|
|
On December 16 2015 02:14 Sn0_Man wrote:
back of the napkin math puts RF at like 10-15k burning dps per totem at juuuust before maps. Plus maybe ancestral bond burn and cursing.
Doesn't seem too OP although it does seem pretty strong.
Actually this sounds fun as fuck imma theorycraft the build for better numbers.
It's way worse.
We're talking about 20k dps *before* Rapid Decay. And it stacks with Searing Bond.
This thread has the math you're looking for. Kinda sad that some other guy is making it popular and he just found out RD works. Pretty sure that interaction will be considered a bug by GGG, otherwise the build is way too broken.
|
level 81 is hardly "juuuuuust before maps" though
but yeah add rapid decay and :O
also the first poster is retarded xd "faster casting rounds out the 6L" *facepalm*
|
I'm pretty sure GGG confirmed that Rapid decay wouldn't work with RF.
|
That's because RF doesn't have the duration tag
because RF totems have the duration tag (due to totems having a duration of course), it apparently works. Which is likely going to get patched out.
|
Haha, that's a nice interaction. Does indeed seem likely to get patched out pretty quickly. You can still have 6 good links with that new destruction support gem though right?
|
Required: Spell Totem, Righteous Fire
OP: Increased Burning Damage, Rapid Decay
Other: Increased AoE (quality of life), Conc Effect (single target damage), Minion and Totem Elemental Resistances (for builds that don't wish to run purity of fire such as blood magic builds)
very little else actually has effect (item rarity i guess). Increased burning damage is surprisingly weak since it's an "increased" modifier and your tree ends up with lots of increased modifiers already but it's still definitely worth a link.
E: theres increased duration I suppose...
|
On December 16 2015 05:06 Sn0_Man wrote:
Required: Spell Totem, Righteous Fire
OP: Increased Burning Damage, Rapid Decay
Other: Increased AoE (quality of life), Conc Effect (single target damage), Minion and Totem Elemental Resistances (for builds that don't wish to run purity of fire such as blood magic builds)
very little else actually has effect (item rarity i guess). Increased burning damage is surprisingly weak since it's an "increased" modifier and your tree ends up with lots of increased modifiers already but it's still definitely worth a link.
E: theres increased duration I suppose...
Empower maybe? For more Totem Life and higher level of the support gems... probably quite good, but hey, also quite expensive & not available at lvl 2/3 early on in the leagues.
|
Empower only affects the active skill which is righteous fire. You don't bother levelling RF cuz it's dmg doesn't increase so nope, no empower.
|
On December 16 2015 05:26 Sn0_Man wrote:
Empower only affects the active skill which is righteous fire. You don't bother levelling RF cuz it's dmg doesn't increase so nope, no empower.
Oh was it only the active gems, it's been a while since I used it.
Given I play a curse build atm, it's quite sad to read the recent patch notes. Yes, what Ghudda did wasn't fine. But GGG's nerf completely destroyed the quality of both Temporal Chains (lol cast speed) and Curse on Hit (lol inc duration). The other two aren't great either. Should've just capped both Slow and Hit Chance at 75% or w/e, like almost everyone seems to think.
|
Has anyone actually crunched the binomial probabilities on Talisman altars? I see a lot of "in the long run, putting in 1x (rarity A) and 4x (not rarity A) will yield the same number of rarity A as putting in 5x rarity A," which is technically true, but I don't think it converges very rapidly for the higher tier talismans.
E.G. The probability of getting 0 of the next tier talismans when you do ten trials at 0.2 chance of success is a whopping 10%, and you'll get 0 or 1 ~38% of the time. You only get more than 2 ~32% of the time.
Some math I may have gotten wrong:
+ Show Spoiler +Usually you can say your mean reliably approximates your proportion when your number of trials times your proportion is greater than 5 (or some say 10).
n*.2 trials > 5 requires n=25 for a given tier, and 25 tier 3 trials (requiring 125 tier 3 talismans) requires over 15000 tier 1 talismans.
On the other hand, if you use 4 rare and 1 non-rare, it only takes 6.25 trials for your mean to approximate the proportion. That's less than 4000 tier 1 talismans (4500 if you want to round up to 7 trials).
|
Wow I never would have believed it, but you do appear to have mathematical proof that PoE uses a lot of RNG and outcomes vary wildly.
No but seriously. It's obvious that there's going to be variance when you don't do 5/5 talismans every time. So what? Decide if you can accept that variance or not and trade up accordingly.
|
Your problem is using a post that says in the long run and then complaining about low sample size. You can't have both.
|
On December 16 2015 07:02 Blitzkrieg0 wrote:
Your problem is using a post that says in the long run and then complaining about low sample size. You can't have both.
My point is that "the long run" for doing 1 + 4 is substantially different from "the long run" for doing 2 + 3 or "the long run" for doing 3 + 2 or "the long run" of doing 4 + 1. In fact it's 2x, 3x, and 4x as long, respectively. It's infinitely longer than "the long run" for 5 + 0 as well, but that's kind of obvious.
"In the long run" you also always have a 50% chance of doing better and a 50% chance of doing worse than expected, too, so why not just take the sure thing? It's not like fusing where according to Chris you are paying more than the expected value in fusings.
|
On December 16 2015 07:09 TheTenthDoc wrote:Show nested quote +On December 16 2015 07:02 Blitzkrieg0 wrote:
Your problem is using a post that says in the long run and then complaining about low sample size. You can't have both. My point is that "the long run" for doing 1 + 4 is substantially different from "the long run" for doing 2 + 3 or "the long run" for doing 3 + 2 or "the long run" of doing 4 + 1. In fact it's 2x, 3x, and 4x as long, respectively. It's infinitely longer than "the long run" for 5 + 0 as well, but that's kind of obvious. "In the long run" you also always have a 50% chance of doing better and a 50% chance of doing worse than expected, too, so why not just take the sure thing? It's not like fusing where according to Chris you are paying more than the expected value in fusings.
Yes, but how much easier is it to acquire 1 rare and 4 not rare talismans than waiting around for 5 rare talismans? Since you can't manufacture rare talismans like most other crafting in the game due to corruption, you want to push your rare talismans through faster and it does not change the rate so you aren't losing efficiency.
On December 16 2015 07:17 Sn0_Man wrote:
i mean there's always the possible upside too of being a lucksack
thats the whole point of expected value: in order for you to be able to come below EV, you have to be able to come above too
The value here doesn't have the same meaning with sockets versus talismans. One provides a guarantee where you know you'll get five sockets. The other merely guarantees a rare corrupted item that could roll with any affixes you can't change. Just getting a rare item is meaningless. You also need that rare item to be good so going for the riskier crafting method has more value. You'll need to be lucky no matter what.
|
i mean there's always the possible upside too of being a lucksack
thats the whole point of expected value: in order for you to be able to come below EV, you have to be able to come above too
|
On December 16 2015 07:15 Blitzkrieg0 wrote:Show nested quote +On December 16 2015 07:09 TheTenthDoc wrote:On December 16 2015 07:02 Blitzkrieg0 wrote:
Your problem is using a post that says in the long run and then complaining about low sample size. You can't have both. My point is that "the long run" for doing 1 + 4 is substantially different from "the long run" for doing 2 + 3 or "the long run" for doing 3 + 2 or "the long run" of doing 4 + 1. In fact it's 2x, 3x, and 4x as long, respectively. It's infinitely longer than "the long run" for 5 + 0 as well, but that's kind of obvious. "In the long run" you also always have a 50% chance of doing better and a 50% chance of doing worse than expected, too, so why not just take the sure thing? It's not like fusing where according to Chris you are paying more than the expected value in fusings. Yes, but how much easier is it to acquire 1 rare and 4 not rare talismans than waiting around for 5 rare talismans? Since you can't manufacture rare talismans like most other crafting in the game due to corruption, you want to push your rare talismans through faster and it does not change the rate so you aren't losing efficiency. Show nested quote +On December 16 2015 07:17 Sn0_Man wrote:
i mean there's always the possible upside too of being a lucksack
thats the whole point of expected value: in order for you to be able to come below EV, you have to be able to come above too The value here doesn't have the same meaning with sockets versus talismans. One provides a guarantee where you know you'll get five sockets. The other merely guarantees a rare corrupted item that could roll with any affixes you can't change. Just getting a rare item is meaningless. You also need that rare item to be good so going for the riskier crafting method has more value. You'll need to be lucky no matter what.
"You'll need to be lucky no matter what" doesn't really mean anything. It's not like getting that 1/5 chance at a rare talisman means that talisman is somehow more or less likely to be dumpster trash. If I trade in 2000 rare amulets at 5 to 0 and you trade in 2000 at 1 to 4, we'll both get an equal number of good amulets (probably). Edit: You just need 1600 more stone circles to do it (and you also need the other 8000 talismans).
I guess this will all change when the stone circles are more common because then the number of attempts doesn't actually matter then (since each method takes an equal number of rare talismans to approach the mean) and the important thing will be packs and Rigvald spamming, where the 1 to 4 method can enable that better.
More people being in maps and rolling IR on them will also change it a bit I think.
|
|
|
On December 16 2015 07:27 TheTenthDoc wrote:Show nested quote +On December 16 2015 07:15 Blitzkrieg0 wrote:On December 16 2015 07:09 TheTenthDoc wrote:On December 16 2015 07:02 Blitzkrieg0 wrote:
Your problem is using a post that says in the long run and then complaining about low sample size. You can't have both. My point is that "the long run" for doing 1 + 4 is substantially different from "the long run" for doing 2 + 3 or "the long run" for doing 3 + 2 or "the long run" of doing 4 + 1. In fact it's 2x, 3x, and 4x as long, respectively. It's infinitely longer than "the long run" for 5 + 0 as well, but that's kind of obvious. "In the long run" you also always have a 50% chance of doing better and a 50% chance of doing worse than expected, too, so why not just take the sure thing? It's not like fusing where according to Chris you are paying more than the expected value in fusings. Yes, but how much easier is it to acquire 1 rare and 4 not rare talismans than waiting around for 5 rare talismans? Since you can't manufacture rare talismans like most other crafting in the game due to corruption, you want to push your rare talismans through faster and it does not change the rate so you aren't losing efficiency. On December 16 2015 07:17 Sn0_Man wrote:
i mean there's always the possible upside too of being a lucksack
thats the whole point of expected value: in order for you to be able to come below EV, you have to be able to come above too The value here doesn't have the same meaning with sockets versus talismans. One provides a guarantee where you know you'll get five sockets. The other merely guarantees a rare corrupted item that could roll with any affixes you can't change. Just getting a rare item is meaningless. You also need that rare item to be good so going for the riskier crafting method has more value. You'll need to be lucky no matter what. "You'll need to be lucky no matter what" doesn't really mean anything. It's not like getting that 1/5 chance at a rare talisman means that talisman is somehow more or less likely to be dumpster trash. If I trade in 2000 rare amulets at 5 to 0 and you trade in 2000 at 1 to 4, we'll both get an equal number of good amulets (probably). Edit: You just need 1600 more stone circles to do it (and you also need the other 8000 talismans). I guess this will all change when the stone circles are more common because then the number of attempts doesn't actually matter then (since each method takes an equal number of rare talismans to approach the mean) and the important thing will be packs and Rigvald spamming, where the 1 to 4 method can enable that better. More people being in maps and rolling IR on them will also change it a bit I think.
It isn't, but if I'm finding stone circles and rare talismans at the same rate then I can do the 1:4 method to constantly attempt to upgrade them without degrading the efficiency.
The only point here is that it doesn't degrade the efficiency. No where has it ever been stated that it more efficient to do them 1:4. You asked why not take the sure thing and I replied that the sure thing isn't actually a sure thing. Guaranteeing a rare talisman isn't useful because you aren't guaranteeing that said rare talisman actually rolls well. The expected value of a rare talisman is pointless, because acquiring a rare does not inherently have value. You need that rare talisman to be good as well.
|
I'm not surprised, GGG has shown willingness to acknowledge when they're wrong and meet the players halfway/ all the way. They're a great company honestly.
|
When you start Dominus and see SRS everywhere but none of the mobs have a talisman...you know you're gonna have a good time. The very next run had the same but with skeletons x_x.
|
|
|
|
|
|
EPT
