As anybody who's run a batch of wands might know, Here's a few points:
- You will never roll 0matk
- You will never get under 14 durability
- 5 matk seems to have a 10% chance despite being theoretically 25%. Its also a 50/50 if the crit or dura are good as well
Across all the items I've made r1, it seems that its impossible to roll some things inherently due to rank. At r1 its an estimated 15% of the roll minimum. This means 1matk savage wands, 1matk Hermits, 2 of each stat purple hearts, and 3 luck rabbits feet. This is a universal given at r1 that the minimum roll is raised from the bar. Its an average 15% across all items, but its higher for others, and lower in some cases. 0->1 in wands is a 20% base, but droseras can't go 1->2 prot because it can only roll 0, 1, and 2. +33% increase would be far too much.
Averages also seem to be around the median. Take this sample of numbers for instance. They're my final amount of rabbits feet before master:
[CODE]5 3 2 7 8 3 5 4 4 14 9 6 3 7 10 6 2 6 10 7 6 10 5 12 2 13 11 6 8 7 6 12 3 6 7 8 3 4 11 3 14 12 7 3 8 11 2 8 3 8 11 5 10 9 2 9 8 4 9 8 3 8 12 9 9 4 13 6 2 11 6 6 10 14 2 9 7 13 7 7 12 11 5 2 2 10 14 5 12 8 2 5 2 6 5 3 4 9 2 14 5 12 10 12 14 13 3 2 2 4 3 2 2 6 3 3 3 2 11 2 6 11 12 7 6 9 12 14 10 14 4 13 13 5 14 7 11 10 6 11 13 9 12 14 8 13 6 2 14 5 11 9 12 7 13 12 10 9 3 7 13 7 2 4 5 10 4 10 2 4 10 2 5 5 13 6 10 14 13 2 11 3 12 2 3 5 2 13 2 4 11 8 8 14 8 8 12 4 4 13 4 13 4 7 2 9 11 9 12 7 7 2 12 10 8 12 6 11 8 10 11 6 9 7 11 2 14 9 3 14 10 3 11 12 8 12 4 7 8 11 5 11 7 9 6 13 13 2 9 10 11 4 11 7 10 10 9 12 14 8 7 7 6 5 14 6 2 3 11 13 14 12 11 6 13 10 12 9 3 9 11 3 7 9 4 10 14[/CODE]
More specifically, the rolls as it is impossible to roll a 2 luck rabbit's feet, and 1 is the inherent minimum. Only droseras actually break this law, as you can't roll 0 prot, but 1 prot is the minimum at all ranks. Going up to 2 would be a vast improvement to the minimum, being the roll of 0, 50%, and 100%. The average here is 7.8. This is a 60% average roll under 13 possible numbers. I normally say its a 50% average roll, but that's apparently wrong. How I wish I had the data set from my previous acc makes. They'd be rather helpful here.
Now, I rolled a rather small sample, probably because these are some of the most painful accs to roll: Purple Hearts. A 40 sample as seen here, the sum of the inherent bonus from 1~9 of each stat, combined.
[CODE]17 21 22 16 14 20 16 30 19 19 24 24 25 29 22 27 23 17 29 19 19 23 22 23 17 24 17 10 22 32 20 20 20 17 21 20 24 13 29 20[/CODE]
Now, the average here is 21.5 The lowest rolls being 10, 13, 14, 16, and 16. Now, this would seem natural, 21.5 divided by a possible 36 gives a 59.72% as you only can roll from 2-9 for each stat. 1 is a given, and you cannot roll +0, so the possible numbers are 2~10, Nine different variants. Seems average, nothing wrong here. Or is it? What could change this?
Apparently, the minimum roll of an individual stat. The purple hearts rolled no 2stat at all, I checked my old stockade to see if it was 3 minimum naturally, and found several purple hearts with 2s. The minimum roll was increased from 1/9 to 2/9, closer to 20%. This means that its not 59.72%, but 67.1875%. The midway mark is ([85x.5]+15=57.5) naturally. This would drag it if its 15~100% to 20, or 25% to 100, up to 60% at 20%, and 62.5% at 25%. But as we can see, that's wrong.
This means that the median is not a fixed 50% at all, its actually higher than that normally, placed at 60%, but increased to almost 70% Put this to say, A Celtic tribolt wand with its insane rolls, a minimum of not 8matk, but quite possibly 12, and an average of lower 20s. And its massive durability roll of 1-40, turns into a minimum of possibly 25~28, and 45~50 would be a commonplace.
Purple hearts though are insanely difficult to measure steady calculations off of. If I had more mutant rabbits feet, I'd make more lucky accs, but I lack the rabbits feet. I'll likely make them as they're the easiest and a single-stat standard, but getting enough for a sizable standard, a massive amount of 400 for an 80 batch probably would return sufficient results. Purple hearts are RNG on 4 different places, overtaken only by hermits, divine weapons, and dragon gear minus the circlet. They're also insufferably hard to perfect, as the highest roll I have ever gotten was a 33 or 32 total stats rolled, which is akin to two 10s and two 8s.
How it wrangles the average up probably is a different hook, and [S]calculating the chances of a perfect item are rather difficult.[/S]
Out of 436 zeder IIs, without mastery, I had a single one get 30/9/19/23, the other was a point or two less in bal/crit. A perfect for every stat depends, but taking your standard pair of guns it'd be okay to ballpark it at 1/500 nonmastered. Mastery affecting that to about 1/425. Ayyyy my mastery package would have net me a 100% perfect pair of zeders or two of em with mastery. For simpler items its a lot more likely. A single stat is 1/10. My head is in the water with 2~3 diff stats. ¯\_(ツ)_/¯