Ingredient Maths And Thanos Legionnaire Set Bonus Questions

What's the set bonus for the Thanos Legionnaire armor? I want to know it for a possible melee mage build and wynndata doesn't have it.
EDIT: Wynndata got updated and it has the set bonus now.

Spoiler: Things I assume about ingredients

About crafting ingredients and how they reach their effect:
I know that unique, rare, legendary and mythic items get a number assigned to each of their IDs between 30% and 130%. (when positive, between 70% and 130% when negative.)
This gets rounded up. Meaning that 1/4 to 2/4 mana regen things, for example. Have a base of 1/4 mana regen at 100%. If you randomly get 101%, you get 2/4 mana regen, despite actually only having 1.01 mana regen.

When looking at the ingredients, considering their random id-like nature, it makes sense for that they use the same system of percentages away from a base and rounding upwards. But when we look at the ranges it seems like the percentual numbers change each time.

Spoiler: For example

You've got a skyraider coin which gives 35-45 melee damage. With a base of 40 melee damage, this is a 12.5% change in both directions. Which can be rounded down to 12% without changing the numbers it effects, that is important because there is no weapon that has (for example) 78.5% of the base. Cataclysm can get -5500 HP on a roll of 110%, but never -5525 HP, since that'd be a roll of 110.5%.

There's also Farcor's trust, which gives 3000-3400 HP. With a base of 3200 HP, this is a 6.25% change in both directions.
Which is weird, because you can not round this down to a 6% deviation. That'd make this
3008 - 3392 HP

Farcor's has a 15-30% boost in fire defense. Since I believe that there can't be any non-full numbers as the base of 100% and it always rounds up, I assume the base here should be 22%
This gives it about a 35% deviation from the base of 22%.
Thus teaching us that it's effect-specific, not ingredient specific.

But what I really wanna ask is beneath this spoiler

Does ingredient effictivenes add before or after the effect of the ingredient is calculated?

Spoiler: What I mean and why it matters

Let's say we add a Borange Fluff (+15% ingredient effectiveness to everything.) to a recipe that also has Cat Food (1/4 to 2/4 mr) in it.

The cat food should have a base of 1/4 mana regen, with about a 50/50 chance of it being 2/4 mana regen.

Spoiler: Why 50/50?

There's a deviation of X% from 100%, meaning that it can be anywhere between 100-X and 100+X as well as 100 itself, everything with 100+X will be 2/4, everything with 100-X or 100 will be 1/4

Spoiler: How does this work with things that add 0-1 of something?

If an ingredient like the shiny mineral deposit should have 0-1 of an ID, it's base is 0.
0+X rounded up with X<1, will be rounded up to 1 and 0-X with X<1 will be rounded up to 0.

Which is good news for the black hole ingredient, since the only way to get a -1 tier attack speed to +1 tier attack speed if my assumptions through observation have been correct, is to have an X of 1
since 0-1 = -1, but 0-0.99 would still round up to 0, and 0.01 would already round up to 1, so for black hole, you get a 1% chance of it being -1 tier attack speed, a 50% chance of it giving no +tier attack speed and a 49% chance of it giving +1 tier attack speed. Which is much better than 1/3 for each of them, duntcha think?

This idea does have a flaw, however. Since X% of 0 still is 0. This is the main reason I doubt if it even uses a % based system like identifications, but it could also be that bases can be in decimals for ingredients or that the X isn't in percentages when the base is 0 or whatever, idk.

If the borange fluff adds its effect after the previously stated effects have happened, it'll increase a 1/4 mana by 15%, rounding 1.15 up to 2/4 mana regen and rounding 2.30 up to 3/4 mana regen.

If the borange fluff adds its effect at the same time, the percentages become 115 - X to 115 + X, meaning that the borange fluff only makes it more likely that you get 2/4 mana regen, rather than guaranteeing 2/4 or higher.

Taken to its extremes, this is the difference between a recipe like 3 x black hole, (2 x Panda kings crown for durability) and a borange Fluff getting up to +6 tier attack speed or up to +3 tier attack speed.

So I assume that it's the | 115 - X --> round up to 115 + X --> round up | one (during calculation of previous chances)
Rather than the | ((100 - X) --> round up) x 1.15 to ((100 - X) --> round up) x 1.15 | one (after calculation of previous chances.

If any of my assumptions are incorrect, please do say so as well. I've elaborated them all precisely because I don't know if they're correct.

________________________________

Can't edit polls ;_; The second options says that I was explaining my thoughts too much.

 

Ingredients:

I'm 99% certain that ingredient values modified by effectiveness round naturally, rather than always rounding up (otherwise what is the point of Major's Badge?)

I can't think of any way to test whether effectiveness is applied before or after the effects calculation that doesn't rest on uncertain premises. It's possible (I'd even lean toward probable) that the ingredients system doesn't work with base and modifiers the way your assumption is written, and is instead a simple linear distribution over each stat's range of values - given that there are certain limitations to the present ID system, such as the inability to have a max +2 item, I could see whoever came up with the ingredient stat distribution system avoiding the base/modifier method for that reason, or other reasons.

I think the Thanos Legionnaire set bonus is listed in the Item updates thread (don't remember for sure).