[3.11]QUEEN OF NECROMANCY - STRONGEST ZOMBIES EVER? 55M+ DPS 7K+ LIFE MAX BLOCK 90% CHAOS RES
Hmmm - Chayula breeches seem to have increased this league or else I'm just luckily running into them more often
Killed my 2nd Chayula in a breech just now. |
|
Let's calculate the average costs of crafting methods. I am really not sure this is all correct, so you are welcome to check it over.
I am using hankinsohl's results for probabilities of rolling each of the mods:
Spoiler
" Among prefixes: 1000/34175 ~= 1/34 for +1 gems Among suffixes: 2000/90225 ~= 1/45 for any maim (so half that, ~1/90 for lvl 20 maim) Relative costs today: 1c is 6 alt 1 annul is 23c 1ex is 185c craicic chimeral for magic imprint is exactly 1ex I am only counting the cost to get to a rare weapon with only the two desired mods. Indeed, final crafting bench cost (4ex for multimod and +2 support gems) is the same for every method. I assume the weapon is exactly ilvl 75. Method A: full alt spam, then beast imprint -> 4.76ex Average cost is 1540 alt + 3 craicic chimeral + 3 annul ~= 4.76ex today + chance to get a good suffix on regal (high str roll, attack speed for leap slam...). Then, you don't need to do the imprint step, though you loose the minion speed craft. 1. Alt spam until you have a blue weapon with both desired mods. Augment when you have one of them alone. Average cost: 1540 alts ~= 257c ~= 1.39ex (Using hankinsohl's results, to hit both mods you need 1540 alt on average for ilvl 75. Augments are neglected, as you only augment when you have one good mod so that's a few dozens at most.) 2. Imprint with beast, regal, annul. If you annul the regaled mod, you win. Otherwise, use imprint to go back to blue item and repeat. Average cost: 3 craicic chimeral (= 3ex) + 3 annul (= 69c ~= 0.37ex) (Each try costs 1 craicic chimeral and 1 annul and has 1/3 chance of success.) Method B: alt-regal-annul for rare with only maim, then mastercraft and slam +1 gems -> 6.24ex Average cost is 309 alt + 3.7 annul + 5.5ex ~= 6.24ex today - currently a lot more expensive on average than method A (but this can change if craicic chimeral and alt get more expensive but exalt doesn't) + more deterministic: you won't spend days without getting anything. The big step has 2/3 odds, so failing it many times in a row is much less probable than using many thousands of alt without getting anything + you are sure to get lvl 20 maim ASSUMPTION (not very important, as it only affects 1. which is the cheap part):
Spoiler
When you roll maim on a blue weapon, you have about 2/3 that the item also has a prefix, and 1/3 chance that it has no other mod. This is just a guess. It is approximately what I have observed. This would be true if a blue item had 1/2 to have a single mod and 1/2 to have two mods: this means 1/4 to have only a prefix, 1/4 to have only a suffix, and 1/2 to have both. Then among items with a suffix, 1/3 have only the suffix and 2/3 have both. 1a. Alt spam until you have maim lvl 20. Then regal and annul until you have a rare item with a single mod which is maim lvl 20. If you fail the annul, scour and start again. Average cost: 309 alt + 3.7 annul ~= 51.4c + 85.4c ~= 137c ~= 0.74ex
Spoiler
- 1/4 of the alt you use give just a prefix: this is useless, you just roll over it (you may want to augment or regal it if it is +1 gems to try and shortcut to step 2 of other methods). - The other 3/4 of the alt give an item with at a suffix (and maybe a prefix), which has 1/90 chance to be maim 20. So on average, you roll maim 20 every 90*4/3 = 120 alt. Every 120 alt, you get a "try". - 1/3 of the time, maim is the only mod on the blue item, so you pay a regal, an annul, and it has 1/2 odds. - 2/3 of the time, there is also a prefix, so after regal you must annul twice. In this case, you have 1/3 odds to fail the first annul, 1/3 to get the first annul right but fail the second one, and 1/3 to get both right. So 1/3 of these tries cost 1 annul, 2/3 cost 2 annul, and they have 1/3 odds to succeed. Combining both cases, a "try" costs on average 120 alt + (1/3 + 2/3 * (1/3 + 2*2/3)) annul ~= 120 alt + 1.44 annul, and has 1/3 * 1/2 + 2/3 * 1/3 = 7/18 odds to succeed, so average cost is 18/7 * (120 alt + 13/9 annul) ~= 309 alt + 3.7 annul ~= 51.4c + 85.4c ~= 137c ~= 0.74ex 1b. Same as 1a, except you only regal if maim 20 is the only mod on the blue item. Average cost: 720 alt + 2 annul = 120c + 46c = 166c -> worse than 1a
Spoiler
Odds to get a single suffix are assumed to be 1/4, so odds to get maim alone is around 90*4 = 360 alt. Then you use a regal and an annul and have 1/2 odds of success. Average cost is 2 * (360 alts + 1 annul). 2. Mastercraft "can have multiple crafted mods" (2ex) and "cannot roll attack mods" (1ex). Then slam an ex: you have 2/3 odds to get +1 gems and be done. If you miss, craft "suffixes cannot be changed" (2ex): your item is back to maim 20, "can have multiple crafted mods" and "cannot roll attack mods", so you can immediately try and slam again. Repeat until the slam succeeds. Average cost: 5.5ex
Spoiler
Initial setup is 3ex worth of crafts. Then first try is 1ex, each following try is another 3ex (2ex craft + slam), and each try has 2/3 odds. You have 1/3 odds to miss the first try and have to pay 3ex. You have 1/9 odds to miss two tries and have to pay the next 3ex, etc. Average cost is 3ex + 1ex + (1/3)*2ex + (1/9)*2ex + (1/27)*2ex + ... = 4ex + 3ex*(1/3 + 1/9 + ...) = 4ex + 3ex*(1/(3*(1-1/3))) = 4ex + 3ex*(1/2) = 5.5ex. Method C: alt spam for one good mod, try to hit the second mod on regal -> 3.56ex? Average cost: ~3700 alt + 1.8 annul = ~617c + 41.4c = ~3.56ex (I have ignored the cost of regal, but that shouldn't be more than 30c as they're really cheap.) I am really not sure about my reasoning for this one. I could be completely wrong. And this relies on the assumption above about odds for 1-mod vs 2-mod magic item, which is only a guess. Anyway, this is provided you always regal when you have one good mod, even if there is also a bad mod (in which case you annul if you hit the regal). Deciding to only regal single-mod items seem much worse: ~6167 alt = ~5.56ex if my reasoning is correct.
Spoiler
Sorry, this is a draft rather than an explanation, but here it is if someone still wants to have a look at it. Method C.a: regal only if you have one good mod alone, no bad mod with it. Let wm the weight of maim (any level), wg the weight of +1 gems, wp the total weight of prefixes, ws the total weight of suffixes. wg = 1000, wm = 2000, wp = 34175, ws = 90225 Use the same assumption as above: 1/4 of alt give only prefix, 1/4 only suffix, 1/2 both. Odds alt +1 gems then regal maim: 1/4 * wg/wp * wm/(wp+ws) Odds alt maim then regal +1 gems: 1/4 * wm/ws * wg/(wp+ws) Total if only regaling single-mod items: 1/4 * wg*wm * (1/wp+1/ws) / (wp+ws) x = 1/4 *1000*2000*(1/34175+1/90225)/(34175+90225) 0.0001621565852904113 1/x 6166.87875 alt ~= 1028c ~= 5.56ex Method C.b: always regal when you have one good mod, even if there is also a bad mod. Get the first right mod three times as often as previously. Indeed, when you see e.g. +1 gems on a blue item, there are 1/3 odds for it to be 1-mod and 2/3 for it to be 2-mod. Odds for regal are the about the same. But if you get the right mods this way, you must use an annul with 1/3 odds of success. Get a "try" every ~6167/3 alt. There are 1/3 odds you regaled from a 1-mod item and so you have finished, and 2/3 odds you have to annul with 1/3 odds, so 1/3 + 2/3 * 1/3 = 5/9 odds to be done. Also, 2/3 of these "tries" cost 1 annul. Average cost: 9/5 * (6167/3 alt + 1 annul) = ~3700 alt + 1.8 annul = ~617c + 41.4c = ~3.56ex | |
When I crafted my weapon I did a mix of ‘A’ and ‘C.’
- Alt slam + augment - If 1 of the mods was desired, regal/scour - Only imprint if +1 socket and 20 Maim (never did hit) - Twice I hit +1 socket and 18 Maim, 1st failed the annul, second hit strength and I kept it. The weapon is subpar from top tier crafted, but it’s certainly worth more than I spent. Right now the only way I could afford to craft another is if I sold my current weapon. | |
So which are the other 4th spectre to choose, cos this thread build dsnt have +2 curses only one, and im using despair. Diabolist as the author recommends, its useles cos curse caster only.
@execreth - @eledhyr - @arthasspaw - @edherisctf - @douchebagontheway_hd
@the_pest - @ioneshotbosses_eightkai | |
" Ruins Hellion from Act 8 Grain Gate for taunt. |
|
" The idea behind Diabolist is that grants an AoE curse whereas the AG will only curse his target. If you don’t want his curse to override the Diabolist’s, equipping him a source of +1 curse should work. If your Dispair is from an AoE source, then it’s relatively moot. 4th Spectre should be a Diabolist for AoE Enfeeble, Ruins Hellion for taunt and 15% increased damage buff, or a second Host Chieftan for higher uptime on power charges. Someone earlier recommended a Reanimator instead of Diabolist. I couldn’t say which is better. | |
" That's not MY problem. I got the Maim + 1 socketed gems in like 350 alts. The problem is that i can't get the Annulment step to work. I'm not "complaining" i was just asking for an alternative to this since i've had bad luck using the crafting method. I'll just keep trying until it works |
|
Since we're discussing crafting, this can be easily crafted with minimal RNG at less than 5ex. The damage it gives is 25% less than the ideal weapon. I used a staff base because the implicit is actually useful.
From League Starter to Shaper – Cheap, Tanky, Fast and Fun Physical ST – Very Detailed Guide
https://www.pathofexile.com/forum/view-thread/2077519 Gannicus, God of the Arena – Deathless Uber Elder for Everyone https://www.pathofexile.com/forum/view-thread/2627521 |
|
" >> That's not MY problem. I'm confused by your response. You're (apparently) dissatisfied with the imprint method of crafting because your annulment orbs have so far failed to remove the bad mod and you've asked for an alternative crafting method to avoid this step (if I've misinterpreted your post let me know). An alternative method to craft your weapon has been outlined which does not use the annulment step at all - but has a chance of failure nonetheless. Seems to me that this alternative method fits your request exactly... the "That's not MY problem" remark leaves me puzzled. Last edited by hankinsohl on Sep 26, 2019, 11:12:47 AM
|
|
" This agrees with my calculation exactly, because I assumed we only want Maim level 20. If we are also happy with level 18, the probability goes up by a factor of 2. So these are the same numbers I got. |