6-socket Orb of Fusing Community Log (done)

"

ScrotieMcB wrote:

"

tidus1492 wrote:

http://tinyurl.com/poeexpcalc

no need for community log on 6L its something you can calculate

Whoever made that has the chance of 6L at 0.33%, or 1 in 300; based off the data we have so far, I can be about 95% sure that that percentage is wrong. It also has 5L chance at 1.63%, also 95% chance of being wrong. Best explanation: he pulled numbers out of nowhere or from a faulty source, and plugged them into a spreadsheet.

He probably pulled them out of the past. At one point the probability to 5-or-6 link was 1 in 40, and the probability to 6 link was 1 in 303. The data being collected in this thread is strong evidence that this has been changed a good deal.

"

MesostelZe wrote:

"

ScrotieMcB wrote:

"

tidus1492 wrote:

http://tinyurl.com/poeexpcalc

no need for community log on 6L its something you can calculate

Whoever made that has the chance of 6L at 0.33%, or 1 in 300; based off the data we have so far, I can be about 95% sure that that percentage is wrong. It also has 5L chance at 1.63%, also 95% chance of being wrong. Best explanation: he pulled numbers out of nowhere or from a faulty source, and plugged them into a spreadsheet.

He probably pulled them out of the past. At one point the probability to 5-or-6 link was 1 in 40, and the probability to 6 link was 1 in 303. The data being collected in this thread is strong evidence that this has been changed a good deal.

Hmm. Still doesn't quite explain the 306 instead of 303. I think it's more likely that the author did something I did until shortly before the creation of this thread: assume that link rates for Fusings match socket rates for Jeweler's. The Jeweler's odds*, from 1 to 6, are 100:90:80:30:5:1, which line up exactly with the formulae in that spreadsheet.

*Got it right this time, odds-probability nitpicker guy.

What is the reason to collect the data in this thread ?

Its like trying to collect data about the average eating patterns of people in a clinic for obesity. People who get a 5l after a few fusings (not to low since it is special too) are much less likely to post in a thread like this then people who used hundreds without and results.

Since people on the "edges" of the bell curve are much more likely to post here you are building a statistic on the data you should usually just ignore entirely.

"

ScrotieMcB wrote:

In order for the data to be accurate, you must regularly post your all of your statistics, not just the 5Ls/6Ls you create. The reason for this is if the only data people posted was of successes, the data would be biased away from failures that never become successes, making the odds seem higher than they really are - it's important to keep that from happening.[/acronym]

However, by not including miracle/low attempt 6L, you are biasing for the contrary and focusing on unlucky people who have dumped a lot of fusings. I for example, have one of those 4 fusing miracle stories and got a 6L. Since I have a 6L, i haven't needed to use more fusing on 6L attempts. This thread tries to find the rough 6L probability by adding/summarizing all the available data. Taking a 4 fusing 6L event, is data!

"

Droggeltasse wrote:

What is the reason to collect the data in this thread ?

Its like trying to collect data about the average eating patterns of people in a clinic for obesity. People who get a 5l after a few fusings (not to low since it is special too) are much less likely to post in a thread like this then people who used hundreds without and results.

Since people on the "edges" of the bell curve are much more likely to post here you are building a statistic on the data you should usually just ignore entirely.

Although people aiming for 5L aren't likely to post "5L success story" here, people shooting for 6L are likely to trip over it a few times and report some good numbers on that.

When it comes to the whole bell curve thing... I see your point. People with average results are the least likely to brag about them, while the abnormally lucky and unlucky are both biased into reporting here. Early on, I'm hoping I can get a mix of both, so that the average of the outliers is still close to the average. Later, I'm hoping public awareness of this thread will lead to people who wouldn't normally post their average results to keep track of them and come here in order to help the project.

"

"

ScrotieMcB wrote:

In order for the data to be accurate, you must regularly post your all of your statistics, not just the 5Ls/6Ls you create. The reason for this is if the only data people posted was of successes, the data would be biased away from failures that never become successes, making the odds seem higher than they really are - it's important to keep that from happening.[/acronym]

However, by not including miracle/low attempt 6L, you are biasing for the contrary and focusing on unlucky people who have dumped a lot of fusings. I for example, have one of those 4 fusing miracle stories and got a 6L. Since I have a 6L, i haven't needed to use more fusing on 6L attempts. This thread tries to find the rough 6L probability by adding/summarizing all the available data. Taking a 4 fusing 6L event, is data!

I don't know how you came to that conclusion... but nevermind.

The point is that we want all data. Big numbers, average numbers, even very small miracle numbers like yours. We do require posting the item itself for verification, but if you comply with that I'll gladly add your data to the mix.

150+ Fusings used
5Link : 1

Total fusings used on 6s
58
5l:1

Nuker, I'd need an exact number, not 150+.

Wugue, I'd need the item linked in-post.

141: nothing
153: nothing

70-80 fusings
3x 5l * stoped @ the 3th 5link