[3.24] Ice Dancing Queen - Icestorm/Cyclone/CI - Viable For Everything

"

King_Haart wrote:

I would think an ES+EV chest would be better due to our Ascendancy giving us ES for EV on our chest, so even 300/1500 would translate to more overall ES, *plus* higher evasion.

i was thinking which base i would buy. Since we dont stack evasion and not running grace i got this armour. Its hard enough to find a good endgame armour ^^

"

King_Haart wrote:

You would also want a Hunter helm with 10% Int.

yea, didnt think of it. Thanks!

"

King_Haart wrote:

Your characters aren't public on your profile so I can't see the Jewels.

i edited my post and show my jewels, but i thought my profile was public.

"

King_Haart wrote:

EDIT:

Just came across this guy in heist, this might've been worth something in another league but probably not this one:

i guess my rings are better, like you said because of the lake thing

TL:DR:
-----------------------------

So, to summarize:


If you have 2 good affixes and want to minimize risk, craft nothing. You win 1/2 of the time.

If you want to push for 3 positive, good affixes or more:

If you have 2 good affixes, craft another. You win 1/4 of the time.
If you have 3 good affixes, add nothing. You win 1/4 of the time.
If you have 4 good affixes, craft another. You win 1/4 of the time.
If you have 5 good affixes, add nothing. You win 1/4 of the time.
If you have 6 good affixes, annul one. You win 1/4 of the time.


------------------------------

I took a look at the public profiles on poe.ninja, looking for what were clearly crafted/reflected rings and amulets. I assumed the really good stuff would be used rather than put up on trade.

Of the non-crafted ones, there are some 4/0, 4/1, and one 5/0 rings (number of +/number of negatives), as well as some 2/3, and a lot of 2/2 and 3/2.

Of the crafted ones, I see one 5/0 amulet, one 4/1 amulet, and a bunch of 3/2's.

If my analysis is correct, the probability of 5/0 is 6.25% and for 4/1 is 31.25%. So you'd expect to see more 4/1's than are present.

---------------

I think the error in my previous analysis is that I always need two specific affixes to be positive. That will happen 50% of the time (since either ++ or -- works, but +- or -+ do not.) So the probability is a bit different than the table above (5/0 is still correct at 6.25%, but 4/1 is 15.625% instead as half of the 4/1 cases are bricked.)

I'll go modify the counting program and see if my comments in the preceding are actually correct and post a corrected table soon.

----------------

To make it very clear, "bricking" means I didn't get +int +all atributes, or +int +resist, for example. +int -all attributes is clearly useless as the mirror is -int +all attributes, making both choices bad.

----------------

OK, here are the corrected results, assuming you require 2 specific affixes to be positive:

2/0: 0.5

2/1: 0.25
3/0: 0.25

4/0: 0.125
3/1: 0.25
2/2: 0.125

5/0: 0.0625
4/1: 0.1875
3/2: 0.1875
2/3: 0.0625

6/0: 0.03125
5/1: 0.125
4/2: 0.1875
3/3: 0.125
2/4: 0.03125

n/m means you start out with n+m good affixes and end up with n positive and m negative affixes.

----------------------------

If you have N good affixes, should you craft another good affix? The analysis already shows that the number of bad affixes is irrelevant.

Let's assume affixes have the same weight. So 2 good and 2 bad (2/2) is worth 0. Recast the above:

2:
2 net: 0.5

3:
1 net: 0.25
3 net: 0.25

4:
4 net 0.125
2 net: 0.25
0 net: 0.125

5:
5 net: 0.0625
3 net: 0.1875
1 net: 0.1875
-1 net: 0.0625

6:
6 net: 0.03125
4 net: 0.125
2 net: 0.1875
0 net: 0.125
-2 net: 0.03125


So, how do I maximize the probability of getting 3 or more net affixes? (Note you can stick with the 50% chance of getting 2 starting with only 2 good affixes.)

Start with 2, add a good, 0.25 chance.
Start with 3, add nothing, 0.25 chance.
Start with 3, add a good, 0.125 chance (for 4.)
Start with 4, add nothing, 0.125 chance.
Start with 4, add a good, 0.25 chance.
Start with 5, add nothing, 0.25 chance.
Start with 5, add a good, 0.15625.
Start with 6, 0.15625.

So I made an error crafting a good affix on my 5 good affix amulet on the first try (I ended up with 0 net.) Good to know.

The above table also explains why 4/0 and 5/0 items are rare. There aren't that many 4 or 5 good affix items being reflected, first of all. And of these, only 1/8 of the 4's and 1/16 of the 5's win.

Hi,

have this helmet

should I remove the craft and slam a Crusader's ex on it?

"

Snowbliz wrote:

Hi,

have this helmet

should I remove the craft and slam a Crusader's ex on it?

Depends. What's the helm's ilvl?

Just ensure there are 3 prefixes before doing any slamming.

"

Graiaule wrote:

"

TheChosenOne_ wrote:

Can you or anyone else look at my build and see what needs to improve to get it more tanky? I only got 4 divines, 146c and like 2 divines worth of stuff to sell I think

https://poe.ninja/pob/HO3

Basically replace every single item except your boots.

In more or less priority order, and very little crafting RNG:

1. Get a sadist garb with ~350ES/~1400+ evasion, 50+ int if possible and 1 free suffix, and craft 6% increased attrs (search for "of Hillock's veil to get the craft.)

2. Get an ilvl 75-79 Hubris circlet with 43+ int and a free suffix and slam it with a hunter orb. Make sure it has no free prefixes before you do that.

3. Buy a 15% incr int belt with no quality (1 div), scour, 4x intrinsic, and slam with deafening spite until you get a good resist and/or ES.

4. Tranquility needs to be on your boots not your amulet, as stampede is the BiS (so it won't be replaced.)

5. Look for 140+ int mirrored rings with either a good all resist or a good single resist more than 87%.

6. Get a glove with 51+ int and the resists you need. Craft 3% dam per 100 int and 15% chance to unnerve with eldritch orbs. That'll cost you less than 100c unless you're unlucky.

7. Keep an eye out for 18% staffs. If they remain too expensive, buy the 4 div cards instead and keep turning in until you get two 18% ones.

8. Flasks should be focused on rare/unique enemies except the bleed flask.

In the above process you want to get 87% to each elemental resist. So there'll be some tweaking. If you can get 87% on both rings, it's really easy to get 87% on your other gear.

You've got a lot of work ahead of you. Check out other posters' gear also for ideas.

----------------------------------

NOTE: To make it VERY clear, that's just the FIRST round of upgrades. Later you'll upgrade your chest, and likely rings and amulet and gloves and maybe even helm.

-----------------------------------

I'm going to start crafting with reflection of Kalandra to see if I can upgrade my amulet and rings. About 3 divine a try, so seems worthwhile. I seem to have maxed out what I can purchase in trade (not really seeing any upgrades there, and I'm just going to avoid the 10%/14%/1% per 15 int amulet for now.)

[See next post for the detailed analysis of using Kalandra's Reflection.]

thanks for posting this. also we cant run elemental reflect? i dont know whats killing me so easily in a un ided map. maybe its elemental reflect

"

TheChosenOne_ wrote:

also we cant run elemental reflect? i dont know whats killing me so easily in a un ided map

We can but we need protection against it, e.g., from a Watcher's Eye and the Pantheon and probably one other place I forgot where. I had 95% protection the previous league.

"

Snowbliz wrote:

Hi,

have this helmet

should I remove the craft and slam a Crusader's ex on it?

NO! NEVER!

%int in helmets is Hunter influence mod. You need Hunter's orb for that.
Crusader is for chest and amulets

If you don't have helmet with %int I would Hunter it, especially if the helmet is in ilvl75-79 range.

thx for the answer, I meant Hunter's exalted, lol

The Helmet is ilvl 85, do you think I should try?

Tried on several helmets 75-79 Deafening Essence of Spite but never had good mods.

I dont find the Deafening Essence of Spite in my lake or maps:(

For ele reflect I first ran with a ring for a bit (easy to get >50% on that and the other 50% from Yugul in Pantheon), but then switched to Watcher's Eye as I found one with 50% reflect mitigation plus 6% spell block for a pretty cheap price.

At some point I might upgrade to a different WE for bossing/no-leech cases, but will probably keep the other one in my inventory just in case.

Took heart of Ice as per the leveling tree and went to see where my next point should go and the only other trees are completely different

feeling defeated

I dont understand, halp?

here is the tree that was recommended that seems to dead end after heart of ice: https://pastebin.com/PWizzxua