I have found 2 scepters for comparison, they are mostly identical except the blue one has extra +1% increased spell damage and +20% increased cold damage. The yellow one, however, has +11 melee cold damage.
Ice thorns have scaling tags: Physical, Cold, Spell, Attunement.
If I understand correctly, “melee cold damage” should not apply, as there is no “melee” tag. So it sounds like the blue scepter should result in a higher overall DPS for Ice Thorns, because of it’s extra spell and cold damage.
However, it does not. The result is that the yellow one has a much higher DPS (72 vs 61).
Here are the screenshots demonstrating all of this (album, click for more images): https://imgur.com/a/lgL84hd
The tooltip is also adding in the 18% chance to ignite on hit on the yellow sceptre which is presumably pushing it over the DPS of the blue sceptre (since the adaptive spell damage on both is so low).
I’ve noticed a similar tooltip bug with Smite. When playing Void Knight you can take a passive node that converts Smite from Fire to Void damage, which might cause some problems? I’ve uploaded screenshots to imgur that should be self explanatory, but here is my description anyways:
Using Torch of the Pontifex makes my Smite tooltip DPS jump from 300 to 876, but I’ve tested my damage on the dummy and it doesn’t change anything. To test it I’ve just used an Item that adds the same flat amount of spell damage as the Torch and hit the dummy a bunch of times. My dummy damage was the same with either item, but the tooltip DPS was vastly different.
The tooltip is also adding in the 18% chance to ignite on hit on the yellow sceptre.
It’s not for me to decide, but I wonder if it’s at all a good idea to overcomplicate the DPS calculation in this way. There are tons of different mechanics in the game, and I really doubt that DPS calculation would be able to keep up with all of them.
For example, there’s an item - Gambler’s Fallacy, which increases crit rate but then decreases it after any crit happens. Technically, this also on average affects the DPS, but I would be very surprised if this is calculated correctly (if developers even attempted to add this effect to DPS calculation).
So it sounds to me like there are 3 options on DPS calculation:
Very simple take damage, add “+ % damage type”, divide by animation speed. This is what most games do like PoE, Diablo, etc.
Super-complicated perfect calculation of every effect. Which is pretty much impossible to do without running some sort of full combat simulation in memory.
Hybrid approach where some effects are counted towards DPS (like % chance to ignite, + crit rate, etc), while others don’t (gambler’s fallacy, other complex effects). Which is what currently happens in LE and is IMHO very confusing.
Using Torch of the Pontifex makes my Smite tooltip DPS jump from 300 to 876
I think this is exact same “feature”. The DPS is counted not based on a “single hit”, but kind of like “if you were continuously hitting an enemy and triggering all the expected effects with respective proc chances - how much on damage would you inflict per second on average”.
And again, IMHO, this is overcomplicating things a lot.