[3.6] Acid Molten Strike Assassin - Uber Elder-viable Poison Build

"

magicrectangle เขียน:

You're wrong about GMP's damage penalty also. It doesn't apply to poison. That's why it is a sensible gem choice for poison molten strike, but not for direct damage molten strike.

When you use GMP, you can't use another gem, like VT. Hence the immense damage loss per projectile. VT brings with it about 50% more damage. All GMP does with respect to this build is add 4 more projectiles, nothing more. No damage loss, no gain (outside of the projectiles). Not even sure why you're arguing. Even if you hit with all the extra magma projectiles (which means a 40% increase), each one is still down 50%. Even straight math you lose. The only reason to run it is if your damage is already high enough for trash that having a larger spread means better clears. For boss fights, there is never a reason.

Do you get it? 10x150=1.5k vs 14x100=1.4k. Even with your BS math on the projectiles hitting, you're always at a loss. Ya you have more, they all hit for less.

How are you not getting this?

"

_Tiem เขียน:

When you use GMP, you can't use another gem, like VT. Hence the immense damage loss per projectile. VT brings with it about 50% more damage. All GMP does with respect to this build is add 4 more projectiles, nothing more. No damage loss, no gain (outside of the projectiles).

Fair enough, I misunderstood you there.

"

Not even sure why you're arguing. Even if you hit with all the extra magma projectiles (which means a 40% increase), each one is still down 50%. Even straight math you lose. The only reason to run it is if your damage is already high enough for trash that having a larger spread means better clears. For boss fights, there is never a reason.

I've never claimed GMP was better than VT, so you're barking up the wrong tree there. That said, you've overlooked some advantages of GMP here.

GMP gives more life gain on hit, it gives faster stacking of duration for noxious strikes, and it gives better straight up damage if you don't have the hat enchant yet. All are reasonable justifications for somebody to choose to use the gem, given their particular playing circumstances.

I haven't used it myself since I got the hat enchant.

"

Do you get it? 10x150=1.5k vs 14x100=1.4k. Even with your BS math on the projectiles hitting, you're always at a loss. Ya you have more, they all hit for less.

How are you not getting this?

You've shifted the conversation away from molten strike mechanics without either explaining what's wrong with my mechanics explanation (beyond calling it "BS"), or acknowledging that it is correct. I'd like you to do one or the other, because third parties reading this thread deserve to know how the game actually works, rather than wondering to themselves which of us was correct.

Here's the wiki entry on molten strike:
https://pathofexile.gamepedia.com/Molten_Strike

It mentions the melee deadzone, but nothing about a ball-vs-ball deadzone. The wiki references a few dev posts about molten strike, none of which say anything about a ball-vs-ball deadzone either. I also didn't find anything about any such thing from googling it.

mark_ggg says projectiles are fired a random distance - though that description is admittedly incomplete since it says distance traveled is random, but doesn't say how angle of fire is chosen

Now I suppose it is possible for the angle to be determined in some other-than-random way, even though the range is determined randomly, but that would be pretty odd. Certainly a randomly determined range can't have a deadzone. It could have a dead angle, but there's no evidence of that, that I could find.

Given the lack of evidence for anything else, a random angle should probably be assumed. That being the case, we have a random (but not uniform in density) ball distribution within the firing area (which is 5-20 units, barring AoE modifiers).

This method of calculating ball distribution has also been assumed by the maker of the molten strike dps notebook, which is considered by most people in the community to be authoritative.

Given a random distribution, the expected value of the number of balls inside a given hitbox is directly proportional to the total number of balls fired. Even though the density of balls is not uniform (increasing the closer to the source you get, so long as you don't go inside the deadzone), there's no hidden dependence of radius (and therefor density) on ball count, so there's no skew to the expected value there. In other words, how close the boss is to your melee hit does not depend on ball count.

What percentage of the target area is covered by the boss's hitbox is irrelevant in this proportionality. Obviously if you make the boss very large, and all projectiles hit, it is trivially true. But you can scale the hitbox to any size or shape within the target area, and adding 40% more balls will still cause 40% more balls to hit the boss, on the average.

"

babsbegood เขียน:

Hi, can someone please tell me if this claw is better than wasp nest, im not so sure because of the low attack speed:

Btw thanks for the guide.

It is. I've also added something in the weapon section so you should be able to compare it yourself now.

"

_Tiem เขียน:

Each projectile has a deadzone to minimize overlap and abusing projectile stacking. You can fill up that 270 degree radius but each projectile will simply occupy a free zone to make the spread larger but still remain even.

You can very easily disprove this ingame.

Both of these screenshots were taken with three projectiles, on one attack they have maximum spread, on the other all three of them overlap.

The spread is completely random, however, the more projectiles you get, the less likely it is to observe the edge case of every single projectile overlapping and the more even it will appear.

"

magicrectangle เขียน:

GMP gives more life gain on hit, it gives faster stacking of duration for noxious strikes, and it gives better straight up damage if you don't have the hat enchant yet. All are reasonable justifications for somebody to choose to use the gem, given their particular playing circumstances.

Factoring in the duration benefit, GMP's damage bonus actually comes close to 96%.

Even assuming you don't need that duration, comparing it to VT is still 10% DPS vs 40% life recovery which makes it worth considering even if you have enchantment.

But yeah, the lag.

I understand why wither is being used but have you looked at running despair with the totem?

If you really want despair run a witchfire flask.

"

SHEPUUURD เขียน:

The spread is completely random, however, the more projectiles you get, the less likely it is to observe the edge case of every single projectile overlapping and the more even it will appear.

I didn't say they never overlap, but that their chance too wasn't great (or diminished thanks to their dead zones). A better and more robust test would be how many times it took you to attack to see that kind of overlap. Extreme edge cases don't provide much support to run GMP. So you got them all overlapping that one time...

In any case, I still can't see much of an argument against always having stronger projectiles and way less lag.

On another note, why didn't you take Perfect Agony? It seems like one of the most valuable keystones for a build like this and while 7 nodes away, you can double down jewel socket on the way and opt out of the one above Blood Drinker, bringing the cost down to 4. You can also opt out of Sleight of Hand which doesn't bring much of anything (albeit strong for 2 nodes) which drives the cost down to a meager 2 points. On my POB, it pumps up the poison damage by 18% (then subtract 4% from Sleight of Hand), well worth it.

https://pastebin.com/qcjuK83P

Incidentally, I'm up to 4x your initial builds damage now (without Wither). With Wither, I can almost touch 400k.

"

NoxXxy เขียน:

I understand why wither is being used but have you looked at running despair with the totem?

Too slow. Wither is for boss fights and brings way more damage with it at max stacks (nearly doubles your damage). Despair brings a modest 15% more damage, not much. Given that dropping a Spell Totem takes a bit of time, it would be arduous to always have to drop one at every pack. If you need help clearing trash at any level, you need better gear. You should never need to pump resources into clearing trash mobs once a build is dialed in. It becomes all about the boss fights.

"Incidentally, I'm up to 4x your initial builds damage now (without Wither)."

Because, again, you don't seem to understand your own assumptions or understand the build as well as Shepuurd.

You're not worth arguing with, but for any one else trying the build please ignore _Tiem's PoB's. Just follow the main build guide. There are really no strictly better improvements to make other than getting an amulet with the non-chaos added as chaos mod. Just keep in mind that to do so AND keep your boot slot resist-free for uber elder means you'll need like a 20EX amulet in all likelihood.

"

grandanes เขียน:

"Incidentally, I'm up to 4x your initial builds damage now (without Wither)."

Because, again, you don't seem to understand your own assumptions or understand the build as well as Shepuurd.

You're not worth arguing with, but for any one else trying the build please ignore _Tiem's PoB's. Just follow the main build guide. There are really no strictly better improvements to make other than getting an amulet with the non-chaos added as chaos mod. Just keep in mind that to do so AND keep your boot slot resist-free for uber elder means you'll need like a 20EX amulet in all likelihood.

Well haters gone hate what can you do.

"

_Tiem เขียน:

"

SHEPUUURD เขียน:

The spread is completely random, however, the more projectiles you get, the less likely it is to observe the edge case of every single projectile overlapping and the more even it will appear.

On another note, why didn't you take Perfect Agony? It seems like one of the most valuable keystones for a build like this and while 7 nodes away, you can double down jewel socket on the way and opt out of the one above Blood Drinker, bringing the cost down to 4. You can also opt out of Sleight of Hand which doesn't bring much of anything (albeit strong for 2 nodes) which drives the cost down to a meager 2 points. On my POB, it pumps up the poison damage by 18% (then subtract 4% from Sleight of Hand), well worth it.

It's already granted by Coralito's Signature.

If you pretended that flask didn't exist, it's actually still not a good cluster. For most builds, 30% effectiveness of crit multi is pretty bad, since you still have to invest into it.
Assassin is an exception here because he does more poison damage on crit and gets a lot of free crit chance awell.

Using your pastebin, Perfect Agony is only 2.4% damage per point (2.8% with diamond flask)
Let's assume that jewel is a 2-point investment, which you'd take in that case. You could subtract one skill point from the travel cost.
In that case, it's still only 2.8% (3.3%).
Sleigh of Hand gives you a total of 3.9% damage, and grants the QoL of having attack speed.

"

_Tiem เขียน:

Incidentally, I'm up to 4x your initial builds damage now (without Wither). With Wither, I can almost touch 400k.

You also have 1/4 of my life recovery. Obviously I'm giving up damage by using LgoH rings.

My PoB is using a Wasp Nest an Thief's Torment, obviously it's easy to surpass.
That said, activating Frenzy Charges where you can't generate them and swapping Enfeeble for Temp. Chains doesn't mean you're dealing more damage.

Putting just the gems and configuration on even terms with my current character, you're dealing 4% more damage...
You can go higher if you get a fossil helmet with -chaos resist and a better claw.
Granted, my character doesn't have perfect gear either and is geared for taking on Uber Elder, not for dealing as much damage as possible. (and also not for Delving)