and probably some more fleshed out third-party sources like tunklab and LEtools) weāll just have to make these kinds of choices based on feel.
i think if you do some dummy testing without any of your buffs, youāll have a better idea of how much impact the stat has. even if there are some multiplicative effects of stacking your buffs with pen, it couldnāt hurt to get a baseline idea of the impact it has on damage.
but if pen gives me a % dmg buff ontop of other ones, and they multiply, doing the test without the buffs will give a bad result. EDIT: for example closed circuit is 15% dmg.
the idea with my suggestion would be to remove as many confounding variables as possible (the other buffs) and use the data you glean from testing using only penetration as a means to establish a baseline for the effect it gives. it isnāt to get a āgood resultā but rather to make the effect easier to identify.
there will be some effects that you canāt necessarily avoid (iirc closed circuit works if lightning blast has chains remaining, itās been a long time since i played with LB so i might be remembering wrong) but by removing the buffs you do have control over you have a clearer picture.
you could even get really meticulous and do your tests wearing only using a white weapon base and 2 different amulets of the same base type (ideally with an implicit that doesnāt affect your damage) with only either the pen affix or the spell damage affix. the passive and skill tree nodes will still be in effect of course, but they will be the same for both tests, so they can be mostly disregarded. this is turning into the ābig math projectā you referred to, but iām just trying to demonstrate more clearly what i was suggesting.
I would think that pen would almost always be much better. Increased damage will be additive to all other non-local skill based damage increases. So in a simplistic example, if you already have 1000% inc damage, an extra 100% dmg will only actually increase your damage by 10%. (Disregarding More modifiers) Pen should act as a second More modifier as it affects damage taken, not damage done. So if after all your buffs, you do 1000 dmg with 10% pen, youāll actually do 1100 dmg (10% More).
In the above example, 10% pen would be equal to 100% increased damage. The break-even point will differ for each build based on individual numbers. My numbers tell me to take pen every time you canā¦
I may be wrong - if so, someone please tell me!
3 Likes
thatās more along the lines of the way I was thinking about it lets see what others think!!
i think your assumption is correct, that itās treated as a āmoreā multiplier (any form of āincreased damage takenā is, as far as i understand it), so you might be correct that itās usually better than a larger ā%inc damageā mod. i think thatās always going to vary on a build-by-build basis though. your post may better answer OPās question as to whatās better for endgame, since it would be the last place an already-high-damage build could find a significant damage increase.
iād value some input from someone more mathematically literate than myself, and anyone who has a better grasp of whatās going on āunder the hoodā in LE.
Penetration is a multiplier on damage, theres no way its not better than additive damage
Target Dummy probably has no resistance anyway as Armor Shred does fuck all to it, that means Pen is even more valuable
The only way Pen can actually be worse is if your attack does multi elemental and your Pen is ālightningā, then Spell damage would boost all of the other damages assuming it was a spellā¦
Also going from 500% spell damage to 550% isnāt actually a 10% more multi either, probably about 3-4% because you have other additives going on, Pen just multiplies the final hit before it lands vs the mob res
@trefenwyd is correct, the incremental benefit you get from 1 more affix slot of increased damage depends on the sum of all of your increased damage. How much more modifiers you have are irrelevant as they are multiplicative and donāt change in this example.
Penetration is an additional more modifier as it affects how much damage the target takes rather than how much you do, frailty is good for a similar reason.
thatās more or less what i arrived at after working through my thoughts in the above posts. thanks for the clarification!
do you know if the value of pen is additive with the value of resistance shred as being the āfinal stepā in the calculation? seems like it would be. (edit: assuming shred reduces enemy res to a negative value here)
Yes, shred can (& generally does) reduce resist to zero and I would assume that pen is additive with shred.
Given that the vast majority of mobs donāt have resist, what would be the point of shred if it didnāt reduce resists to negative figures?
i think you mistook what i wrote. i wasnāt asking if it was possible for shred to do that, merely pointing out the assumption i was making in my previous sentence. i know most mobs donāt have resists unless from mods currently, but that doesnāt mean there wonāt be more resistant mobs in the future. iām just trying to make sure i have an understanding of the fundamentals.
i was mostly asking out of interest for the build iām working on currently. iām playing harvest lich and have options for necrotic shred (harvest skill tree) and necrotic pen (lich passives) so i just wanted to make sure my understanding of it was correct before making decisions about the build. 
i had a slight doubt that having 100% shred chance on harvest would devalue the 30% pen i can get from the tree. if they were multiplicative with each other, i would reconsider my current strategy of investing into both since there would be diminishing returns and i could put points into other things instead.
Ah.
It kinda does, but it depends how much you can get, plus the pen would affect non-Harvest damage as well. Also, what else would you be giving up to get the pen from the passives & would that be equivalent to a bigger more multiplier (details would help)?
as far as i can tell, thereās not much (or anything, really) in the passive tree that would give more damage than the pen node. iām level 60ish with the build now so the skill points are starting to slow down and iām having to be a little more judicious about where i put the points. originally i was planning to go low-life with exsanguinous but i havenāt really invested in ward in the tree yet. so if it were the case that having both the pen and the shred was not as good a choice as i thought they were, it would make it easier to make some cuts to those in favor of ward. as it stands iāll just slowly make my way to ward viability.
i think iāll do a write up on the build once i get the monolith done with it. i have no interest in arena pushing so i dunno if itāll be viable there but itās a fun build and seems pretty tanky. completely deathless so far. (i think itās only the first or second time iāve got a build through the campaign without at least one death. i might have done it with the void hammer build i played a few patches ago but i canāt remember.)
Personally I run no ward passives but I have flame ward, im lvl 85 and beat all monoliths and empowered monoliths. I have health % on my chest suffix along with 12 vitality on the prefix. I use unread scroll instead of aegis, the 6% base crit is way better than an aegis if you have a little lifeleach with any build that benefits alot from crit. I have 5% lifeleach (gloves and 1 ring). after that I just get good resists and 100% crit avoidance (40% from a blessing and 2 items with t5) so I donāt get one shot. My resists are 138/131/121/110/70/113/125. I think ward is way over-rated since the nerf. Flame ward however is extremely good and I have 19% cdr so I can use it more often. The aura and added dmg is good too. 5% cdr from passive and 14% on helm. Iāll probably be doing a video on my build soon. I do run 9% cold pen on my ammy but iāve been thinking about swapping it for crit multiplier or spell dmg, Itās hard to tell if itās helping as much as it used to end game.
sorry for the confusion, i was getting pretty off topic and talking about my current harvest lich build. i did play a spellblade (based on llamaās guide) recently and also chose to go for flame ward. itās a very good skill. i donāt think i invested much in ward retention on that build, but i did pick up some ward on hit. i havenāt played the build since i finished the first level 100 timeline with it, so itās still around level 75, but it was definitely one of the hardest hitting and hard to kill builds iāve tried. i think ward is a great defensive layer, and it was made stronger by the change to resistance. (protection was much less useful for ward builds because you could stack ward so much higher than life that you needed ridiculous protection values to get any noticeable damage reduction from it)
it sounds like youāre going for the shatter strike approach, which i tried out but it just felt worse than flame reave/firebrand. cold pen and freeze multi are probably very necessary to make that feel good, but i didnāt get far enough with it to be sure. i put it down around level 45, but i might go back to it at some point.
Iām playing a glacier build thatās different then any build Iāve seen yet. Iāll probably make a video on it like I said. Itās much different than the ones Iāve seen.
Assume 1000 damage on a mob with 75% resist.
Your base damage: 1000
Monster base resist: 75%
1000 - (1000 x .75) = 250 damage taken
10% penetration
1000 - (1000 x .65) = 350 damage taken
Mob takes 100 more damage if you have 10% penetration.
10% spell damage
1100 - (1100 x .75) = 275 damage taken
Mob takes 25 more damage if you have 10% spell damage added.
Pound for pound 1% penetration is 4x worth 1% increased damage. The breakpoint where increased damage outperforms penetration is 4%. So if you have to choose between 1% penetration or 5% spell damage, choose the spell damage.
Math could be wrong so someone correct me if Iām wrong. The assumed 75% resist can be any number. The key is to have it be uniform so you can plug in 0% resist if you want (-10% resist with 10% penetration).