Guide Ingredient Ids Are Not An Equal Distribution + Other Findings

A bit technical stuff below, but tl;dr: Ingredient IDs use integer values of 0-100 to determine the roll of each ID. This is similar to ID rolls on regular items. Attack Speed and level of equipment are leaning towards a bell curve as well but more data points would be needed. Durability values are inconclusive.

---------------------------------------

@nbcss wanted to know whether ingredient IDs were based on using a base integer value to determine the final ID value upon crafting an item. He chose using Infernal Flesh, which has a particular value of 600-800 HP. If it was based on an integer value, then each crafted item will only have an even HP value. After 10 rolls by @Rypreorow, this was exactly the case. This means that the base integers used are between 0-100, because in order to get an odd number, any number of base integers other than 101 base integers would be needed.

Now I came up with the question, are these base integers equally distributed across all possible ID values? By this, I mean that if an ingredient has 3 possible ID values, will the base integers be equally distributed as 33/33/34 or as a bell curve? I chose Rotten Flesh, which has 2-4% Main Attack Damage. 96 data points were collected. The results are below:



As seen from the chart, the values follow a bell curve. Another person, @Nieke, also found similar results:

• 3 possible values: 17:27:20
• 4 possible values: 9:24:20:11

Notice the 4 possible values dataset. The peak is more of a plateau, which indicates that it follows very similarly to how item IDs are determined. So our conclusion is then that these base values are first equally distributed between the maximum and minimum values, and then are rounded to the nearest integer value. For example, given values 2-4, 25 base integer values fall between 2-2.5, 50 base integer values fall between 2.5-3.5, and 25 base integer values fall between 3.5-4.

More data points would be nice for 4 or more possible values but I think this is pretty conclusive.

An example of what is described above:

We know that there are base integers from 0-100. Let's assume that an ingredient has a value between 1-10. From there, the base integers are evenly distributed between 1-10. These are then rounded to fit the integer value between 1-10. This then gives our chance to obtain said value.



---------------------------------

For data purposes, I also collected the combat level, damage value, attack speed, and durability. Unexpectedly, I also found a bell curve for the combat level as well:



Attack speed could also be a bell curve but more data points would be needed. Durability is inconclusive:



Because of Excel problems, I couldn't find a way to convert the fake dates back into damage values without manual typing. I'm too lazy to check and compare to attack speed, but I have attached the sheet over on the prof discord if you want to take a look. I can't attach it here :(

 

arent base rolls for item ids uniform
or am i wrong

 

Yeah I think the rounding memo was missed a bit here

 

Can’t understand anything you say so I’ll just like and nod and pretend I get it

 

Ah, I see what you mean. That's an interesting discovery.

 

I did? Can you explain so I don't mislead people?

 

would be more clear with a general-case graph with more than 4 possible values (even 5 is better than 4 bc it at least shows the plateau better, denying a bell curve)

sorry if it's a big ask

 

Something like this we suspect:



We know that there are base integers from 0-100. Let's assume that an ingredient has a value between 1-10. From there, the base integers are evenly distributed between 1-10. These are then rounded to fit the integer value between 1-10. This then gives our chance to obtain said value.