[3.25] Ice Dancing Queen - Icestorm/Cyclone/CI - Viable For Everything
" 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 ^^ " yea, didnt think of it. Thanks! " i edited my post and show my jewels, but i thought my profile was public. " 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. Последняя редакция: Graiaule. Время: 20 сент. 2022 г., 20:41:00
| |
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. | |
" 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 Последняя редакция: TheChosenOne_. Время: 20 сент. 2022 г., 23:50:51
|
|
" 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. | |
" 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. Последняя редакция: Cjin. Время: 21 сент. 2022 г., 7:22:53
| |
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 Последняя редакция: snfflz. Время: 21 сент. 2022 г., 13:58:37
|