Orb of Fusing - How does it work?

"
ChanBalam wrote:
So what number needs to be rolled to get the first fusing? A "1"? A "6"? Or perhaps a 1 on the first fusing; a 2 on the second, etc. and finally a 5 on the fifth?

Firstly, small amendment to my explanation: it rolls between 1 and the number of sockets, not 0 (or it might roll between 0 and the max number of possible links; I don't know which, but it doesn't matter).

Let me use an example. Say you have a random 6-socket item and use a fuse on it. On the top-left socket, it rolls a 3. So the first three sockets are linked. It then jumps to the next socket after that group of links, which is the middle-left socket. There it rolls, say, a 1. That socket remains unlinked. It then jumps to the bottom-left socket, and rolls a 2, linking the final sockets.

To get a 6L, you have to roll a 6 on the first socket. That's it.

Also, this is why 6-socket items can occasionally become "backwards" 5Ls. That occurs when the first socket rolls 1 and the second rolls 5.
Have you done something awesome with [url=http://pathofexile.gamepedia.com/Sire_of_Shards]Sire of Shards[/url]? PM me and tell me all about it!
"
Namcap wrote:
"
dudiobugtron wrote:
"
Ludvator wrote:
you are unlucky a bit

More than one item 6L in 100-500 fusings and you say *un*lucky?


That's the part I find the funniest.

Guys I won the lottery 4 weeks in a row but then I didn't win for 2 weeks. What's going on? Am I cursed?

Meanwhile I'm sitting there with 1 manual 6 link ever in my history of PoE (couple thousands of fusings over different items and leagues).


The thread was not meant to sound like a whine thread. It was more or less a question if it might be a possibility that there is a hidden modifier in each item which influences whether an item gets linked easily or almost never.

I've played a lot during the last years. First item I linked ever was a Shavronne's back in open beta which I spent like 5000 fusings on. I thought it is horribly hard to get a 6 link and of course it was harder back then because there was no option to fus on 20% quality. A lot of people ran lotteries (which seems quite uncommon now).

After this experience I decided to only use fusings until an item gets 5 linked and start on a new one. I 6 linked maybe 4-5 items since then. Some with little effort, some with higher costs - but all of them before they got 5linked for the first time.

"
viperesque wrote:
"
ChanBalam wrote:
So what number needs to be rolled to get the first fusing? A "1"? A "6"? Or perhaps a 1 on the first fusing; a 2 on the second, etc. and finally a 5 on the fifth?

Firstly, small amendment to my explanation: it rolls between 1 and the number of sockets, not 0 (or it might roll between 0 and the max number of possible links; I don't know which, but it doesn't matter).

Let me use an example. Say you have a random 6-socket item and use a fuse on it. On the top-left socket, it rolls a 3. So the first three sockets are linked. It then jumps to the next socket after that group of links, which is the middle-left socket. There it rolls, say, a 1. That socket remains unlinked. It then jumps to the bottom-left socket, and rolls a 2, linking the final sockets.

To get a 6L, you have to roll a 6 on the first socket. That's it.

Also, this is why 6-socket items can occasionally become "backwards" 5Ls. That occurs when the first socket rolls 1 and the second rolls 5.


Interesting. Never found something like this when I searched for information about linking.

I still feel it should somehow slightly increase the chance of linking an item the more fusings you use on it. Not much, just slightly. Mainly to not feel like you "wasted" thousands of fusings if you're really unlucky.
Last edited by DerDave on Aug 5, 2016, 4:56:50 AM
"
viperesque wrote:
"
ChanBalam wrote:
So what number needs to be rolled to get the first fusing? A "1"? A "6"? Or perhaps a 1 on the first fusing; a 2 on the second, etc. and finally a 5 on the fifth?

Firstly, small amendment to my explanation: it rolls between 1 and the number of sockets, not 0 (or it might roll between 0 and the max number of possible links; I don't know which, but it doesn't matter).

Let me use an example. Say you have a random 6-socket item and use a fuse on it. On the top-left socket, it rolls a 3. So the first three sockets are linked. It then jumps to the next socket after that group of links, which is the middle-left socket. There it rolls, say, a 1. That socket remains unlinked. It then jumps to the bottom-left socket, and rolls a 2, linking the final sockets.

To get a 6L, you have to roll a 6 on the first socket. That's it.

Also, this is why 6-socket items can occasionally become "backwards" 5Ls. That occurs when the first socket rolls 1 and the second rolls 5.


Interesting!

Does this mean that getting a "top" 5L is more likely than a "bottom" 5L?

P(Top 5L) = P(First socket rolls 5)

P(Bottom 5L) = P(First socket rolls 1) * P(Second socket rolls 5)
Yes. I'm told that that's why you observe fewer backwards 5Ls in the market; they're slightly less likely to occur than forwards 5Ls.
Need game info? Check out the Wiki at: https://www.poewiki.net/

Contact support@grindinggear.com for account issues. Check out How to Report Bugs + Post Images at: https://www.pathofexile.com/forum/view-thread/18347
.
Last edited by Entropic_Fire on Oct 26, 2016, 9:37:20 PM
"
adghar wrote:
Yes. I'm told that that's why you observe fewer backwards 5Ls in the market; they're slightly less likely to occur than forwards 5Ls.


Yeah, the percentage of "1-5" items relative to the number of "5-1" and "5" items is exactly the same as your chance to roll "1" on first socket. The knowledge has a practical value too, if we assume your chance to roll 1 is 25%, that means an item with 6 sockets has a 25% better chance to get a 5L than one with 5 sockets (not to mention a chance for an accidental 6L).
Wish the armchair developers would go back to developing armchairs.

◄[www.moddb.com/mods/balancedux]►
◄[www.moddb.com/mods/one-vision1]►
If the chance of rolling a 6 link is one in six on the first socket, then how do the remaining rolls on the other sockets affect one's final result?

I see how you get 1:1250 from 0.1667 x 0.1667 x 0.1667 x.1667 x 0.1667. So if you get a six on roll one/socket one, don't you also have to get a six on the remaining 4 rolls?

Do the existing, starting links on an item affect the rolls or results?
"Gratitude is wine for the soul. Go on. Get drunk." Rumi
US Mountain Time Zone
ChanBalam: You're misunderstanding viperesque. The rolls are not equally weighted (i.e. you said "if it is 1 in 6" but it is not 1 in 6).

Viperesque did specify that the community does not know the exact odds; 1/1250 is a guessing estimate. A guesstimate, if you will.

You should instead think of a 1250 sided die, and only 1 side reads 6. Let's say 250 sides of that die read 1, maybe 500 read 2, 250 read 3, 200 read 4, and 49 read 5.

If you roll a 6 on the first socket, the other sockets are irrelevant (no rolls are made on them because all sockets have already been linked).

If you roll a 1 then any 6 you roll is wasted, because only 5 sockets remain. I believe what happens then if you roll a 6 is to make it a 5L, but it might just reroll. I'm pretty sure that has no effect on the actual probability (it will do that weird door switch paradox thing though).
Need game info? Check out the Wiki at: https://www.poewiki.net/

Contact support@grindinggear.com for account issues. Check out How to Report Bugs + Post Images at: https://www.pathofexile.com/forum/view-thread/18347
.
Last edited by Entropic_Fire on Oct 26, 2016, 9:38:45 PM
"
Entropic_Fire wrote:
What were my chances of linking this in 43 fuse with the 1/1100 figure?

It was 3,7%

P = N * p * q^(N-1)

N = the number of opportunities for event x to occur
p = the probability that event x will occur on any particular occasion
q = the probability that event x will not occur on any particular occasion.
Wish the armchair developers would go back to developing armchairs.

◄[www.moddb.com/mods/balancedux]►
◄[www.moddb.com/mods/one-vision1]►
Last edited by raics on Aug 5, 2016, 1:55:01 PM

Report Forum Post

Report Account:

Report Type

Additional Info