Damage reduction is not additive, it is multiplicative (therefore having diminishing returns). I'm not sure where you got your info, but I have tested this stuff a lot on Ascension and it's 100% got diminishing returns - Dispersion with 10% dmg reduction does not reduce your damage by 100%, but by 91%.

What does this mean? It means that smaller incremental amounts of damage reduction are considerably worse as you gain dmg reduction. For example: 2x 6% reduction talents nets you only 8.8% reduction vs a single talent 12% which will you 12%. So in your guide, taking spell warding is much better than taking guarded by the light and frozen core, even though it is the same cost.

I do not know how some talents interact with this, for example, does demonic embrace and soul link count as 12% or 2x 6%? Depending on which one really determines whether demonic embrace is a worthwhile talent...

"Damage reduction is not additive, it is multiplicative (therefore having diminishing returns). I'm not sure where you got your info, but I have tested this stuff a lot on Ascension and it's 100% got diminishing returns - Dispersion with 10% dmg reduction does not reduce your damage by 100%, but by 91%."

As you mentioned, yes that is correct! Which is also mentioned beside the graph with the example results

In regards to the last thing you mention, I'm on the same boat, which is why the additive math isn't exactly correct and would take further testing to see its true value

Sorry I mistyped last time: 2 x 6% is 11.6% vs. straight 12% - dunno where I pulled that 8.8% from xD

Yeah I think we agree but your paragraph next to the graph above is incorrect as it is calculating the opposite of diminishing returns. This would be the corrected version:

The effects of percentage based damage reduction scale logarithmic-ly, the more you have the less valuable additional damage reduction becomes. For example, lets say someone is hitting you for 100 damage, if you have 0% damage reduction and you add 1% that 100 damage is reduced to 99 damage, a 1% effective reduction. However if you already have 90% damage reduction and you add another 1% that 100 base damage which was already reduced to 10 is now further reduced to 9.9. That change in incoming damage from 10 to 9.9 is a 0.1% reduction in actual damage taken by adding this extra 1% of damage reduction.

Looking at this more in-depth, this effect explains why picking up defensive stance is a terrible choice when trying to maximise damage reduction. Defensive stance offers 10% dmg reduction for only 2AE at the cost of 10% healing and damage. If you're sitting at 50% damage reduction, then this extra 10% only bumps you up to 55%. So the relative loss of 10% healing and damage (which is additive) is a large gap and really not worth it.

The more damage reduction you have, the less effective each te spent. At 0%, a te can net you 2%, at 50% a te will only net you 1%, and at 25% each te will only give you 0.5% dmg reduction. As a player, you effectively have to decide how valuable each point of te is to you. A general rule of thumb in ascension is if a te gives you <2% of stats, it's likely a terrible pick.

Talents like Improved Blink, Ardent Defender and Nether Protection, are still good, although you're not going to have 100% uptime on the damage reduction, most of the time these sorts of talents will proc when you need them most.

Damage reduction is not additive, it is multiplicative (therefore having diminishing returns). I'm not sure where you got your info, but I have tested this stuff a lot on Ascension and it's 100% got diminishing returns - Dispersion with 10% dmg reduction does not reduce your damage by 100%, but by 91%.

ReplyDeleteWhat does this mean? It means that smaller incremental amounts of damage reduction are considerably worse as you gain dmg reduction. For example: 2x 6% reduction talents nets you only 8.8% reduction vs a single talent 12% which will you 12%. So in your guide, taking spell warding is much better than taking guarded by the light and frozen core, even though it is the same cost.

I do not know how some talents interact with this, for example, does demonic embrace and soul link count as 12% or 2x 6%? Depending on which one really determines whether demonic embrace is a worthwhile talent...

"Damage reduction is not additive, it is multiplicative (therefore having diminishing returns). I'm not sure where you got your info, but I have tested this stuff a lot on Ascension and it's 100% got diminishing returns - Dispersion with 10% dmg reduction does not reduce your damage by 100%, but by 91%."

DeleteAs you mentioned, yes that is correct! Which is also mentioned beside the graph with the example results

In regards to the last thing you mention, I'm on the same boat, which is why the additive math isn't exactly correct and would take further testing to see its true value

Sorry I mistyped last time: 2 x 6% is 11.6% vs. straight 12% - dunno where I pulled that 8.8% from xD

DeleteYeah I think we agree but your paragraph next to the graph above is incorrect as it is calculating the opposite of diminishing returns. This would be the corrected version:

The effects of percentage based damage reduction scale logarithmic-ly, the more you have the less valuable additional damage reduction becomes. For example, lets say someone is hitting you for 100 damage, if you have 0% damage reduction and you add 1% that 100 damage is reduced to 99 damage, a 1% effective reduction. However if you already have 90% damage reduction and you add another 1% that 100 base damage which was already reduced to 10 is now further reduced to 9.9. That change in incoming damage from 10 to 9.9 is a 0.1% reduction in actual damage taken by adding this extra 1% of damage reduction.

Looking at this more in-depth, this effect explains why picking up defensive stance is a terrible choice when trying to maximise damage reduction. Defensive stance offers 10% dmg reduction for only 2AE at the cost of 10% healing and damage. If you're sitting at 50% damage reduction, then this extra 10% only bumps you up to 55%. So the relative loss of 10% healing and damage (which is additive) is a large gap and really not worth it.

The more damage reduction you have, the less effective each te spent. At 0%, a te can net you 2%, at 50% a te will only net you 1%, and at 25% each te will only give you 0.5% dmg reduction. As a player, you effectively have to decide how valuable each point of te is to you. A general rule of thumb in ascension is if a te gives you <2% of stats, it's likely a terrible pick.

Talents like Improved Blink, Ardent Defender and Nether Protection, are still good, although you're not going to have 100% uptime on the damage reduction, most of the time these sorts of talents will proc when you need them most.

Do more pvp centric stuff! This page was really helpful!

ReplyDelete