6-socket Orb of Fusing Community Log (done)

"

HikaruYami написал:

It's painfully obvious from all of the posts in recent months talking about how many fuses it took people to get 6Ls that the community has no idea what it's doing or how math works (Scrotie seems to, though).

You need to be prepared to record complete data far in advance of posting it. While it's true that we can back-solve an estimate of probability from what people are doing, it's a far worse estimate than what Scrotie's initial post was talking about.

Like, you've saved 3000 fusings. You reeeeally want that 6L. You want to present data about the use of these 3000 fusings. What you absolutely _do not do under any circumstances_ is say "welp, I'll just post the number it took me to get that 6L once". Never do that. That is unreliable data.

If you use all 3000 and got no 6Ls, first off, _you have to post that data if you resolved to collect it in the first place_. If you cannot be trusted to do that, then you have no place in data collection for anyone, anywhere, ever, because your "positive" data will only ever skew the samples.

If you get the 6L you want before using all 3000, find something else worth 6L'ing and use the rest of them. Keep doing this until you run out of all 3000. This is the only way to collect good data. Idiots posting just the amount that it took _them_ to successfully 6-link _once_ is warped positive data that only destroys any kind of hope for a statistical model.