This is an unofficial explanation of how item power is calculated. It has been created due to confusion about minimum rolls vs item power after TU10.1 went live.
If you want to discuss actual loot quality since 10.1, there are plenty of threads on that topic: Example 1, Example 2
Background
The TU10.1 patch notes said this:
On State of the Game #165, this graph was presented to illustrate the loot power distribution for heroic content - note the 60% minimum average roll.Increased minimum item power and chances for higher power items for several difficulties, resulting in higher average rolls overall
Explanation
When asked on twitter about an item that had less than 60% rolls, Nikki Kuppens responded:
In other words, item power is NOT the same as minimum rolls. For example, these knee pads have a headshot damage roll that is less than 60%
To calculate item power:
- Weapon Damage = 100%
- Armor Regen = 89% (4413/4925)
- Headshot Damage = 49% (4.9/10)
Item power = 79%
This system allows for items like this backpack:
That's an 88% item = (1 + 0.66 + 1)/3 which can be recalibrated to a 100% item. Whether a 100% blue backpack is desirable is a topic for another day
Why are the bars the "wrong" size?
In the example above, the status effect roll is 6.6% vs a maximum of 10%. Or 66% of the maximum. Yet the bar is less than half
This is because the bars don't start at zero, but at the minimum roll for a given attribute.
If the minimum roll is 5% at L40, then the bar can show a maximum range of 5%. To calculate how big the bar should be: 1.6/5 = 0.32 or 32%
[edit] added calculation example for bars
What about Skill Tier?
Speculation - items that roll skill tier as the core attribute ignore that when calculating item power (as it is always maxed).
For example:
Status Effects = 100%
Skill Haste = 66% (8/12)
Item Power = 83% = (100 + 66) / 2
This could explain why skill items seem to roll better for me. As always, your RNG may vary.
Note: all items were taken from heroic content within the last 24 hours.
I take it that this is in relation to a certain video posted by a certain individual recently. I really would not pay any attention to someone that calculates a percentage value by repeatedly dividing by 2 until they get close to the value they see and then goes on to say that it's roughly something or other. All they are explaining at that point is why they don't understand the system.
The only valid point they made was that the bar doesn't even approximate the value.
No idea who you're talking about, and I'm happy to remain ignorant. It was inspired by people still posting about loot with less than 60% minimum rolls on heroic.Originally Posted by RichardOshea Go to original post
I address this at the bottom of the first post, but will add an exampleThe only valid point they made was that the bar doesn't even approximate the value.
Having said that, there may be a bug with heroic control points when the player or civilian leader dies. Anyone experiencing purple loot in those circumstances should add their details to that thread.
IMO, the subject matter experts, at Massive that make this game, don't know how any of this works, let alone people piecing it together on the internet.
This game just needs to die and Massive needs to release a remastered Division 1. That will by time to maybe get a dev company that isn't a "lowest bidder" tier company and make Div3.
And what piece of my explanation do you think is incorrect? The math has been laid out for you, and citations have been provided.Originally Posted by JUGRNAUT01 Go to original post
This is helpful but I am pretty sure that when they showed the graph they said this was based on challenging content and that heroic and legendary had higher average rolls. They never said by how much. I think for transparency purposes it would be helpful to players to know where the minimums maximums and averages were for each difficulty level. Also does it affect the number of loot per difficulty level. That would help players gauge what content and difficulty is worth their time.
I think most players complain because none of this is clear to players outside of forums that explain this. I appreciate how hard it is to explain as I have a background in statistics. But they need to do better in my opinion as they gloss over simple concepts that they assume people know about.
"Item Power" itself is a poor term to be using as the relative power of an item is a subjective value judgement. You might have an item with high rolls but those rolls might be for attributes that you feel are useless for the character/build you're creating. It takes a statistical value and imposes upon it a psychometric value that changes depending on the recipient of the item.
FWIW:Originally Posted by ladybug101 Go to original post
I doubt Massive will ever publish the actual ranges, so we'll have to hope someone can datamine them.
Given the subjective nature of value, it is the best we've got.Originally Posted by RichardOshea Go to original post
Correction - this will only be true for TU11 onwards.Originally Posted by dagrommit Go to original post
Citation: SOTG #168
I'm glad I labelled that comment as speculation![]()